Geomorphology of debris flows and alluvial fans in Grand Canyon National Park and their influence on the Colorado River below Glen Canyon Dam, Arizona

Item Type Dissertation-Reproduction (electronic); text

Authors Melis, Theodore S.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 01/10/2021 07:54:56

Link to Item http://hdl.handle.net/10150/191214 GEOMORPHOLOGY OF DEBRIS FLOWS AND ALLUVIAL FANS IN GRAND CANYON NATIONAL PARK AND THEIR INFLUENCE ON THE COLORADO RIVER BELOW GLEN CANYON DAM, ARIZONA

by

Theodore Steven Melis

Copyright 0 Theodore Steven Melis 1997

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF GEOSCIENCES

In Partial Fulfillment of the Requirements For the Degree of DOCTOR OF PHILOSOPHY IN SCIENCE In the Graduate College THE UNIVERSITY OF ARIZONA

1997 2

THE UNIVERSITY OF ARIZONA c) GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have

read the dissertation prepared by Theodore Steven Melis

entitled Geomorphology of Debris Flows and Alluvial Fans in Grand Canyon National Park and Their Influence on the Colorado

River Below Glen Canyon Dam, Arizona

and recommend that it be accepted as fulfilling the dissertation

requirement for t - Degree of Doctor of Philosophy

7,3)(7/ //4?' Date

7 /174 1-c-4 Date

Cuilfv 3/3I/ Date

Kathe ne K. chboeck Date 2,6k; 8 97 Simon Ince Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I herebycertify, that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation req ire ént Lf(t7(17 Dissertation Director Victor R. Baker Date 3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special peimission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the copyright holder.

SIGNED: 4

ACKNOWLEDGMENTS

Where do I begin? First, I need to thank the many individuals who helped with the field and office work that made this report possible. The list of persons I need to thank is long, but not so long that each one cannot be mentioned here; having paid my page charges in years of tuition fees. The professionalism of the numerous guides who piloted boats during the project on the Colorado River made the field work efficient, safe, and fun. These guides included: Tony Anderson, Jeff Arronson, Jane Bernard, Steve Bledsoe, Gary Bolton, Tom Brownold, Kelly Burke, Michael Collier, Tim and KC Deutschlander, Brian Dierker, Dave Edwards, Kenton Grua, Chris Geneous, Bob Grusey, Tom Hanson, Matt Herman, Les Hibbert, Jon Hirsch, Liz Hymans, Mark Jensen, Jan Kempster, Cindy Krzniack, Zeke Louck, Michael Lowery, Tom Moody, Nelbert Niemi, Lars Niemi, Pete Reznick, Suzanne Rhodes, Glenn Rink, Rachael Running, Dennis Silva, Drifter Smith, Jeanette Smith, Kelly Smith, Paul Swanstrom, Al Tinnes, John Toner, John and Suzette Weisheit, Greg Williams, Tom Wise and Mike Yard.

I was ably assisted in my field work by numerous individuals, including Jim Boyer, Clay Bravo, Bryan Brown, Rondo Buecheler, Kelly Burke, Nathan Carlton, Jeff Clark, Todd Colletti, Jim Courson, Steve Cox, Dar Crammond, John Elliott, Steve Eudaley, Cassie Fenton, Juan Martinez-Goytre, Paul Grams, Peter Griffiths, Mike and Terry Grijalva, Martha Hahn, Lauren Hammack, Joe Hazel, Ralph Hopkins, Shirlee Jensen, Don Jensen, Matt Kaplinski, Jan Kempster, Karel Malloy, Lisa Michaels, Eric Martin, Tom Martin, Hilary Mayes, Greg McCormick, Travis McGrath, Lizette Melis, Betsy Menand, Carlos Mendoza, Holly Metz, Kip Metzger, Johnny Moore, Cynta de Narvaez, Joel Pederson, Elise Pendall, Bill Phillips, Betsy Pierson, Dave Pratt, Sarah Rathburn, Karen Reichardt, Gordon Rodda, Reneé Rondeau, Dave Rubin, Jim Pizzuto, Tim Randle, John Rhis, Robert Ruffner, Mauricio Schabes, Jack Schmidt, Christine Smith, Greg Spoonenberg, Doug Stephens, Larry Stevens, Deborah Tosline, Justin Turner, Judy Wiess, Toni Yocum, Bob Webb, Tamara Wiggins, Betsy Wirt, and Stephanie Yard. Special thanks also go to Janice Bowers, Diane Grua, Mia Wise, Jim Hasbargen, Mimi Murov, Glenn Rink, Reneé Rondeau, Ray Turner, and Steve Wiele for their hard work and devotion of months of time to field work on the Colorado River in Grand Canyon. Mimi Murov was an especially important field assistant, not only through her actions, but also because of her ceaseless curiosity and desire to understand the processes we all studied. In attempting to answer her insightful questions, I often realized how little I really understood about the subtleties of Grand Canyon. Ed Holroyd, of the Bureau of Reclamation, Denver, Colorado, supplied the digital elevation models used for estimation of tributary drainage area early on in the project. Hilary Mayes and Peter Griffiths were pivotal in producing the color plates showing drainage areas, study sites and debris-flow frequency included here.

Bob Webb was instrumental in introducing me to the topic of Grand Canyon debris flows and their effects on the Colorado River below Glen Canyon Dam. He challenged me to attack the varied and complex problems of documenting and interpreting debris-flow data in all the nooks and crannies of Grand Canyon; most of which didn't even have names. Accepting the challenge, I began a seven year long journey through a maze of places, people and ideas that have not yet ceased to amaze me. In addition to the professional satisfaction of participating in these studies, I retain the memories of numerous river trips 5 through Grand Canyon with people who, intentionally or not, changed me forever.

Much of this research involved repeat photography. My trip to Glen Canyon with Ray Turner and Bob Webb in December, 1989, changed my perceptions of how important photographic archives can be to natural scientists. One of my most memorable field experiences during this project was participating in the 34-day long Robert Brewster Stanton Centennial Rephotography Trip in early 1990. I thank Bob Webb for including me in the crew roster of what must stand as one of the best river trips ever. Photographers who helped with the repeat photography on that first trip and many afterwards include Jane Bernard, Gary Bolton, Tom Brownold, Michael Collier, Dave Edwards, Jim Hasbargen, Liz Hymans, Ralph Hopkins, Steve Kimsey, Sara Light-Waller, Dominic Oldershaw, Steve Tharnstrom, and Ray Turner. Sara Light-Waller, Dominic Oldershaw and Dave Ring performed much of the darkroom work necessary for the large-scale repeat photography work; I extend many thanks for their patience and hard work. Chuck Sternberg produced the line illustrations found throughout the manuscript, and Hilary Mayes produced the many color plates located at the end of this volume; thank you both.

Thanks also to Vic Baker, Jim Bennett, Constance McCabe, Yehouda Enzel, Dick Iverson, Waite Osterkamp, Tom Pierson, Steve Reneau, and Kevin Scott for their review of the research design and field work in April, 1991. My perspectives on debris flows also benefitted greatly by field experiences during a 1994 visit to the Peoples Republic of China, and I thank Jon Major for making that trip possible for me. I also want thank Dick Iverson, Jon Major and Rick LaHusen for valuable insights gained into debris-flow behavior during a 1995 visit to the U.S. Geological Survey experimental debris-flow flume, near Blue, Oregon.

A particularly important event that occurred during the course of this project was the spring, 1996, Glen Canyon Dam Beach/Habitat-Building Flow. For the first time ever, a large dam on the Colorado River was used to flood a river under controlled conditions while scientists measured the benefit to downstream ecosytem resources. I was fortunate in having an opportunity to be integral in the planning and implementation of that experiment, and I am forever grateful to those persons who allowed me to be involved in that important event; one which was truely a hydrologist/geologist dream come true.

Finally, I especially thank David L. Wegner of the Bureau of Reclamation's Glen Canyon Environmental Studies for his unwavering support of this research while waging a fourteen-year long environmental and bureaucratic war against the powers-that-be; a war which he had ultimately both won and lost by the close of 1996. 6

DEDICATION

This volume is dedicated to my family, all of whom have supported me in my education and pursuit of a career in the geosciences. To my wife, Lizette, who never wavered in her support of my Grand Canyon studies, despite my many prolonged absences away in the field. To my parents, Elizabeth and Theo, who have always been there to encourage me while I pursued one long-range educational goal after another. To Bush,

Mauricio, Lisa, Kim, Chip, Judy, Fanny and Willy who never stopped asking me questions like -- "so what exactly is it that you are trying to do down there?" And finally, I dedicate this work to our soon-to-arrive children; may the wonder of Gemini be with you both always (see the ring). If there was ever a good reason to complete this work, and complete it on time, then you two were it. 7

TABLE OF CONTENTS LIST OF FIGURES 11 LIST OF TABLES 15 ABSTRACT 17 CHAPTER 1- INTRODUCTION 19 Historical overview 19 What are debris flows? 20 Relevance of the study 23 Debris flows and the Colorado River 25 Components of the study 30 Purpose and scope of the study 31 The Grand Canyon study area 35 Units and place names 36

CHAPTER 2- BACKGROUND AND PREVIOUS WORK 38 Earliest Grand Canyon Studies 38 Earliest eyewitness account of a Grand Canyon debris flow 43 Earliest published accounts of highly-sediment-laden flows 47 Conceptual framework for later Grand Canyon Studies 52 Renewed interest in Grand Canyon geology 56 Recent studies of rapids and debris flows in Grand Canyon 60 Era of environmental research on the Lower Colorado River 65

CHAPTER 3- FREQUENCY, MAGNITUDE, AND INITIATION OF HISTORIC GRAND CANYON DEBRIS FLOWS 73 Abstract 73 Introduction 74 Previous related work 80 Historic Grand Canyon debris flows 81 Methods 84 Designating drainage areas 84 Evidence for Grand Canyon debris flows and associated flow regimes. 86 Documenting debris-flow frequency, magnitude, and initiation 87 Results 105 Geomorphically-significant tributaries of the Colorado River 105 Debris-flow frequency 108 Debris-flow magnitude 112 Source areas, initiation mechanisms, clay mineralogy, and precipitation 125 Discussion and conclusions 145

CHAPTER 4 - DISCHARGES, FLOW TRANSFORMATIONS AND HYDROGRAPHS 153 8

TABLE OF CONTENTS, Continued

Abstract 153 Introduction 154 Methods 154 Physical characteristics of debris flows and fans 157 Peak discharge and volume 157 Particle-size distributions and water content 159 Fan morphology and stratigraphy 161 Hypothetical hydrographs of debris flows 164 Discussion and conclusions 168

CHAPTER 5- DEBRIS FLOWS AND ALLUVIAL FANS: THEIR INFLUENCE ON THE REACH-VARIED GEOMORPHOLOGY OF THE COLORADO RIVER IN GRAND CANYON, ARIZONA 171 Abstract 171 Introduction 172 Previous work 176 Conceptual models of fan evolution in Grand Canyon 176 Re-examining previously-designated geomorphic reaches 183 Objective 188 Methods 189 Results 194 Evaluating the geomorphic reaches of Schmidt and Graf (1990) . 194 The basis for newly-proposed geomorphic reaches 197 The influence of fan-eddy complexes on fine-sediment resources 205 Discussion and Conclusions 220 Implications of channel-reach variability on riverine resources 224

CHAPTER 6- REWORKING OF AGGRADED COLORADO RIVER DEBRIS FANS BY THE 1996 CONTROLLED FLOOD BELOW GLEN CANYON DAM, ARIZONA 228 Abstract 228 Introduction 229 Previous and related work 233 Methods 236 Debris-fan study sites 236 Documenting debris-fan reworking 239 Results 244 Hydrographs of the 1996 controlled flood 244 Relation between the hydrograph and reworking at Lava Falls 245 Areas and volumes of debris-fan deposits 249 Armoring of the distal margins of debris-fan deposits 254 Constrictions, velocities, water-surface fall, and stream power 258 Transport of marked particles 261 Discussion and conclusions 269 9

TABLE OF CONTENTS, Continued

CHAPTER 7- SUMMARY AND CONCLUSIONS 281 Summary of key results 289 Implications for management of river resources below Glen Canyon Dam . . . 292 A river restored? 294

APPENDIX A - Locations of observed streamflow floods and debris flows in Grand Canyon 297

APPENDIX B - List of geomorphically-significant tributary canyons of the Colorado River in Grand Canyon that produce debris flows 305

APPENDIX C - Additional drainage-basin characteristics for 529 geomorphically-significant tributaries of the Colorado River in Grand Canyon 331

APPENDIX D - List of photographic views of Glen, Marble and Grand Canyons made by Franklin A. Nims and Robert B. Stanton between July, 1889 and March, 1890 358

APPENDIX E - List of original photographic views made by Eugene C. LaRue, and replicates made between Glen Canyon Dam and Lake Mead 384

APPENDIX F - List of photographic views made by other photographers of the Colorado River corridor in Grand Canyon that were replicated in this study 397

APPENDIX G - Characteristics of boulders transported by recent debris flows in tributaries of Marble and Grand Canyons 422

APPENDIX H - Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona 432

APPENDIX I - Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon 455

PLATE 1 - Geomorphic Reach 1: River Miles 0.0 to 11.3 462

PLATE 2 - Geomorphic Reach 2: River Miles 11.3 to 22.6 463

PLATE 3 - Geomorphic Reach 3: River Miles 22.6 to 35.9 464

PLATE 4 - Geomorphic Reach 4: River Miles 35.9 to 61.5 465

PLATE 5 - Geomorphic Reach 5: River Miles 61.5 to 77.4 466

PLATE 6 - Geomorphic Reach 6: River Miles 77.4 to 115.7 467 10

TABLE OF CONTENTS, Continued

PLATE 7 - Geomorphic Reach 7: River Miles 115.7 to 125.5 468

PLATE 8 - Geomorphic Reach 8: River Miles 125.5 to 140.0 469

PLATE 9 - Geomorphic Reach 9: River Miles 140.0 to 160.0 470

PLATE 10 - Geomorphic Reach 10: River Miles 160.0 to 213.9 471

PLATE 11 - Geomorphic Reach 11: River Miles 213.9 to 225.8 472

PLATE 12 - Colorado River Tributaries Within Grand Canyon National Park and Vicinity, Arizona 473

PLATE 13 - Drainage Areas that Experienced Debris Flows Between 1872 and 1996, Colorado River: Lees Ferry to Diamond Creek 474

PLATE 14 - Probability of Debris Flows for Colorado River Tributaries: Lees Ferry to Separation Canyon 475

REFERENCES 476 11

LIST OF FIGURES

FIGURE 1.1, Location map of the Colorado River through Grand Canyon National Park 21

FIGURE 3.1, The morphology of a typical debris fan and eddy complex along the Colorado River in Grand Canyon 77

FIGURE 3.2, Stratigraphic column of rocks exposed in Grand Canyon 83

FIGURE 3.3, Typical channel bend where superelevated debris-flow evidence is often found 93

FIGURE 3.4, Longitudinal variation of a typical flow-surface profile found in a superelevated channel bend in Grand Canyon 96

FIGURE 3.5, Example of replicated photographs showing typical impacts to the Colorado River from tributary debris flows 109

FIGURE 3.6, The 137Cs activities of historic debris-flow deposits sampled in Grand Canyon 113

FIGURE 3.7, Peak discharge versus drainage area for recent debris flows associated with different storm types 114

FIGURE 3.8, Peak debris-flow velocity versus channel gradient for debris flows . 117

FIGURE 3.9, Debris-fan volume versus drainage area for recent debris flows associated with different stolin types 119

FIGURE 3.10, Peak discharge versus debris-fan volume for recent debris flows in Grand Canyon 121

FIGURE 3.11, Peak debris-flow discharge versus sand content for recent debris flows 123

FIGURE 3.12, Percentage of each failure type that has initiated recent debris flows in Grand Canyon 128

FIGURE 3.13, Relations between intense daily precipitation and station characteristics near Grand Canyon 132

FIGURE 3.14, Graphical long-term trends in intense daily precipitation at four stations near Grand Canyon 134

FIGURE 3.15, Timing versus magnitude of recent debris flows in Grand Canyon . . 138 12

LIST OF FIGURES, Continued

FIGURE 3.16, Standardized time series for seasonal precipitation at stations near Grand Canyon during the twentieth century 140

FIGURE 4.1, Location map showing recent debris-flow study sites throughout Grand Canyon 155

FIGURE 4.2, Relations between peak discharge and debris-fan volume versus tributary drainage areas for recent debris flows 158

FIGURE 4.3, Particle-size distributions from the surfaces of "simple" and "complex" debris fans recently deposited in Grand Canyon 160

FIGURE 4.4, Plan views of three characteristic debris-fan types associated with recent debris flows in Grand Canyon 162

FIGURE 4.5, Particle-size distributions of initial, dam-release reworked, and pre-dam flood reworked debris-fan surfaces found in Grand Canyon 165

FIGURE 4.6, Idealized Type I, II, and III hydrographs that describe recent tributary debris flows in Grand Canyon 166

FIGURE 5.1, Location map of the Colorado River through Grand Canyon showing the locations of major debris flows that occurred in the twentieth century 173

FIGURE 5.2, Replicate aerial photographs showing a typical "fan-eddy complex," at river mile 62.6-R before and after being impacted by a debris flow in 1990. 178

FIGURE 5.3, Source lithology and particle-size distributions of sediment deposits at the mouth of tributary 62.6-R, and the cobble bar located immediately downstream 179

FIGURE 5.4, Reach-averaged fan-eddy complex attributes for eleven geomorphic reaches described by Schmidt and Graf (1990) along the Colorado River in Grand Canyon 196

FIGURE 5.5, Lowess smoothed-curve plots of fan-eddy complex attributes between river miles 0 and 225 of the Colorado River in Grand Canyon 198

FIGURE 5.6, Reach-averaged fan-eddy complex attributes for revised geomorphic reaches of the Colorado River in Grand Canyon 204 13

LIST OF FIGURES, Continued

FIGURE 5.7, Average percentages of tributaries with alluvial fans for previous and revised geomorphic reaches of the Colorado River in Grand Canyon 206

FIGURE 5.8, Numbers of alluvial fans and rapids associated with fan-eddy complexes along the Colorado River in Grand Canyon 208

FIGURE 5.9, Statistical distribution of local mainstem constrictions caused by alluvial fans along the Colorado River in Grand Canyon 210

FIGURE 5.10, Source lithologies of clasts on 118 reworked fans sampled along the Colorado River within revised geomorphic reaches 211

FIGURE 5.11, Alluvial-fan height (with percent-boulders), and dominant lithologie sources for reworked fans in fan-eddy complexes along the Colorado River 213

FIGURE 5.12, Lithologie diversity of clasts on reworked alluvial fans of the Colorado River between Lees Ferry and Diamond Creek in Grand Canyon 215

FIGURE 5.13, Relative percentages of shoreline-types found along the Colorado River between Lees Ferry and Diamond Creek in Grand Canyon . . . 219

FIGURE 6.1, Map of the Colorado River in Grand Canyon showing the locations of debris fans 231

FIGURE 6.2, The morphology of a typical debris fan and eddy complex along the Colorado River in Grand Canyon 232

FIGURE 6.3, Graph of the annual flood series for the Colorado River near Grand Canyon, Arizona 234

FIGURE 6.4, Hydrographs of selected Colorado River floods and the 1996 controlled flood 246

FIGURE 6.5, Aerial photographs of Lava Falls Rapid before and after the 1996 controlled flood 247

FIGURE 6.6, Hydrographs showing the relation between stage, discharge, and reworking at Lava Falls Rapid during the 1996 controlled flood . . . . 248

FIGURE 6.7, Replicate photographs of the debris fan at 127.6 Mile 250 14

LIST OF FIGURES, Continued

FIGURE 6.8, Particle-size distributions of pristine debris flows, reworking by combined powerplant releases and tributary inflows, reworking by the 1996 controlled flood, and armoring from pre-dam floods in Grand Canyon 255

FIGURE 6.9, Graphs of particle-size distributions and lithologies at Lava Falls Rapid before and after the 1996 controlled flood 256

FIGURE 6.10, Graphs of water-surface profiles through selected rapids reworked by the 1996 controlled flood 262

FIGURE 6.11, Graphs showing the presence of marked particles of different sizes on four debris fans before and after the 1996 controlled flood 264

FIGURE 6.12, Graphs showing the presence of marked particles placed on the Palisades Creek debris fan 267

FIGURE 6.13, Graph showing the relation between streampower, date of debris flows, and changes in area and volume of aggraded debris fans reworked by the 1996 controlled flood in Grand Canyon 273

FIGURE 6.14, Graph of changes in percent constrictions of debris fans in Grand Canyon reworked by the 1996 controlled flood 274

FIGURE 6.15, Schematic diagram showing reworking of debris fans by the Colorado River 279

FIGURE 7.1, Spatial relationships between debris-flow probability, river geomorphology, and other important riverine resources of the Colorado River, between river miles 0 and 225 295 15

LIST OF TABLES

TABLE 3.1, Summary of tributary drainage-area statistics in Grand Canyon . . . . 107

TABLE 3.2, List of major historic debris flows that have impacted the Colorado River 110

TABLE 3.3, Debris-flow peak discharge estimates from selected study sites 116

TABLE 3.4, Volumes of recently-aggraded debris fans in Grand Canyon 118

TABLE 3.5, Percent sand content of selected recent debris flows 122

TABLE 3.6, Summary of clay mineralogies contained in recent debris-flow deposits 129

TABLE 3.7, Results of statistical analyses on long-term trends in intense daily precipitation at sites near Grand Canyon 137

TABLE 3.8, Timing and volumes of recent debris flows relative to El Nitio/La Nina Events 144

TABLE 5.1, Previously designated geomorphic reaches of the Colorado River through Grand Canyon, based mainly on occurrence of bedrock lithologies at river level 187

TABLE 5.2, Reach-averaged channel characteristics for previously designated geomorphic reaches of the Colorado River through Grand Canyon, based on measurements of fan-eddy complexes at 529 tributary confluences 195

TABLE 5.3, Revised geomorphic reaches and subreaches of the Colorado River through Grand Canyon, based on fan-eddy characteristics measured near 529 tributary confluences 202

TABLE 5.4, Shoreline characteristics, by river mile (0 to 225), associated with revised geomorphic reaches of the Colorado River through Grand Canyon 217

TABLE 6.1, Characteristics of debris fans aggraded after 1983 in Grand Canyon 237

TABLE 6.2, Changes in the areas of debris fans reworked by the 1996 controlled flood in Grand Canyon 251

TABLE 6.3, Changes in the volumes of debris fans reworked by the 1996 controlled flood in Grand Canyon 253 16

LIST OF TABLES, Continued

TABLE 6.4, Changes in particle-size distributions of debris fans reworked by the 1996 controlled flood in Grand Canyon 257

TABLE 6.5, Changes in the constrictions of the Colorado River by debris fans reworked by the 1996 controlled flood in Grand Canyon 259

TABLE 6.6, Velocities measured in rapids and riffles during the 1996 controlled flood in Grand Canyon 260

TABLE 6.7, Changes in water-surface profiles and stream power around selected debris fans reworked by the 1996 controlled flood in Grand Canyon 263

TABLE 6.8, Movement of marked boulders from five debris fans during the 1996 controlled flood. 265 17

ABSTRACT

Debris flows in at least 529 Grand Canyon tributaries transport poorly-sorted clay- to boulder-sized sediment into the Colorado River, and are initiated by failures in weathered bedrock, the "fire-hose effect," and classic soil-slips often following periods of intense rainfall coincident with multi-day storms. Recent debris flows had peak-discharges from

about 100-300 m3/s. Twentieth-century debris flows occurred from once every 10-15

years in eastern tributaries, to once in over a century in western drainage areas. System-

wide, debris flows likely recur about every 30-50 years, and the largest recent flows were

initiated during Pacific-Ocean storms in autumn and winter.

Three idealized hydrographs are inferred for recent debris flows based on deposits

and flow evidence: Type I, has a single debris-flow peak followed by a decayed recessional

streamflow; Type II, has multiple, decreasing debris-flow peaks with intervening flow

transformations between debris flow and non-debris flow phases; and Type III, may have

either a simple or complex debris-flow phase (begin as either Type I or II), followed by a

larger streamflow peak that reworks or buries debris-flow deposits under streamflow

gravel deposits.

From 1987 through 1995, at least 25 debris flows constricted the Colorado River,

creating 2 rapids and enlarging at least 9 riffles or rapids. In March-April, 1996,

reworking effects of a 7-day controlled flood release (peak = 1,300 m3/s) on 18 aggraded

debris fans in Grand Canyon were studied. Large changes occurred at the most-recent

deposits (1994-1995), but several other older deposits (1987-1993) changed little. On the

most-recent fan deposits, distal margins became armored with cobbles and boulders, while

river constriction, flow velocity, and streampower were decreased. Partial armoring of fan

margins by relatively-low mainstem flows since the debris flows occurred, was an 18 important factor limiting fan reworking because particles became interlocked and imbricated, allowing them to resist transport during the flood. Similar future floods will accomplish variable degrees of fan reworking, depending on the extent that matrix- supported sediments are winnowed by preceding mainstem flows. 19

CHAPTER 1

INTRODUCTION

Historical overview

John Wesley Powell pioneered scientific exploration of the Colorado River through

Grand Canyon during river expeditions in 1869 and 1871-72 (Powell 1895). During those trips Powell marveled at the scale of erosion evidenced by hundreds of tributary canyons along the river; many of them over 1.5 kilometers (km) deep. Although he and his crew faced many hardships during river exploration, none surpassed those encountered while navigating the Colorado River's large rapids. Powell observed the spatial relationship between the rapids and tributaries throughout Grand Canyon, concluding that the whitewater obstacles were formed by large boulders deposited in the river during tributary floods (Powell 1895). Though correct in his conclusion, Powell lacked the firsthand experience needed to recognize tributary debris flows as the key sediment-transport agent creating the Colorado River's most dramatic features -- its rapids. Remarkably, some later workers refuted the nearly one-to-one spatial relationship between tributaries and rapids in

Grand Canyon (Leopold 1969).

In the 1870s, Grove Karl Gilbert observed that the landscape of the Colorado

Plateau is dominated by erosion. Set in the semi-arid climate of the southwestern United

States (U.S.), the deeply incised Colorado Plateau covers large areas of New Mexico,

Colorado, Arizona, and Utah and features its most renowned landform -- Grand Canyon

(fig. 1.1). During the early period of geologic studies in the U.S., Gilbert's insightful writing on Grand Canyon included:

To the student of erosion are exhibited the most distinguished monuments of its action; and he is given an opportunity to partially isolate certain of the 20

conditions which control the rapidity of erosive action, by viewing their influence where that influence is at a maximum (Gilbert 1876).

Gilbert's statement captures the strong allure Grand Canyon has for process- minded geomorphologists. Of the many erosional processes at work on the Colorado

Plateau and in Grand Canyon, the hillslope process of debris flow is one of the most interesting and important, despite its relatively infrequent occurrence. Although debris- flow deposits abound within Grand Canyon tributaries and along the Lower Colorado

River, many "students of erosion" since Gilbert have failed to recognize the role that debris flows play in shaping the geomorphic framework of the Colorado River.

Nowhere else on the Colorado Plateau are debris-flows more important than along the now-regulated Lower Colorado River in Grand Canyon National Park. This unique mass-wasting process not only creates and maintains the river's spectacular whitewater, but is also key in eroding Grand Canyon's tributaries deep into bedrock. When tributary debris flows reach the mainstem, they impact the riverine resources of the Colorado River below Glen Canyon Dam (fig. 1.1), and modify the channel. Understanding the role that debris flows play in controlling the Colorado River is more important now than ever, owing to the effects of regulated flows imposed by large dams. This is especially true below Glen Canyon Dam, where debris flows dominate the form and hydraulics of the channel most dramatically. Debris-flow impacts on the Colorado River below this large dam are the main focus of this report.

What are debris flows?

Debris flows are poorly-sorted mixtures of water, mud and rocks that normally contain less than 30-to 40 percent water, and often contain sediment from clay-to boulder 21

FIGURE 1.1. The Colorado River through Grand Canyon National Park and vicinity,

Arizona, with debris-flow study sites (from Melis et al. 1994). 22 size (see: Griffiths et al. 1996; Hammack and Who! 1996; Webb et al. 1996; Blair and

McPherson 1994; Hammack 1994; Melis et al. 1994; Melis and Webb 1993; Bull 1991;

Webb et al. 1989; Costa 1984; Johnson and Rodine 1984; Cooley et al. 1977; Fisher 1975;

Blackwelder 1928; Rickmers 1913). Debris flows commonly initiate in steep,

unconsolidated hillslope sediments following intense or prolonged rainfalls (Costa 1984).

They may also occur following bedrock failures in cliffs (Griffiths et al. 1996), and are

commonly associated with volcanism where ash and water are abundant during eruptions

(Pierson et al. 1990). Once liquified, the poorly-sorted sediment originating from slope

failures is usually transported through existing channels and may travel long distances at

varying velocities (Rodine and Johnson 1976).

Sedimentary deposits resulting from debris flows are unique, easily identified once

recognized, and often occur as channel-margin levees and boulder lines oriented sub-

parallel to flow on alluvial fans (Costa 1984; Johnson and Rodine 1984). The deposits are

normally very-poorly sorted and devoid of sedimentary structures, with coarser clasts

supported by a clay to-pebble sized matrix (Costa 1984). Alluvial fans are distinctive

landforms that usually represent the cumulative deposition of many debris flows over time,

and are found in arid to-alpine geomorphic environments throughout the world (Blair and

McPherson 1994; Costa 1984; Bull 1977;1963; Sharp 1942; Blackwelder 1928; Rickmers

1913). Most alluvial fans in Grand Canyon differ in overall form from classic alluvial fans

found in other semi-arid to-arid environments. River-reworked Grand Canyon fans are fundamentally different because they highly constrained in size by the narrow valley of the

Colorado River, and composed mostly of boulders rather than poorly-sorted sediment

(Melis et al. 1994; Melis and Webb 1993). The distinctive coarsened texture of most fans

along the mainstem reflects reworking and grading of original debris-flow deposits by 23 streamflow floods on the mainstem that occurred before regulation (Melis and Webb 1993).

Such reworking of modern debris-flow deposits still occurs following regulation, but to a much reduced degree (Webb et al. 1996; Webb and Melis 1995; Melis and Webb 1993;

Graf 1980).

Debris flows were recognized as an important geomorphic process by some geologists during the early 20th century (Sharp 1942; Sharpe 1938; Blackwelder 1928;

Rickmers 1913), but most workers remained unaware of their importance until relatively recently (Blair and McPherson 1994). Debris flows are most renowned for their abilities to transport large volumes of sediment, and carry very large boulders over long distances through low-gradient channels (Beatty 1989; Costa 1984; Rodine and Johnson 1976).

Transport and deposition of very large boulders in non-glaciated environments created dilemmas for many early geologists who depended solely on fluvial processes for their interpretations (see Trowbridge 1911). Today, boulder transport by debris flows is more widely recognized and well documented; especially in deserts (Beatty 1989).

Relevance of the study

Grand Canyon provides an ideal field setting for studying interactions of fluvial and hillslope processes that occur between a large regulated river and its unregulated tributaries.

From available historic and field-derived data on flow and features of the Colorado River before closure of Glen Canyon Dam in 1963, insights are gained on how these interactions

occurred in the pre- and post-dam eras. The unregulated Colorado River in Grand Canyon

was formerly the place where poorly-sorted sediment delivered periodically by tributary

debris flows was regularly reworked and graded by mainstem floods. That natural

interplay between the mainstem and its tributaries was disrupted by reduced annual peak 24 flows from Glen Canyon Dam after 1963, however, little information on these process interactions was available until recently. Regulation now allows debris flows from unregulated, ephemeral, bedrock Grand Canyon tributaries to exert greater control on the

Colorado River's geomorphology and ecosystems than occurred before construction of

Glen Canyon Dam. These impacts are causing changes to the river and its ecosystem that are still only partly understood.

The debris-flow dominated framework of the river through Grand Canyon is relatively unique (Webb et al. 1988; Schmidt and Rubin 1995), and contrasts with other most other bedrock-controlled rivers not impacted by debris flows. Debris flows continue to erode Grand Canyon tributaries despite regulated mainstem conditions, (Hereford and

Huntoon 1990), but sediment delivered to the mainstem can no longer be carried from the basin.

The several modes of sediment transport have occurred in Grand Canyon at variable rates owing to changing climate through geologic time. Erosion of Grand Canyon's bedrock tributaries appears to be a somewhat intractable problem at first because present arid to semi-arid conditions in the region result in mostly ephemeral streams. However, debris flows are key in eroding bedrock channels under a wide range of climate conditions

(Howard 1995; Costa 1984). Such conditions, including a ready supply of clay-to boulder-size particles in combination with short but intense rainfalls, are elements readily found in Grand Canyon.

Large boulders up to several meters in diameter are produced during rockfalls when cliff failures occur throughout Grand Canyon (Hereford and Huntoon 1990; Ford et al.

1974), and are carried to the river channel because of unique debris-flow properties

(Rodine and Johnson 1976). Boulders that accumulate in ephemeral tributary channels 25 weather very slowly under the semi-arid conditions in Grand Canyon and greatly exceed the competence of streamfloods there, yet these normally-stable particles are easily transported into the Colorado River by occasional debris flows. Transport of boulders by tributary debris flows in Grand Canyon tributaries is likely one of the most important factors driving vertical incision of these small drainage basins (Howard 1995).

Regulated flows below Glen Canyon Dam now severely limit the Colorado River's competence to rework boulders and finer sediment deposited at hundreds of tributary mouths by debris flows downstream. This fact has only short-term implications for the ongoing evolution of Grand Canyon and the Colorado River because the expected service life of Glen Canyon Dam is short compared with geologic time. However, perturbation of the fluvial processes that formerly enabled the Colorado River to accomplish its work -- significantly-reduced flooding -- complicates management of riverine resources and complex ecosystems below Glen Canyon Dam over time scales of human impacts. Impacts from regulation of the Colorado River will continue for a significant period along some 380 km of riverine habitat between Lakes Powell and Mead, Arizona (fig. 1.1), based on the predicted service life of the dam (— 700 years). Impacts to riverine processes over this time period may be significant for the river's ecosystems and species dependent on habitats below Glen Canyon Dam, as well as for humans who also use these resources.

Debris flows and the Colorado River

The hundreds of reworked, bouldery, debris fans deposited along the Lower

Colorado River in Grand Canyon are of particular interest because they impose local base- level controls on the river channel along at least 380 km (Melis et al. 1994; Melis and Webb

1993; Webb et al. 1988; Cooley et al. 1977). Such locally-derived channel controls create 26 a highly-segmented river profile comprised of alternating pools and rapids/riffles where debris flows encounter the mainstem (Howard and Dolan 1981;1979). The resulting stair-

step form of the channel isolates individual channel segments (pools) from one another at

rapids where stable debris fans act as check dams. This segmentation of the mainstem

channel affects storage of sand, navigation, and characteristics of the river ecosystem, such

as primary productivity and physical-habitat attributes important to native and non-native

species of plants, animals and insects (Valdez and Rye! 1996; Blinn et al. 1995; Melis et al.

1995; Schmidt and Rubin 1995; Melis et al. 1994).

Grand Canyon debris flows are also of special interest because they debouch from

mainly low-order, ephemeral streams, yet impose great control on the channel form and

fluvial processes of a high stream-order perennial river (Webb 1996; Webb et al. 1996;

Webb et al. 1988). Similar relationships between large North American rivers and their

minor tributaries are relatively rare (Webb 1996; Schmidt and Rubin 1995; Melis et al.

1994; Webb et al. 1988). This is true because the two factors that allow debris flows to

dominate the Colorado River -- the immense vertical relief in even the smallest drainage

areas, and the unique combination of shales, limestones and sandstones present in Grand

Canyon stratigraphy (Griffiths et al. 1996; Melis et al. 1994) -- are not common in most

river basins. Before regulation of the river through Grand Canyon, only the largest

boulders deposited in the mainstem were stable during large river floods. However, drastically-reduced annual peak flows now allow finer sediment to accumulate at tributary

confluences and persist there for longer periods following debris flows.

Being a dominant mode of sediment transport in at least 529 tributaries of the

Colorado River in Grand Canyon from Lees Ferry to Diamond Creek, Arizona (fig. 1.1;

river miles 0 to 225; see: Melis et al. 1994; Melis and Webb 1993; Webb et al. 1989;1988), 27 debris flows also provide an important source of sand and finer sediment to the river downstream of Glen Canyon Dam (Schmidt and Graf 1990, Howard and Dolan 1981).

Since 1963, approximately 87 percent of the sediment formerly transported by the mainstem is annually impounded by Lake Powell (Andrews 1991). This fact has

substantially increased the importance of inputs of fine sediment by debris-flows

downstream of Glen Canyon Dam. However, by supplying boulders that exceed the

competence of mainstem discharges, debris flows also maintain large rapids. These

features directly control the navigability and sediment budget of the river downstream of

Lake Powell; a situation that affects over 20,000 whitewater enthusiasts who annually use

the Colorado River for recreation (Stevens 1990).

In terms of the river's sedimentology, alluvial fans constrict the mainstem channel

at hundreds of sites in Grand Canyon, resulting in zones of recirculating flow, termed

eddies, where reduced flow velocities act to trap fine sediments. Eddy-stored sand and silt

is mobilized during high flows and deposited along channel shorelines, forming the

substrate of the aquatic/riparian river ecosystem. Debris flows are also important because

resulting debris fans provide stable foundations in eddies where sand bars deposit and

remain stable long enough for communities of riparian plants and animals to establish

(Meus et al. 1995; Schmidt and Rubin 1995; Stevens et al. 1995; Valdez and Ryel 1995;

Schmidt et al. 1993; Rubin et al. 1990; Schmidt 1990; Schmidt and Graf 1990).

In addition, endangered fishes, such as the native humpback chub (Gila cyphus),

evolved within the eddy-dominated reaches of the Colorado River during the last several

million years. Recent studies suggest that these native fishes rely heavily on the debris-fan

controlled eddies as their primary adult habitat (Valdez and Ryel 1996). Localized impacts

on aquatic habitat, including elevated levels of mainstem aquatic production, have been 28 recently documented following Grand Canyon debris flows (Melis et al. 1995). These impacts are still not fully understood, but are apparently related to the increased constricting effects that debris-flows impose on rapids, leading to enhanced conditions of light penetration promoting primary productivity in the low-velocity pools upstream (Melis et al.

1995).

Sand bars deposited on debris fans are frequently used as campsites by thousands

of river-runners who annually boat on the Colorado River. Fluctuating releases associated

with hydro-powerplant operations at Glen Canyon Dam after 1963 eroded many sand bars

along the Colorado River and disrupted the dynamic equilibrium that formerly existed there

between debris flows, larger tributary inputs of fine sediment and mainstem sand bars

(Webb 1995; Schmidt et al. 1995; Melis et al. 1994; Melis and Webb 1993; Schmidt and

Graf 1990; DOT 1989). The spacing and three-dimensional geometries of debris fans, as

well as the morphology of the mainstem and it tributaries, influence the size, shape and

stability of sand bars along the margins of the channel under varying river stages and

sediment supplies (Schmidt and Rubin 1995; Melis et al. 1994. Because of this fact, a

greater understanding of debris flows and their effects on the formation, evolution and

stability of sand bars (i.e. effects on the geomorphology of the river corridor) is needed to

predict future trends in the sediment mass-balance and related resources in Grand Canyon.

As of 1997, a program termed "Adaptive Management," was implemented on the

Lower Colorado River. One of the many goals of this program is to link what is known

about the river's geomorphology with the biotic components of its ecosystem (see: DOT

1995; Gunderson et al. 1995; Lee 1993; Walters 1986; Holling 1978). Implementing a

long-term, ecosystem-based monitoring and research program in Grand Canyon requires a

detailed understanding of the river's geomorphic framework, as well as its hydrology, 29 sediment balance and natural range of variability of the processes at work there. Because debris flows periodically contribute large volumes of fine sediment and large boulders to the Colorado River, these events can significantly affect the sediment balance of the river on local to-reach spatial scales and over short and long time periods (DOT 1995). Since storage of fine sediment along the river is mostly controlled by the hundreds of pool and

debris fans/rapid complexes located along the river, changes in those controls by future

debris flows will naturally alter the sediment balance of the channel, and in turn, alter the

aquatic habitat. Monitoring such physical changes is also important as numerical models

developed for flow and sediment transport on the river are refined.

Since debris-flows were first recognized as having important impacts on resources

below Glen Canyon Dam (DOI 1995), Glen Canyon Dam operations have been modified

(August, 1991) to better preserve downstream sediment resources. Sediment, in the form

of alluvial fans, sand bars and talus is considered to be the foundation of the river's

ecosystem (DOT 1995). With ongoing debris flows, debris fans become more constricting

and sediment storage in upper pools is expected to increase throughout the system.

Information about impacts of tributary debris flows is being integrated into a long-term

monitoring and research plan for the river resources below Glen Canyon Dam. This

program is being designed and implemented with the hope of modifying dam operations,

within limits set by law, to preserve and restore downstream resources toward

sustainability.

In many cases today, science is being used to provide resource managers with

information useful in answering applied questions (Stanford et al. 1996). Ultimately, this

study was motivated by a need to establish links between a unique geologic process --

debris flow -- in Grand Canyon National Park and riverine resources managed by the 30

National Park Service, the Bureau of Reclamation, and other Federal and State agencies.

As stated earlier, interest in Grand Canyon debris flows was initially stimulated by environmental concerns for downstream resources affected by operations of Glen Canyon

Dam. Some of the information presented here was incorporated into an environmental impact statement that examined all dam-related effects on varied resources, entitled

Operation of Glen Canyon Dam Environmental Impact Statement (DOT 1995). Information on debris flows in that document is mostly derived from Chapter 3 of this volume, which was originally published by Melis et al. (1994) as a USGS Open-File Report. This debris flow study was one of an overall series of projects conducted by the U.S. Geological

Survey (USGS) in cooperation with the Bureau of Reclamation during the 1980s and 90s.

The Glen Canyon Dam EIS was also part of a coordinated effort to evaluate alternative Glen Canyon Dam operations as a means of preserving and restoring natural river resources in Grand Canyon, while allowing the dam to continue operations for power generation.

Components of the study

Historical data on the occurrence of debris flows in Grand Canyon is limited and has been difficult to obtain because of the isolated setting and inaccessibility of many of the

Colorado River's tributaries. The USGS sediment study made during the preparation of

the Glen Canyon Dam EIS used boat access to these tributaries. I examined debris-flow

settings, debris fans, and rapids throughout Grand Canyon. I also made detailed studies of

the effects on the Colorado River of historic debris flows that occurred between 1939 and

1994 in 39 tributaries. 31

In many cases, I was able to examine sites of recent debris flows within hours-to

weeks of their occurrence, but unfortunately have never witnessed one of these sediment

floods except during recessional flow. This thesis describes what is presently known

about the process of debris flow in Grand Canyon based on field studies made between 1989 and 1996.

Purpose and scope of the study

The purpose of this study is to document the historic occurrences and explain the

significant impacts of Grand Canyon debris flows on the Colorado River and its resources

below Glen Canyon Dam. Most of the chapters presented here were prepared jointly with a

variety of colleagues for submission as journal/proceedings articles or USGS publications.

The information included here pertains to: 1) a literature review of historic observations and

writings that slowly led to our present understanding of debris flows as a geomorphically-

important process, 2) debris-flow frequency and magnitude in Grand Canyon (1872 to

1996), common initiation mechanisms documented for debris flows in the Canyon, and

hydroclimatology of debris flows (1939 to 1996); 3) a conceptual model to explain the

hydrologic and sedimentologic characteristics of recent debris flows, including information

on peak discharges, flow transformations and hydrographs; 4) the role of historic debris

flows and older debris fans in controlling the structure of the mainstem channel and its

fluvial processes, including storage of fine sediment; and 5) the reworking effects of a

controlled-flood released from Glen Canyon Dam in spring, 1996 on recently-aggraded debris fans and rapids.

Chapter 2 is a literature review summarizing the key studies, eyewitness descriptions, and related historical work leading up to the present recognition of debris 32 flows as an important geomorphic process affecting the Colorado River in Grand Canyon

National Park. The purpose of this summary is to provide a broad perspective on the importance and unique qualities of debris flows relative to Grand Canyon and its river.

This is done by examining the history of how debris flows and their deposits were initially

misinterpreted by early workers, but later recognized and documented by geologists as

important agents of landscape development throughout the western U.S. -- and eventually

in Grand Canyon. I also briefly examine the recent studies of alluvial fans/processes that

are controversial, and try to relate them to the overall perceived importance of debris flows

in the geologic record. As part of this review, I purposely include extensive excerpts from

early articles on debris flows and their environments to provide background on how much

progress has been made in our understanding of hillslope processes in settings such as

Grand Canyon. This background information on historic debris-flow studies is important

with respect to earlier interpretations of the geomorphic framework of the Colorado River

in Grand Canyon. My objective being to offer the interested reader a better-appreciation for

the efforts made by early workers to educate their colleagues about debris flows.

Additional excerpts provide insights on how early workers struggled to interpret debris-

flow dominated terrains without the advantage of a first-hand understanding of this unique

hillslope process.

Chapter 3 was written by Melis, Webb, Griffiths and Wise, and was originally

published as U.S. Geological Survey Water-Resources Investigation Report 94-4214

(Melis et al. 1994). This chapter presents a summary of baseline data on system-wide

frequency of Grand Canyon debris flows between 1872 and 1997 that were derived mainly

from repeat historic photography, assessment of aerial photographs, and historical written

accounts. Debris-flow magnitude is discussed, and data are presented that were derived 33 from field studies for debris flows between 1939 and 1996. Historical data on debris-flow frequency and magnitude offer a uniquely-detailed perspective on the relative importance of recent debris flows under modern conditions of river regulation.

This chapter also summarizes recently-observed mechanisms for triggering Grand

Canyon debris flows, based on recent floods in 39 tributaries of the Colorado River between 1939 and 1997. Initiation mechanisms for debris flows are often complex in

Grand Canyon, and can vary greatly depending on the lithologic, climatic and physiographic variability of tributary settings. Grand Canyon has many varied settings for debris-flow initiation, mainly owing to its diverse and interlayered lithologies, as well as its extreme topographic relief.

Finally, this chapter provides some information on the hydroclimatology of historic

Grand Canyon debris flows relative to regional and local instrumental climate data from

1939 through 1996. Debris flow initiation requires some form of either intense rainfall or runoff in the vicinity of steep, unconsolidated hillslope detritus. I examine the important controlling variables of debris-flow frequency and magnitude, such as available sediment and intense precipitation. In doing this, I outline the types of storms that have initiated historic debris flows, document trends of intense 20th-century precipitation as they relate to recent debris flows and discuss how such trends may be reflected in magnitude and frequency patterns of historical debris flows.

Chapter 4 was written by Melis, Webb, and Griffiths as a proceedings-volume article for the First International Conference on Debris Flow Hazards and Mitigation: Mechanics, Prediction, and Assessment; a meeting sponsored by the American Society of

Civil Engineers scheduled for summer, 1997, in San Francisco, CA (Meus et al. 1997).

This chapter presents a conceptual model for explaining recent debris-flow behavior in 34

Grand Canyon. The model is based on peak discharges, debris-fan volumes, and

sedimentologic and hydrologic field data derived from field sites where debris flows reached the Colorado River. The combined analyses of these data suggest three distinct flow behaviors associated with recent debris flows along the Colorado River.

Chapter 5 describes the reach-varied physical geomorphology of the Colorado

River channel relative to regional and local geology and its alluvial fans. Much of the information in this section should be of interest to geologists and hydrologists familiar with debris-flow dominated, rivers in bedrock canyons , but the information also relates broadly to management objectives for sustainablility of riverine resources related to operation of

Glen Canyon Dam. The reach-varied characteristics observed along the mainstem channel relate to the regional structure of the Colorado Plateau, the river's debris fans, and the bedrock geology exposed at river level and within tributaries. In this chapter, I examine how these complex factors such as river-level lithology, particle size of fan sediments, and fan geometry combine to control the Colorado River through Grand Canyon, and how such changes might influence physical and biological resources of the river. This is the first time such an assessment of reach variability has been attempted at a detailed level using river- channel measurements for the Colorado River in Grand Canyon.

Chapter 6 was written jointly by Webb, Melis, Griffiths and Elliott, and was originally published as a USGS Open-File Report (Webb et al. 1997). It was later rewritten as a chapter in an American Geophysical Union monograph on results of the 1996

Glen Canyon Dam controlled flood. This chapter describes the system-wide effects of the spring, 1996, controlled-flood release from Glen Canyon Dam on recently--aggraded debris fans and rapids in Grand Canyon. Mainstem floods provided the counterbalance to tributary-imposed controls on the mainstem (fan/rapids complexes) before regulation. As 35 part of recent attempts to restore and preserve resources below Glen Canyon Dam, floods are recognized as an important factor.

Chapter 7 is a summary of the study results and overall conclusions about how ongoing debris flows will likely change characteristics of the Colorado River through

Grand Canyon over time under regulated conditions.

Beyond the environmental implications of Grand Canyon debris flows for downstream resources, this report hopefully adds to the general knowledge on Grand

Canyon geology, and on interactions between rivers and debris flows/fans wherever such interactions occur. While Grand Canyon tributaries reflect erosional processes related to past and present climatic conditions, debris flows in them are highly dependent on lithology, geologic structures and rates of tectonism on the Colorado Plateau. Ideally, the significance of debris flows described here is best appreciated in the context of regional geologic and climatic history. Unfortunately, climatic reconstructions near Grand Canyon are limited, and this study adds nothing to those data. Hopefully, documentation of historic debris flows in Grand Canyon can provide additional insight into the most important factors influencing the geomorphic framework of the Colorado River. Such information is useful to those interested in Grand Canyon geology and geomorphic processes occurring there.

The Grand Canyon study area

The study area includes all "geomorphically significant" tributaries of the Colorado

River between Lee's Ferry and Diamond Creek, Arizona (river miles 0 to-225), excluding the Paria and Little Colorado Rivers, and Kanab and Havasu Creeks (river miles 0.5-R,

61.5-L, 141.3-R, and 156.9-L, respectively). I define and list the names of these 36

tributaries, their drainage areas, and rapids that have formed at their junctures with the Colorado River.

Units and place names

In reporting the results of this investigation, I use the English unit of mile to

describe the locations of tributaries along the Colorado River (river mile), and the unit of

feet for altitude on location maps; metric units are used for all other measures. Use of river

mile has considerable historical precedent and provides a reproducible method of describing

the location of tributaries with respect to the Colorado River. Locations of tributaries are

described in river miles downstream from Lee's Ferry, Arizona, and a descriptor of "L" for

river-left and "R" for river-right. The left and right sides of the Colorado River are

determined as one faces downstream. River-mile designations for tributaries were

estimated from USGS 7.5-minute quadrangle maps in conjunction with the most recent

edition of a well-known river guide (Stevens 1990).

Names of tributaries and other significant landmarks were derived from two

sources. Published USGS 7.5-minute quadrangle maps are the source for most names of

tributaries. For tributaries without names on 7.5-minute quadrangle maps, I use either

informal names, which are designated within quotation marks and are typically obtained

from Stevens (1990), or refer to the tributary as "an unnamed tributary" at a specified river

mile and side. In addition, I also refer to "E-Areas," which are smaller drainage areas that

may have debris flow and flash-flood potential, but did not meet all of the criteria of my geomorphically significant" category for Grand Canyon tributaries.

I typically refer to "Grand Canyon" in broad reference to both Marble and Grand

Canyons. Technically speaking, "Marble Canyon" is the canyon reach of the Colorado

River between Lees Ferry, Arizona and the confluence of the Little Colorado River (LCR, 37 river miles 0 to-61.5); I refer to Marble Canyon only for the purpose of describing specific tributaries in that reach of the river. Grand Canyon, which is designated between the confluence of the LCR and the Grand Wash Cliffs (river miles 61.5 to-about 280), is considerably larger and better known than Marble Canyon. As a result, I therefore refer informally to the first study-reach of Grand Canyon (designated study-reach five in this study), as "Furnace Flats;" the reach located between river miles 61.5 and 78.0.

The study area is isolated and little historical documentation exists for most of these tributaries; especially the ephemeral ones. Consequently, the type and amount of information known for each tributary varies greatly. However, all data presented here provide useful information for understanding the overall role of debris flow in Grand

Canyon's ongoing evolution and management of Lower Colorado River riparian resources. 38

CHAPTER 2

BACKGROUND AND PREVIOUS WORK

Earliest Grand Canyon studies

In the decade following John Wesley Powell's exploration of the Colorado River,

Clarence Edward Dutton conducted pioneering geology studies in the "Grand Canyon

District," known today as the Colorado Plateau. Dutton's investigations of the Colorado

Plateau are renowned (Anderson 1977). Most notable is Dutton's landmark 1882 monograph entitled The Tertiary History of the Grand Cation District; with Atlas. In that report Dutton combined his experiences in the southwest with many ideas and theories of his own as well as those offered by his colleagues Gilbert and Powell. Dutton's 1882 report represents the synthesis of ideas and observations made by these men over three decades, and was the first major treatise on the physiography and erosional history of the canyons and plateaus that dominate the Colorado Plateau; including Grand Canyon.

Although Dutton's Grand Canyon District monograph remains a monumental work, it was by his own admission only partially complete. Dutton's Tertiary History was unique compared with other reports he had written on the geology of southwestern landscapes, and this is best described in his own words:

I have in many places departed from the severe ascetic style which has become conventional in scientific monographs. Perhaps no apology is called for. Under ordinary circumstances the ascetic discipline is necessary. Give the imagination an inch and it is apt to take an ell, and the fundamental requirement of the scientific method -- accuracy of statement -- is imperiled. But in the Grand Canyon District there is no such danger. The stimulants which are demoralizing elsewhere are necessary here to exalt the mind sufficiently to comprehend the sublimity of the subjects. Their sublimity has in fact been hitherto underrated. Great as is the fame of the Grand Canyon of the Colorado, the half remains to be told (Dutton 1882, p. xvi). 39

Although Dutton's Tertiary History addressed many aesthetic and broad geological characteristics of the region, it lacked details on the hillslope and fluvial processes at work throughout the region. This case might be explained by the fact that the science of Geology was relatively new, and the body of first-hand observations and information in geological literature on infrequent processes such as debris flows was limited during the late 19th century.

The following passage from Dutton's Tertiary History chapter, entitled "The excavation of the Grand Canyon," provides insight on his limited understanding of the range of geomorphic processes at work in Grand Canyon:

In common parlance it is customary to say that the river [the Colorado] has cut its cation. This expression states but a small portion of the truth. The river has in reality cut only a narrow trench, of which the width is equal to the width of the water surface. It has also been the vehicle which has carried away to another part of the world the materials which have been cut away by both processes. Opening laterally into the main chasm are many amphitheaters [tributaries] excavated back into the main platform of the country. At the bottom of each is a stream-bed, over which in some cases a perennial river flows, while in other cases the water runs only during the rains. Like the trunk-river itself, these streams, whether permanent or spasmodic, have cut down their channels to depths varying somewhat among themselves, but generally a little less than the depth of the central chasm. These tributaries often fork, and the forks are quite homologous to the tributaries in the foregoing respects. They too have cut narrow gashes no wider than their water run surfaces. Down the faces of the walls and down the steep slopes of the taluses myriads of rain gullies. When the rain comes it gathers in rills, which cascade the wall-clefts and rush headlong through the troughs in the talus. Carrying down the abundance of sand and grit, the waters scour out these little channels in much an and same way as their united streams and cut down their beds in the amphitheaters in the main chasm itself. But the work of corrasion by running water is limited to cutting at any the cutting of very narrow grooves in the rocks, the width of the given time and place being equal to the width of the water-surface of the stream. Corrasion alone, then could never have made the Grand Cañon what it is. The amount of material removed by that process is but a very small fraction of the total excavation. Another process acting conjointly with the corrasion, and in an important sense dependent upon it, has effected by far the greater part of the destruction. This additional process is weathering. In order to comprehend the combined results of the two, it is necessary to study their action in detail (Dutton 1882, p. 230-231). 40

Dutton intuitively knew that "another process acting conjointly with corrasion," was required to explain the immensity of erosion displayed in Grand Canyon. However, it is clear from his report that knowledge of channelized sediment-transport processes at that time was limited to what is termed "normal streamflow" (Pierson and Costa 1987), or strictly fluvial processes. Not having observed or studied debris flows firsthand, Dutton and others naturally relied on the obvious processes of physical and chemical weathering to explain the large-scale erosion displayed in Grand Canyon's tributaries.

In another section of his 1882 report discussing the formation of Grand Canyon rapids, Dutton provides the following explanations for the Colorado River's whitewater:

The rapids are, however, the results [sic] of two independent causes. (1) When the course of the stream lies in the hard rocks the rate of declivity [meaning gradient] is greater. The explanation is obvious. (2) The second cause is of a totally distinct nature. At the opening of every lateral chasm or side gorge a pile of rocks and rubble is thrown out into the main stream. Most of these side gorges are dry throughout the greater part of the year. But when the rains come their narrow beds are occupied with floods of muddy water, rushing downward with great velocity and often in great volume, bowling along fragments of all sizes from a few pounds to many tons. Thus an obstruction like a low dam is built across the river. The declivity of the side gorges is always much greater than that of the main stream. Their slopes are rarely less than 200 feet [61 m] to the mile, except in those tributaries which, like Kanab Creek and Cataract Cañon [meaning Havasu Creek], are of great antiquity. The minimum slopes of the beds of the great amphitheaters in the Kaibab are seldom so small as 200 feet [61 m] to the mile. The power of a great flood rushing down such slopes is indeed formidable. When the torrents reach the river the larger fragments are dropped; for the maximum slope of the main stream (reckoned throughout any stretch exceeding four or five miles in length) never exceeds 25 feet [7.6 m] to the mile, and the water, though enormous in volume at flood time, has less velocity than the torrents of the side chasms. The river has, however, sufficient power to sweep onward masses of considerable size, which are rapidly ground up as they are rolled along (Dutton 1882, p. 241-242).

In a final summary passage Dutton reiterates his ideas regarding the river's rapids:

Reverting here in a summary way to the two classes of rapids formed in the river, we find that they are due to two wholly distinct causes -- 1st, to the unequal hardness or resisting power of the strata in the bed of the stream; -- 2nd, to the accumulation of piles of large fragments washed in through the lateral gorges. The greater number of rapids are due to the latter cause. The general characteristics of 41

the two kinds, however, are strikingly different. Those which occur in the hard crystalline rocks are very long, and with some notable exceptions are less violent and headlong than the others, but they are also complicated with the second class of rapids, i.e., those due to the bowlders [sic] rolled in from the side chasms. If we separate the latter effects, we find the effect of hardness of the strata to be a general increase of slope distributed with some approach to uniformity over many miles of length of the river bed. It is plainly so in the Kaibab and Sheavwits [plateaus], and this effect is not at all obscured by the joint operation of both causes. The rapids of the second class are short and violent, as might have been inferred from the nature of the cause which produces them (Dutton 1882, p. 244).

Many of the ideas expressed by Dutton in these excerpts about the Colorado River

and its rapids likely originated from Powell's earlier first-hand observations. Dutton was

for the most part completing the task of formalizing Powell's ideas when it came to explaining rapids because his own experience with the Colorado River in Grand Canyon

was limited. As for his initial explanation of rapids, the discordant hardness of lithologies exposed at the river is no longer recognized as an explanation of Grand Canyon rapids

(Webb et al. 1996; Hereford et al. 1996; Melis et al. 1994; Melis and Webb 1993; Webb et al. 1989; Kieffer 1985). Present ideas on the creation of Grand Canyon rapids rely instead on Dutton's second proposition, that being the introduction of boulders to the river from side canyons in almost all cases (98 percent).

While the relationship between tributary canyons and rapids was accepted and described by Dutton (as it was also initially by Powell), his explanations for the transport of large boulders into the river relied on: 1) unobserved, extreme flow velocities occurring during tributary streamfloods (fluvial transport); and 2) the observed relative steepness of tributary channels versus mainstem gradient. Like Powell a decade earlier, Dutton's explanations for rapids failed to offer alternative hypotheses for widespread boulder transport from tributaries to the Colorado River as evidenced by the numerous rapids that occur throughout Grand Canyon. Dutton's limited interpretation of rapids is showcased 42 here not to discredit his pioneering work, but to expose the limits of knowledge and observational experience geologists of that era had with respect to highly sediment-laden channelized flows (mass-wasting processes). Dutton had no first-hand experience with debris flows, however, they are a common hillslope process widely recognized in many environments today (Bull 1977; Blair and McPherson 1994), and they have the unique

ability to transport large boulders through low-gradient channels (Rodine and Johnson

1976).

Since Dutton's time, detailed geologic studies of Grand Canyon have resulted in

many theories concerning its formation and evolution (for examples, see Lucchitta and

Young 1986; Hunt 1969). Following Dutton's work, additional Grand Canyon studies

accelerated, resulting in many reports and papers on its lithologie, stratigraphic,

paleontologic, and structural characteristics. However, other Grand Canyon geology

topics, such as hillslope processes related to its continuing weathering and erosion, remained poorly understood well into the 20th century.

The lack of available information on Colorado River tributary floods in the 19th

century greatly hindered Dutton's understanding of the Colorado River and its canyons.

However, even reports made on Colorado River geomorphology as recently as the 1960s

fail to identify the importance of debris flows in Grand Canyon tributaries. Some modern

workers have even refuted the close relationship between tributaries and rapids; a fact

recognized by Powell and Dutton more than a century earlier (Leopold 1969). One

plausible explanation for such oversight, was provided long ago by Chamberlin in his

essay entitled, "The method of multiple working hypotheses," in which he makes the

introduction: "With this method the dangers of parental affection for a favorite theory can be circumvented" (Chamberlin 1890). Schumm (1991) also provides an excellent 43 overview of the many potential pitfalls that await scientists conducting Earth science research. Suitable explanations for Grand Canyon rapids would wait until well into the

1970s - 80s (Webb et al. 1987; 1988; 1989; Howard and Dolan 1981; Cooley et al. 1977).

In defense of early Grand Canyon researchers, accounts of debris flows and studies

outlining their exceptional ability to transport large boulders were not readily available to

scientists such as Dutton, Gilbert, Powell and others. As a result, Dutton could only offer

general explanations for Grand Canyon rapids based on those processes familiar to him. It

was only with continued exploration of the Colorado River following Dutton's reports that

the unique characteristics of debris flows were recognized, and their importance in Grand

Canyon geology and other regions documented.

Earliest eyewitness account of a Grand Canyon debris flow

First witnessed and described by Robert Brewster Stanton in July, 1889, Grand

Canyon debris flows went mostly unnoticed by geologists during the 19th and 20th

centuries. Formal investigations of debris flows and their effects on the Colorado River

began relatively recently with the study by Cooley et al. (1977); work stimulated by the

occurrence of a large debris flow in Crystal Creek (fig. 1.1; river mile 98.3-R) in

December, 1966. That debris flow radically altered navigation of the river by aggrading

Crystal Rapid. The lack of research on Grand Canyon debris flows prior to the 1960s tributaries; owes mostly to: 1) the remoteness and difficult access to most Grand Canyon

2) the relative infrequency of such floods combined with limited use of the Colorado River

for navigation before 1965; and 3) the lack of publications on debris flows before 1920.

Although an unlikely student of Grand Canyon debris flows, Stanton provides, the

earliest account of such a sediment flood (Stanton 1965), based on his observations made 44 while leading the second successful exploration of the Colorado River through Grand

Canyon in summer, 1889. Stanton explored Grand Canyon during two river trips. He lead the first one in spring and summer, 1889, and the second one in winter and spring,

1890 (Webb 1996; Smith and Crampton 1987; Stanton 1965). Unlike Powell -- the geologist and ethnographer, Stanton was a civil engineer specializing in railroad construction and his interest in Grand Canyon and its river were practical in scope.

Stanton's initial interest in the Colorado River centered around surveying its course through the great canyons of the southwest to evaluate the feasibility of a water-grade railroad route between Green River, Utah, and Yuma, Arizona. Later, after experiencing Grand Canyon first-hand, Stanton's interest in the Colorado River broadened to include its natural history.

An important exploratory expedition in its own right, Stanton's second expedition produced the first complete mile-by mile photographic survey of the Colorado River through Grand Canyon (Webb 1996). Stanton's railroad project was fortunately never started, as it would have wreaked untold havoc on the arid-land and riverine environment of the canyon bottom. By 1919 the Colorado River corridor was protected from railroad schemes following the establishment of Grand Canyon National Park (fig. 1.1). Although

Stanton's dream of a Grand Canyon railroad was never realized, the engineer's contribution to our modern understanding of Grand Canyon debris flows was significant, albeit unintentional.

During an emergency retreat from the Colorado River following his failed first attempt to traverse Marble Canyon in July, 1889, Stanton witnessed a debris flow. The engineer and his crew were lucky to survive the tributary flood. Stanton, a meticulous keeper of field notes, later described the awe-inspiring event in his field ledger dated Thursday July 18th, 1889: 45

We go a mile or two further up canyon [South Canyon, river mile 31.6-R] and stop to rest. Great storm comes up. Streams of water and rock pour down sides of canyon in every direction. Shelter behind a leaning rock. This storm shows great care with which the drainage above the road [meaning the proposed railway route] must be provided for. This storm in the side canyon was wonderful sight, and a great revelation to me.

As the rain commenced to fall we heard some rocks roll down the slope behind us, when we looked up, and it seemed as if the whole slopes [sic] of the gorge had begun to move at the top. Little streams of water came over the top, and in a moment they changed into streams of mud; and as they came down they gathered strength and turned to streams of mud and rock, under-mining larger rocks; and starting them they plunged ahead, and in a few moments the whole sides of the canyon seemed to be moving down upon us with a roar and awful rumbling noise; and as the larger rocks plunged ahead of the streams, they crashed against other rocks, breaking into pieces; and the fragments flew in to the air in every direction, hundreds of feet above our heads; and as these came nearer the bottom where we were, it seemed as if we were to be buried in an avalanche of rock and mud. But the rain soon ceased, and the whole canyon resumed its deathlike stillness except for the noise of the little stream of muddy water running in the creek bed at our feet.

It is a noticeable fact, that before today, in all the weeks we have been in these canyons we have never heard as much as a pebble drop or roll down from the side (Smith and Crampton 1987, p. 88).

Stanton's description of the muddy, debris-laden torrent he witnessed in what is now South Canyon (river mile 31.6-R), remains the best-documented eyewitness account of a Grand Canyon debris flow to date. Although Stanton lacked the temilnology needed to describe the flood he witnessed as a "debris flow," by today's standards (Pierson and

Costa 1987), it is clear from his account that he recognized the potential of such floods to quickly transport boulders and finer sediment from steep hillslopes to the Colorado River.

From his experience in South Canyon, Stanton gained valuable insight into Grand

Canyon's unpredictable and dynamic character with respect to hillslope processes and sediment transport; first-hand experience that eluded later researchers for several decades.

Stanton was not only impressed by the debris flow he witnessed, but also by the fact that during two months previously spent traversing the Colorado River he had not heard a 46 single stone move. Had Dutton experienced such a flood, he likely would have refined his interpretation of Colorado River rapid formation.

In a manuscript on the Colorado River written many years after the completion of his Grand Canyon work, Stanton elaborated on the strange flood he witnessed in 1889. In this more detailed passage, Stanton provides useful insights into the nature of the thunderstorm that initiated the July, 1889, debris flow in Marble Canyon. He notes in the following excerpt that the July 18th flood was preceded by intense thunderstorms which occurred all during the previous night. In retrospect, Stanton wrote of the storm and flood:

When a few miles up, there came on another of those sudden and furious storms, so common at times in the dry southwest.

The rain did not fall in showers, or in "bucketfulls [sic]," but in what could almost be termed solid sheets of water. We took shelter under an immense leaning rock, in the bottom of the canyon valley, near the creek. As the rain increased, I heard some rock tumbling down behind us, and, looking up, I saw one of the grandest and most exciting scenes of the crumbling and falling of what we so falsely call the everlasting hills. As the water began to pour over from the plateau above, it seemed as if the whole upper edge of the Canyon had begun to move. Little streams, rapidly growing into torrents, came over the hard top stratum from every crevice and fell on the softer slopes below. In a moment they changed into streams of mud, and, as they came farther down, again changed into streams of water, mud and rock, undermining huge loose blocks of the harder strata and starting them, they plunged ahead. In a few moments, it seemed as though the slopes on both us, sides of the whole canyon, as far as we could see, were moving down upon first with a rumbling noise, then an awful roar. As the larger blocks of rock of the streams, they crashed against other blocks, lodged on the plunged ahead the slopes, and, bursting with an explosion like dynamite, broke into pieces, while flew into the air in every direction, hundreds of feet above our heads, fragments down and as the whole conglomerate mass of water, mud, and flying rocks, came slopes nearer to where we were, it looked as if nothing could prevent us from the of being buried in an avalanche of rock and mud. It was a scene of the wildest fury the elements! Canyon The rain ceased as suddenly as it began, and, in a few minutes, the whole resumed its deathlike stillness, only broken by the rumble of the stream of muddy of the water, running in the creek bed at our feet. It is a noticeable fact that in all weeks we had been in the Canyons, before this day, we had never heard as much the heavy as a pebble drop, or anything roll down from the cliffs, not even during storms in Marble Canyon. This latter is accounted for by the different nature of the walls-the hard limestone and marble (Stanton 1965, p. 90-91). 47

In this later account, Stanton describes key features of Grand Canyon debris flows:

1) they are most commonly initiated by extremely intense, short-duration summer

thunderstorms; 2) their occurrence is associated with steep, unconsolidated hillslopes and is

closely related to the variable bedrock geology and source-sediments, such as shales and

colluvium containing limestone boulders; and 3) they occur with little warning following

wet periods, and are usually of short duration. Unknowingly, Stanton was also the first

person to describe the most common type of Grand Canyon debris-flow initiation

mechanism, termed the "fire-hose effect" (Griffiths et al. 1996; Johnson and Rodine 1984);

a triggering mechanism for hillslope failure described in later sections of this report.

Stanton's eloquent description of what occurred high above him on the slopes of

Marble Canyon in July, 1889, marked the truest beginning of debris-flow observation and

studies in Grand Canyon; nearly a century before the first formal descriptions of debris

flow there were published. However, his insights about the importance of debris flows in

Grand Canyon was not readily available to Grand Canyon geologists until 1987. Stanton's

description of a Grand Canyon debris flow, along with his more than 400 photographs of

the river corridor greatly aided in the present debris-flow study described here. Stanton's

photographic record of the Colorado River inadvertently provided a baseline datum for

assessing historic debris-flow frequency and magnitude in Grand Canyon; a legacy for

which the author is gratefully indebted to him.

Earliest published accounts of highly-sediment-laden flows

One of the earliest descriptions of a highly sediment-laden flood in the southwest

(possible a debris flow, but termed a "sheetflood" by the author), appeared in the U.S. 48 literature in the late 19th century. In his 1897 paper, W.J. McGee provides an excellent eyewitness description of what he terms a sheetflood in southern Arizona:

During the 1894 expedition a moderate local rain occurred while the party were [sic] at a Papago rancheria near Rancho de Bosque, some 15 miles [24 km] north of the international boundary at Nogales: the rainfall was perhaps one-fifth of an inch [0.5 cm], sufficient to moisten the dry ground and saturate clothing despite the concurrent evaporation, and was probably greater in the adjacent foothills of Santa Rita range... The shower passed in a few minutes and the sun reappeared, rapidly drying the ground to whiteness. Within half an hour a roar was heard in the foothills, rapidly increasing in volume... The water was thick with mud, slimy with foam, and loaded with twigs, dead leaflets and other flotsam; it was seen up and down the road several hundred yards in either direction dully half a mile in all... The torrent advanced at race-horse speed at first, but, slowing rapidly, died out in irregular lobes not more than a quarter of a mile below the road; yet, though so broad and tumultuous, it was nowhere more that about than 18 inches [46 cm] and generally only 8 to 12 inches [20 to 30 cm] in depth... Within the flood, transverse waves arose constantly, forming breakers with such frequency as to churn the mud-laden torrent into mud-tinted foam; and even when breakers were not formed it was evident that the viscid mass rolled rather than slid down the diminishing slope, with diminishing vigor despite the constant renewal from the rear. Such were the conspicuous features of the sheetflood -- a thick film of muddy slime rolling viscously over a gently-sloping plain (McGee 1897, p. 100).

Although McGee does not use the term debris flow, he describes many characteristics of the sheetflood that are commonly associated today with debris flows; including a lobate appearance along the flow's leading edge, multiple surging waves, rolling flow behavior, and viscous muddy appearance (Costa 1984). It is unclear whether McGee actually witnessed a debris flow or a sheetflood based on his description, but he did refer to the phenomenon as "a sort of mudflow" (McGee 1897, p. 108). Most important, he recognized this type of flood as different than streamflooding and reported its uniqueness in the literature.

It is not clear exactly when the hillslope process of debris flow was first recognized as a significant sediment-transport process in the southwestern U.S. As is commonly the case, new discoveries often occur as a series of inadvertent observations made by a group 49 of workers who share their experiences. The historical recognition of debris flows as an important geologic agent is not an exception. Although detailed descriptions of viscous, muddy floods and their deposits were published by early workers in the U.S. and abroad before 1930 (Scrivenor 1929; Eckis 1928; Blackwelder 1928; Pack 1923; Rickmers 1913;

Howe 1909), such reports were either not widely read, or such events were considered so unusual that they were dismissed as unimportant by geologists.

Blair and McPherson's recent publications (1994; 1993; 1992), provide compelling evidence for the pervasiveness of debris-flow deposits in the stratigraphic record, and provide comprehensive historical summaries of early published debris-flow observations.

However, their claims about the importance of debris flows in the geologic record are subject to debate. Controversies in the professional literature continue into the 1990s over the relative importance of debris flows versus fluvial processes in alluvial-fan development, and facies models (see Blair and McPherson 1992;1993; Hooke 1993; McCarthy and Cadle

1995; Kim 1995; Blair and McPherson 1995a,b). Such controversies suggest that interpretations of hillslope versus fluvial processes preserved in the stratigraphie record are still evolving.

In an early study describing the alluvial fans of Owens Valley, California published in 1911, Trowbridge provides a compelling example of how unaware most geologists of that era were with respect to debris-flow processes. In his 1911 paper, Trowbridge gives excellent detailed descriptions of numerous channels and ridges comprising fan surfaces; features that are consistent with the debris flows that created the Owens Valley fans:

Low ridges and shallow depressions on the individual fans constitute topographic features of the second order. These are the channels and depositional features of the streams which deposited the fans... The depressions are, as a rule, about 10 ft. [3 m] deep and less than 100 ft. [30.5 m] across, though at a maximum they reach a depth of 20-25 ft. [6.1-7.6 m] and width of not quite 100 ft. [30.5 m]. Their bottoms are usually flat and their slopes as steep as the material will permit. The 50

elevations are less numerous than the depressions, and have less relief. They are seldom more than 5 ft. [1.5 m] above the surrounding plain, and their height in many cases is only equal to the diameter of the individual bowlders [sic] of which they are composed. The ridges consist of mere divides between channels and of lines of bowlders [sic] bordering the channels...

It is clear that these channels on the fans mark the courses of the distributaries from the fan-making streams. The streams branched again and again, some of the distributaries reaching the inter-fan depressions and flowing off through them. It is equally clear that some of the elevations are merely inter-distributary divides. The origin of the ridges bordering the depressions is not so clear. Possibly they are in principle natural levees, built as the waters overflowed their channels. It is understood that these are the streams which deposited last on the surface of the fans. In the building of the bajada, the channels undoubtedly shifted frequently, those of one time being filled up and a new set formed during periods of greater deposition following heavy rains or the rapid melting of snow in the mountains (Trowbridge 1911, p. 711-712 ). To most modern workers, it is obvious from the above excerpt that Trowbridge is describing features of debris-flow deposits commonly recognized today on alluvial fans

(Blair and McPherson 1994; Bull 1977). After admitting his inability to explain the formation of the ridges, it is worth noting that Trowbridge, for lack of another conceptual model, describes the boulder lines as probably forming "as the waters overflowed their channels... following heavy rains or the rapid melting of snow in the mountains;" again inferring strictly fluvial processes. In a later section, the geologist again notes similar features as being an integral part of alluvial fans:

A fourth topographic feature of the bajada consists of almost innumerable lines of bowlders [sic] which, though primarily a matter of the constitution of the bajada, affect the topography in a minor way.

These lines of bowlders have a radiate arrangement similar to that of the channels and ridges. The bowlders are so close together as to make low and discontinuous ridges, which by winding his way among the bowlders one may in some instances be able to cross without climbing. The height of the ridges is determined by the size of the bowlders , and is usually less than 10 ft. [3 m]. These seldom consist of more than two thicknesses of bowlders (Trowbridge 1911, p.713-714).

It is also clear from this second account that the features described by Trowbridge are 51 debris-flow boulder levees; features typically found on alluvial fans (Costa 1984; Johnson and Rodine 1984).

A key factor in recognizing deposits resulting from the hillslope process of debris flow centers around the ability of such events to transport large boulders over long distances over gentle gradients. Trowbridge's thorough descriptions of the alluvial fans in

Owens Valley, lacked only a reasonable process that could explain the odd surface features of the fans; a discontinuity that centered mainly on the transport of large boulders. For the interested reader, I include an extended excerpt from the section of Trowbridge's paper entitled "The transportation of large bowlders [sic]," in order to highlight this point. The following passage is not only historically significant to geomorphology, but also relates directly to Dutton's awkward interpretation of Grand Canyon rapids nearly forty years earlier:

One would hardly travel a mile on the Sierra bajada, or see the heads of the fans at the foot of the Inyo Mountains without asking how the large bowlders [sic] came to their present positions. It is essentially a problem of the means of transportation of the largest bowlders farthest from the mountain, for if they can be explained, the smaller bowlders may be considered to have been carried shorter distances by the same methods. Probably the most difficult problems are offered by the largest bowlders... measuring 10x20x30 ft. [3x6x9 m]...

It is clear that these bowlders came to their present positions through the agency of water. Though their size suggests glaciers as the transporting agents, such an explanation is out of the question... Nor is there anything in the fact of icebergs floating in a lake. The deposits with which the bowlders are associated are not lacustrine, and no lake existed in the valley during glacial times. The surface on which the bowlders lie clearly was made by running water. The bowlders are so clearly a part of the fans, and fans are so clearly running-water deposits, that the bowlders must be considered to have been transported by running water.

It is clear that the bowlders were not transported according to the common and well- known methods of stream transportation. They have certainly not been rolled along the bottoms of streams in the usual way... On the surface of the fans, furthermore, it is impossible for streams to have existed deep enough and strong enough to have so rolled these bowlders. Such streams would have to have a depth about equal to the diameter of the bowlders, be confined to a narrow channel, and flow with a velocity almost inconceivable for a stream... Also the gradient of the fans is 52

relatively low. Where the largest bowlders are, the slope is not over 6°, and west of Lone Pine not more than 3 ° ... They were perhaps carried to the heads of the fans by glaciers, glacial waters, and gravity. From there to their present positions, some special methods are called for.

A clue to a possible manner of transportation for these bowlders is obtained from observations of run-off water at the side of a previously dusty road after a heavy rain. Where the running water is but a small fraction of an inch deep, pieces of rock an inch in diameter are carried down stream. The moving of the large pieces involves the transportation of a very much greater amount of fine material. The movement of the large pieces is accomplished by the removal of fine material from the area immediately down stream from, and under, the lower part of the larger piece. By undercutting in front, and then by gravity and the push of water and sediment from behind, the large piece is pulled and pushed forward into the depression prepared for it...

Obviously this process would operate to best advantage where there was the greatest volume of water, and where fluctuations of water were greatest; that is, along the main channels of the fans, and on the Sierra fans rather than at the foot of the Inyo Mountains... If this method of transportation of large bowlders is not adequate, methods which are, are not known (Trowbridge 1911, p. 740-741).

In this excerpt, Trowbridge comes as close to perceiving the process of debris flow as he could have at that time, given his limited observational experiences. Trowbridge observed

that fine sediment was somehow important in the transport of the boulders, however, his

need to explain the role of fine particles in boulder movement again relied on the familiar

processes associated with fluvial processes. As a result, he came close to arriving at the

methods which "are not known," but failed to make the needed connection. In fact, the

understanding of debris flows improved relatively little for several decades after

Trowbridge completed his Owens Valley study.

Conceptual framework for later Grand Canyon studies

By the start of the twentieth century, William Morris Davis recognized that Grand

Canyon's incision into the Colorado Plateau resulted from a variety of geological processes, some of which he described as simple and others as complex. In his "cycle-of- 53 erosion" theory of landform evolution, Davis outlined ideas on landform evolution that occurred over very long time periods, and under conditions best described today by

Schumm (1991) as "decay equilibrium." Landforms, Davis concluded, evolved in a predictable series of erosional stages, depending on the progressive maturity of the landscape.

Such a theory provided a popular "big picture" approach to explaining landscape evolution, yet lacked detailed explanations of geomorphic processes behind such landform development; a predicament not unlike that faced by Trowbridge in his field studies of the

Owens Valley alluvial fans. Although the Davisian perspective is still embraced by historical geologists to a degree, its popularity declined rapidly with the onset of quantitative geomorphology in the 1940s and 1950s.

In contrast to the Davisian view of landscape evolution, studies by Grove Karl

Gilbert were highly process oriented, and largely ignored the variable of time which Davis' theories centered around. Gilbert's work had laid the foundation of his theory of landforms based on the idea of dynamic equilibrium before the end of the 19th century, although it was several decades before his ideas were fully explored. Although Gilbert's geomorphic theories were temporarily overshadowed during the Davisian era, they were later embraced by workers such as Langbein and Leopold (1964) during development and refinement of their "Quasi-equilibrium" theory on stream processes. Rather than being mutually exclusive, the ideas of Davis and Gilbert compliment one another in the explanation of landforms when both varying time scales and detailed processes are considered.

Despite the advances made in geomorphology during the late 19th and early 20th centuries, minimal understanding and experience with catastrophic mass-wasting events, 54 such as debris flows acted to impede progress in understanding the landscape evolution of the arid southwestern U.S. Tolman's writings on the bolson region of southern Arizona provide another good example of how handicapped early geologists were in their ability to accurately interpret features and processes seen in the alluvial valleys of the southwest:

The distribution of coarse material in the outwash desert deposits is in some ways remarkable... To appreciate what is accomplished in these waterways at a distance from the mountains, it is necessary for one to have seen some such display as I once witnessed on the desert plain... I was driving with no thought of rain, when suddenly I came to one of the numerous washes that cross the road , and found it filled to the brim with a foaming roaring flood, which was as impassable as the Niagara. Looking towards the mountains perhaps ten or fifteen miles distant, I noticed for the first time the black speck of the cloudburst, which happened to be at the headwaters of this arroyo and no other. In a few hours I was able to cross, and on my return found some boulders to approach three feet in diameter. And this was fifteen miles distant from the mountains! On the west slope of the Santa Rita mountains (east of Tucson) there is a ridge of rocks that was piled there by a torrent which later took a different course. These large boulders have remained invincible against the subsequent attacks of the water. This deposit has been mistaken for the moraine of an old glacier... Sometimes great boulders are found at some distance from the mountains. I recall one case where, between a great boulder and the parent mass from which it had been detached, there was a short ridge or knoll fifty feet high. I was asked how the boulder could have traveled up over the hill to its present position without aid of ice. I answered "the surrounding country has been washed away and the hill left as a remnant, since the boulder rolled down from above." (Tolman 1905-6, p. 15).

Despite Tolman's first hand observation of the debris flow he nearly witnessed in 1904, he apparently failed to draw on that earlier experience as he later wrote of the alluvial fan deposits of the bolson region:

The astonishing size of the boulders found in the outwash, with other phenomena described later, has led to a local popular appeal to glacial action. Boulders up to six feet [1.8 m] in diameter are found in the Catalina and Santa Rita outwashes near Tucson, for a distance of half a mile [0.8 km] from the present rock surface, and boulders at least two feet [0.6 m] in diameter six miles [9.6 km] and more from the present ranges. In ocean deposits occasional distribution of material of this size below cliffs only is noted... They [the boulders] are generally subangular, although some are partly rounded by their tumultuous journey... Almost no decomposition of either boulders or smaller material was noted in the various bajada deposits examined... Layers of sized and clean sorted boulders, gravel, and sand are discovered but more often sorting is completely lacking, although stratification 55

is always excellently developed. Pebbles are found in a matrix of sand, and boulders in a matrix of clay. In extreme cases the latter may not be due to water action. For instance, about two miles [3.2 km] northwest of Travertine Point, Salton Basin, California, talus deposits, remnants of old slopes, were filled in with wind dust, and proved treacherous to climb. This deposit formed under extreme conditions of great daily temperature change and small non-torrential precipitation (Tolman 1909, p. 156).

Tolman's interpretation of the alluvial fan features described in the preceding excerpt were fit into the paradigm that he was familiar with, in spite of his first-hand observations of freshly deposited debris-flow boulders made only a few years before. At this point, I must admit that my first encounter with debris-flow levee deposits, as an undergraduate hiking in Grand Canyon, left me confused as to their formative processes.

One of the best and most insightful early accounts of debris flows can be found in

Rickmers' 1913 book entitled The Duab of Turkestan. Early U.S. geologists given the opportunity to read these descriptive first-hand accounts of debris flows would likely have benefitted greatly from this information. However, Rickmers' book was published in

England and may have been unavailable to many geologists in the U.S. at the time of its publication. Rickmers had many opportunities to witness debris flows during his exploration of Turkestan. In the author's opinion, Rickmers' insights into the strange

"mudspates," as he termed them, were far ahead of his time. Rickmers' keen perception of the importance of the muddy floods he observed likely arose from his repeated observations of such floods during his expeditions. His descriptions are worth comparing with previous excerpts on alluvial fan features and processes previously presented:

...I should like to describe the mudspate which is an important form of rock- transport, especially among barren mountains, and which is frequently met with in the Alai-Pamirs... the typical mudspate consists of mire charged with a great number of rock splinters and blocks, but sometimes it may be composed almost entirely of "clean" stones ranging in size from a pepper-corn to large boulders. Thus "mud" signifies that the downpour is tough and plastic like a lava-stream or snow avalanche, while "spate" insists upon the sudden and spasmodic character of 56

the phenomenon. Nor can it be otherwise considering that a perfect suspension of much firm matter and coarse particles in little water cannot be kept up long, because it "sets" immediately when at rest, as everyone knows who has to do with mortar or concrete. Owing to this physical difficulty the mudspate is rare in comparison to the familiar modes of denudation, although in favorable localities it may become very important... When a gentle slope of grit and shingle has been soaked like a sponge by rain or melting snows there may come a time when it bulges out and slides off in the manner of a bog-burst on Irish moors. Slipping into channels and gullies this mass is mixed with more water, attains a higher speed and carries away soft material as well as rocks which it finds on its way. It is during this descent that the mudspate generally acquires its characteristic composition, for only by movement can an even mixture of liquid and solids be maintained. It is neither dry nor is there much free water, but the whole mass appears like a rapid flush of mud, although frequently the rock waste is so rough as no to suggest what is popularly called mud. Enormous boulders will float in this thick porridge like a cork on water or iron on quicksilver.

Operating with a minimum of water the mudspate liquifies itself automatically when, during its descent, it has become too thick. Stopping for a while it dams up the water runlet in the gully and then proceeds again, repeating, if needs be, the process several times. It is as if the mountain were suffering from some internal complaint easing itself in fits and starts.

Intermittent water supply owing to a dry climate, absence of strong vegetation and barren mountain flanks reaching up to the snowline are the conditions which favor the mudspate as a habitual and periodical phenomenon... As in the case of snow- avalanches the flowing mass has the tendency of forming a gutter within itself, not needing one determined beforehand... Great importance must be attached to the mudspate as a geological factor. Firstly as to volume, on which point I quote Conway: "Where I was, all through Nagar and Hunza, the mud-avalanche is by far the most important agent in forming the miscellaneous deposits you meet with in the valleys. The moraines, even though enormous, are secondary, and are not so big as the enormous deposits produced by the mud-avalanches.. .and the volume of debris that they must have brought down in the fortnight was enormous, far greater in proportion that anything carried by the glacier in a similar length of time" (Rickmers 1913, p. 193-198).

Renewed interest in Grand Canyon geology

Although the 1920s brought little new information on Grand Canyon

geomorphology, it was a decade of renewed exploration and documentation of the

Colorado River following the expeditions of Powell and Stanton. Later Grand Canyon research focused mainly on the potential for developing the water resources of the 57 southwestern U.S.; including flood control and storage of the Colorado River. Claude H.

Birdseye and Eugene C. LaRue, of the USGS, conducted extensive topographic surveys of the canyons of the Colorado River during that decade in a systematic search for suitable dam sites. Birdseye's (1924) work provided the first quantitative profile of the river, and demonstrated that the river's fall occurred mainly through Grand Canyon as a series of abrupt drops within rapids.

As part of Birdseye's 1921-23 river expeditions, LaRue photographed most of the

Colorado River; inadvertently duplicating many images originally taken by the Powell and

Stanton parties. The later photographs showed changes caused by a variety of processes along the river corridor, including debris flows that occurred during the preceding half- century. Birdseye's expeditions completed the first fully-instrumented survey the profile of the Colorado River. In addition to his survey data, Birdseye was also accompanied on the

1923 trip by the first geologist to travel the river since Powell -- Raymond C. Moore.

Moore reported on the geological suitability of proposed dam sites as well as many other subjects (Birdseye and Moore 1924).

More than a decade before Birdseye and Moore (1924) made their trips through

Grand Canyon, Eliot Blackwelder had observed and noted debris flows in Utah. Based on those studies, Blackwelder (1928) produced one of the important early papers outlining the geological importance of debris flows in the semi-arid regions of the western U.S.

Blackwelder's article describes a debris flow in Utah that he witnessed. In this landmark paper, Blackwelder emphasized how rarely debris-flow deposits and processes were recognized in the field at that time, and how little on debris flows was taught to young geologists in universities. In his landmark 1928 paper entitled Mudflow as a geologic agent in the semiarid mountains, Blackwelder describes how poorly-documented debris 58 flows were before 1930:

Nevertheless the geologic importance of mudflows has been but little appreciated by most geologists and physiographers. They are not given much space in most textbooks of geology. The writer found mudflows mentioned in only six representative modern textbooks and manuals and only two gave to them as much as a paragraph. Even in those two books they appeared to be regarded as unusual phenomena, confined chiefly to such mountains as the Alps and the Himalayas and to volcanic eruptions. It is safe to assume that students in university courses in geology learn very little about mudflows, and hence most geologists have but very little initial acquaintance with this important element in land erosion. Yet in our semiarid and even in our arid regions they are not rare and peculiar phenomena, but normal agencies of gradation, which aid greatly in forming the debris fans that border many high mountain ranges (Blackwelder 1928, p. 466).

Although Blackwelder emphasized the importance of debris flows, or "mudflows,"

as he referred to them, in mountainous terrains, he failed to mention the occurrence of

debris flow in other settings dominated by inverted topography, such as Grand Canyon.

The reason for this is likely related to the reasons mentioned above, such as inaccessibility

to Grand Canyon tributaries and a lack of navigation of the Colorado River. If Blackwelder had traversed the Colorado River with Birdseye and Moore in the early 1920s,

then he might have become the first geologist to recognize debris flows as key in the

formation of Grand Canyon rapids.

Blackwelder made several important points in his 1928 paper. Of these, the

following is probably most relevant regarding observations of Grand Canyon rapid

formation made by Powell and others several decades earlier: These deposits [referring here to his observations of mudflows] have generally been supposed to be formed by floods, although floods normally assort their loads to some degree and leave their deposits perceptibly stratified. The products of stream action are gravel, sand, or silt, not a bouldery earth or clay. Although short, swift floods may carry boulders and may deposit a poorly classified mixture of particles, their product is not devoid of bedding. Furthermore, it is difficult to understand how a thin flood, that is not confined in a definite channel, can sweep boulders weighing hundreds of tons several miles out on a plain with a gradient of only 4 to 6 degrees. It is now known that these unstratified deposits are the product, not of ordinary water floods, but of mudflows (Blackwelder 1928, p. 471). 59

Such insights would have been a revelation to earlier scientists such as Trowbridge, Powell and Dutton, as well as many others. Even today, many Earth science professionals fail to recognize the importance of debris flows as a sediment-transport agent in shaping environments throughout the U.S. and other areas (Blair and McPherson 1994).

Following Blackwelder's work, several other important descriptive articles on debris flows and their deposits appeared in the geologic literature during the 1940s and

1950s. With respect to understanding Grand Canyon debris flows, the most interesting and significant papers of this period were published by Robert P. Sharp following observations that he made of mudflows in Alaska and California (Sharp 1942; Sharp and

Nobles 1953). These articles provide excellent accounts of debris flows in both alpine and semi-arid settings. Sharp is linked to Grand Canyon Geology in that he was also a member of the 1937 Colorado River expedition led by Ian Campbell of the California Institute of

Technology to study Pre-Cambrian rocks of Grand Canyon.

Recent correspondence with Sharp revealed that he and other scientists on the 1937 expedition accepted Powell's stated relationship between Grand Canyon rapids and

Colorado River tributaries, but also indicate that none or the members of the 1937 party attributed debris flows to formation of the river's whitewater. In a letter to the author describing the 1937 river expedition, Sharp states:

When we went down the [Colorado] river in 1937 I was just finishing up my Ph.D. thesis work in the Ruby-East Humboldt Range of NE Nevada. At that time I considered myself a budding structural geologist (my thesis ended up being about half structure and half geomorphology) and I don't recall ever having heard of debris flows. We accepted the rapids as made by bouldery debris dumped into the Colorado by its tributaries. McKee [Edwin D.] and I never talked about them, and other geologists (Campbell, Maxson, and Stark) on the trip were focused on the Pre-Cambrian rocks. I must admit that the thought of a role debris flows might have played [controlling river geomorphology] never entered my mind... 60

My interest in debris flowage as a major landform process was sparked by a magnificent debris-flow levee near our camp in the Icefield Ranges along the border of the Yukon Territory and Alaska in the summer of 1941...

Wolf Creek Valley (now named Steele) had a stagnant glacier in its lower part replete with lots of muddy, water soaked (highly mobile) glacial till. I sat one day watching small mud-flow lobes moving down a slope. It was then that I realized how the debris-flow levee near our camp formed...

I think that appreciation for debris flows had no single spark; it just grew naturally by accumulation over many years. In this country Blackwelder was certainly a major influence. (R.P. Sharp, written correspondence, 1995).

Sharp's candid reflections of his own early failure to recognize debris flows and their role in creating Grand Canyon rapids provide a good example of the importance of first-hand observations to earth scientists.

Recent studies of rapids and debris flows in Grand Canyon

The earliest description of a Grand Canyon tributary flood attached to the terms

"debris flow or mud flow" appeared in a guidebook to the Colorado River published by

Troy L. Péwé:

A modern debris flow or mudflow deposit can be seen an eighth of a mile upstream along the old trail leading from the modem Lees Ferry. On September 13, 1967 a large local convectional storm dropped considerable precipitation in the area, and out of one of the gulleys from the large cliff to the north of Lees Ferry poured a debris-boulder-mud flow. Blocks of Navajo sandstone and other rocks 1 to 15 feet [0.3 - 4.6 m] in diameter were carried down the short gulley into the Colorado River, partly destroying a wooden dock. Natural levees about 4 feet [1.2 m]wide were formed on either side of the mud flow and incorporate many of the large boulders.

This mud flow breaches the old road that has been there since 1871 and appears to be the only time that a mud flow has come across this road from these gulleys in about the last 100 years. One of the important points to consider is that this information reinforces the idea that a great many of the geological happenings occur over short intervals of time with long periods of quiescence in between. This is especially true in arid regions. A further note which increases the value of this mud flow observation is that such mud flows on a larger scale must have occurred in the past in all of the tributary canyons of the Colorado River (Péwé 1968, p. 15). 61

Péwé's recognition of debris flow as an important agent in Grand Canyon geology and the geomorphology of the Colorado River likely derived from a wide range of professional experience as a geologist throughout the southwestern U.S. and in Alaska where debris flows occur frequently. Péwé 's interest in debris flows along the Colorado River was also likely stimulated by the recent occurrence of several of these events in Grand Canyon during December, 1966.

During the late 1960s, Leopold (1969) conducted research on the rapids of the

Colorado River in Grand Canyon. Based on field measurements and observations made during one river trip, Leopold offered an explanation for the equal-spaced distribution of rapids throughout Grand Canyon. His explanation was based on the idea that packets of coarse sediment were slowly transported through the river system over time under conditions of "quasi-equilibrium," and that they were maintained at equidistant intervals

(non-random) along the river to equalize energy loss. Leopold maintained that rapids are caused by the combination of large, periodic river floods and sediment inputs from upstream tributaries. According to Leopold, rapids form at locations where these coarse sediments piled up along the river channel; the particles remaining almost stationary through time. Smaller materials would be transported through the system over these bouldery blockages by more frequent flows. In his 1969 report, Leopold states that some rapids are created by direct inputs of debris from tributaries, but makes no mention of specific types of floods that occur in tributaries. Leopold later states in his report that:

Many rapids, however, do not seem to be explained by the three types of circumstances mentioned [debris fan constrictions, bedrock constrictions, and rockfalls]. Rather, they are associated with what seems at first glance to be a random occurrence of gravel accumulations, either as a central bar across the channel or as the channelward extension of a lateral gravel bar. In fact, these gravel accumulations are not random when viewed in terms of a long reach of channel. They have a roughly regular spacing as has long been observed in the occurrence of 62

gravel riffles in small streams (Leopold 1969, p. 136).

In his attempt to explain most Colorado River rapids through the concept of quasi-

equilibrium, Leopold failed to recognize the almost one-to one spatial relationship between

Grand Canyon rapids and tributary confluences. Leopold (1969) correctly identified that

more than half of the river's fall to sea level occurred within Grand Canyon rapids and that they were likely stable over geologic time. However, he dismissed the influence of

bedrock geology, and further failed to recognize the significance of tributary debris flows

in creating and maintaining these important geomorphic features of the Colorado River.

This failure may be explained by Leopold's relatively limited experience working in Grand

Canyon, and by his application of sediment-transport principles developed on streams and

rivers not dominated by debris-flow processes. By his own admission, Leopold states:

The main difference between the bed topography in a deep narrow gorge like the Grand Canyon and the common gravel-bedded stream of less mountainous areas is a matter of scale (Leopold 1969, p. 144).

The other main difference that Leopold failed to mention was that the geomorphic

framework of the Lower Colorado River through Grand Canyon is dominated by tributary

debris flows; a fact recognized by later workers (Webb et al. 1988; Cooley et al. 1977).

Another early mention of "mudflow" in association with Grand Canyon erosion

also appeared in an article by Ford et al. (1974) entitled "Rock Movement and Mass

Wastage in the Grand Canyon." In this article, the authors describe various mass

movement processes occurring in Grand Canyon, and summarize some of the known events that occurred during the 20th century. They provide a brief discussion of the characteristics of mudflow and make the distinction between mudflow-type processes and those of streamflow flash floods; citing the earlier observations of Grand Canyon debris 63 flows by Péwé (1968) and Billingsley (Ford et al. 1974, P. 121) Details of Grand Canyon debris flows were first documented in the scientific literature by Cooley et al. (1977). In their landmark paper entitled "Effects of the

Catastrophic Flood of December, 1966, North Rim Area, Eastern Grand Canyon,

Arizona," Cooley et al. (1977) identified flood deposits as being produced by "mudflows" in several Grand Canyon tributaries following an extreme winter storm in 1966. Two of the reported debris flows were large enough to reach the Colorado River. The most dramatic effects of debris flow were reported from Crystal Creek, a large tributary located at river mile 98.3-L (fig. 1.1). The 1966 Crystal Creek debris flow created one of the worst navigational hazards on the river when hundreds of large boulders were deposited by the debris flow, creating Crystal Rapid (Cooley et al. 1977). study, accounts By the time Cooley et al. (1977) began their 1968 Grand Canyon and and descriptions of debris flows throughout the western U.S. were more widely frequently published. Besides those mentioned above, key publications on debris flows, included: mudflows, alluvial fans and alluvial-fan processes published by the mid 1960s Beatty 1963; Blissenbach Allen 1965; Anderson and Hussey 1962; Anstey 1965; 1966; Curry 1966; Denny 1954; Bonney 1902; Bull 1963; 1964; Chawner 1935; Croft 1962; and Foster 1965; 1967; Eckis 1928; Grove 1953; Hooke 1967; Leggett et al. 1966; Mason Pack 1923; Scrivenor 1929; 1956; McKee et al. 1967; Mullineaux and Crandell 1962; 1943; and Winder 1965. Besides Sharpe 1938; Shreve 1968; Waldron 1967; Wentworth insightful recognition and having access to increasing sources on debris flows, the et al. (1977) also owed to their interpretation of Grand Canyon debris flows by Cooley helicopters. increased access to several Grand Canyon tributaries by use of 64

Like Péwé, Cooley et al. (1977) recognized that debris flows influenced the geomorphology of the Colorado River locally at rapids. However, they did not emphasize the idea these local controls at many locations along the mainstem were dominant in molding the geomorphic framework of the Colorado River through Grand Canyon; a conclusion that Péwé (1968) had mentioned in passing in his guidebook. In addition, the initial peak-discharge estimates reported by Cooley et al. (1977) for several of the 1966 debris flows were greatly overestimated owing to their misuse of the manning equation for estimating these non-Newtonian flood flows (Webb et al. 1989). At the time that Cooley et al. (1977) began their 1968 study, little was known about methods for accurately estimating discharge of debris flows; this task remains problematic. the Following the study by Cooley et al. (1977), Dolan et al. (1978) documented rapids along the influence of regional geologic structures on the spacing and occurrence of

Colorado River. They recognized that the occurrence of large tributaries were mainly the river, controlled by regional faults, and that these tributaries delivered coarse debris to et al. (1978) did where alluvial fans and large rapids formed as a result. However, Dolan boulders from not include details of the processes responsible for transport of large

tributaries to the mainstem. the distribution Following after the above-mentioned studies, Graf (1979) analyzed provided another explanation for the of rapids throughout the Colorado River, and concluded that rapids in the whitewater. Based on statistical analyses of their spacing, Graf other reaches. Graf system were randomly distributed in some reaches, and nearly so in stability of downstream (1980), also studied the effects of Flaming Gorge Dam on the Like Leopold, Graf also dismissed the tributary- rapids in the Upper Colorado River basin. source explanation for rapid spacing because some rapids did not coincide with debris 65 sources. This is mainly true in the Upper Colorado River basin where Graf concluded that rapids were likely the remnant sediment deposits of a former river flood regime and climate. In Grand Canyon, Graf concluded that tributary sources did not adequately explain formation of rapids because many tributaries supporting debris fans do not have rapids. However, Graf failed to recognize the significance of debris flows in rapid creation there, ignoring the fact that nearly all rapids in Grand Canyon can be attributed directly to tributary inputs of coarse sediment delivered by tributary debris flows. At any rate, Graf's

1979 paper makes no mention of debris flows as a key mechanism for transporting coarse debris to the Colorado River. Regarding rapids attributed to tributary sources, Graf states:

In some cases, such as Gray and Desolation Canyons, most of the rapids result from fluvial tributary processes, while in other cases, especially Cataract Canyon, mass movement predominates. Of the 410 rapids in the major canyons, 278 are collocated with tributary streams, and approximately 100 are collocated with mass movement sites. The remainder are not located at any obvious source of debris. This wide-ranging test of the tributary hypothesis in the Colorado and Green River systems reveals two major problems. First, there are many rapids not located at an obvious debris source. Material forming these rapids must have been moved by floods on the main streams. Second, some tributaries have rapids at their outlets, while similar streams nearby do not. Thus, the tributary hypothesis provides a useful explanation for the debris in rapids, but an incomplete explanation of the distribution of rapids as geomorphic features (Graf 1979, p. 534).

The shortcoming of both the studies by Leopold and Graf may be their desire to provide a general theory for all rapids on rivers, despite scale and reach-varied geomorphic conditions of process, such as the dominant influence of debris flows in Grand Canyon.

Further, Graf's explanation of rapids based on his analysis of force and resistance included only reaches of the Colorado River in the upper basin.

Era of environmental research on the Lower Colorado River

Knowledge of the geomorphic processes at work on the Colorado River in Grand 66

Canyon was increased greatly during the 1970s by the research of Howard and Dolan

(1981; 1979) and Dolan et al. (1974; 1978). These early environmental studies focused mainly on impacts of flow regulation on sand bars through Grand Canyon National Park below Glen Canyon Dam, but also provided some key insights on the overall geomorphic framework of the river, including the spacing and creation of rapids, mentioned above.

As part of their inventory on the conditions of sand bars, Howard and Dolan (1979) were the first to provide a system-wide assessment of tributary flooding impacts on river resources using aerial photographs; including impacts to existing debris fans and rapids. In their 1979 report, Howard and Dolan (1979) state that between 1965 and 1973 (years of the two sets of aerial photographs existed) 27 percent of all tributary alluvial fans aggraded owing to tributary floods. However, they reported that only 10 percent of the alluvial fan constrictions were increased by 15 m or more, and stated that "Catastrophic narrowing and steepening of rapids is very uncommon, the most notable example being the creation of a major rapid at Crystal Creek (river mile 98.3-R; fig. 1.1) in the mid 1960s (Howard and

Dolan 1979, p. 851). However, like other researchers before them, Howard and Dolan

(1979) also failed to mention tributary debris flows as being a dominant process controlling the river. Also, they did not differentiate debris flows from streamflow floods with respect to delivery of coarse debris to the Colorado River and tributary-fan building.

In a later landmark paper entitled "Geomorphology of the Colorado River in the

Grand Canyon," Howard and Dolan (1981) provide the first integrated view of the

Colorado River through Grand Canyon, including details of both river and tributary processes important to conditions of river resources below Glen Canyon Dam. A section in their paper entitled "Alluvial Fan Boulders," provides the best description up to that time on the relationships between natural and regulated flows, the role of regional structure on 67 river geomorphology, tributary influences and formation of alluvial fans and rapids in

Grand Canyon. Howard and Dolan (1981) do not specifically mention debris flows and their role in rapid formation, but they do cite Cooley et al. (1977), mentioning the impacts of the 1966 flood at Crystal Creek and its modification of Crystal Rapid. This citation indicates that Howard and Dolan were aware that "mud flows" had been documented in

Grand Canyon during just prior to their studies. Whether or not Howard and Dolan realized the significance of debris flows as the primary agent in transporting large boulders from tributaries to the river is not clear from their 1981 paper.

One of the important conclusions from Colorado River studies of this period by

Howard and Dolan is the importance of constricting tributary debris fans in the short and long-term storage of coarse and fine sediment in and near eddies below Glen Canyon Dam

(Howard and Dolan 1981). The important balance between mainstem sediment transport and the combined local conditions of the mainstem channel and debris fans is clearly presented in their 1981 paper:

Because most of the drop in the river occurs in rapids where debris is contributed by tributaries and rockfalls, the overall river gradient is primarily determined by the balance between addition and removal of coarse debris. The rate of addition depends upon the size, quantity, and location of the supplied debris, together with the frequency of tributary flooding or rockfall. The rate of removal depends upon the in situ comminution of the boulders to a transportable size by abrasion, breakage, and chemical weathering. This equilibrium is statistical rather than exact, for both addition and removal of alluvial fan debris occurs sporadically during floods. The characteristic time scale of this balance of addition and removal must be on the order of a thousand years or more. The most distinguishing feature of the equilibrium of fan debris is that the balance is not between sediment input and its unmodified transport our of the system, as in the case of fine sediment, but rather between input and weathering or erosion to a finer grain size.

The morphological context of the debris-fan rapids along with the structurally controlled channel width constitute the major constraints upon transport and deposition of cobbles. The cobble bars, reworked primarily during major floods, have an intermediate time scale of adjustment, probably measured in hundreds of years. Both comminution and downstream transport are involved in the balance of rpmrycoll nf n-raxnal 9r,r1 qrp riprix7g.r1 hnth frnm vrrIrtrprim nrl lc-sr-1111r frrym 68

tributary floods, rockfalls, and comminution of fan debris. The fans, bars, and channel width constraints, in turn, provide the morphological framework for transport and deposition of fine sediment (Howard and Dolan 1981, p. 295).

Valuable insights on the geomorphic framework of the Colorado River provided by

Howard and Dolan (1981), such as the above, laid a foundation for later environmentally-

focused studies of the river throughout the 1980s and 1990s.

In regard to Leopold's (1969, p. 135) earlier statements on the geomorphic

framework of the Colorado River with respect to sediment transport, Howard and Dolan state:

This least-work hypothesis has been successfully applied by several geomorphologists and most directly for sediment transport by Kirkby (1977). However, the existence of morphological constraints may severely limit the efficiency of transport (Chang 1979). The overall channel gradient is dominated by the fall in rapids, so that the river profile is not primarily adjusted for transport of the most abundant size components; that is sand and finer sizes. By contrast, the former low gradient of the river system upstream from Lees Ferry suggests that section was, in fact, adjusted primarily for sand transport (Howard and Dolan 1981, p. 296).

This idea that system-wide river hydrology and sediment transport were controlled by

localized debris fans was new in the early 1980s, and it paved the way for later studies on

sand bar evolution, sedimentology and geomorphology by other researchers during implementation of the Glen Canyon Environmental Studies (GCES) program in 1983. The idea that tributary debris fans were integral in controlling mainstem hydrology and sediment transport was carried forth by Webb et al. (1987; 1988; 1989; 1996) beginning in 1984.

Continuing concerns about the effects of fluctuating releases from Glen Canyon

Dam on downstream physical resources of the Colorado River led to additional river research during the 1980s. Studies on dam effects were sponsored by Bureau of

Reclamation and conducted by several agencies and academic institutions during the period 69

1983 to 1989 under GCES, Phase I. Much of the sediment research during this period was

conducted by the USGS through the Water Resources Division's Arizona District and

National Research Program. As part of these studies, and following after the earlier work

of Cooley et al. (1977), Webb et al. (1988; 1989) made detailed studies of Grand Canyon

debris flow and their effects on the rapids of the Colorado River below Glen Canyon Dam.

During the 1980s, Kieffer (1985; 1987; 1990) made detailed studies of hydraulic

and physical properties of twelve Grand Canyon rapids. In her 1987 report, Kieffer

specifically mentions tributary debris flows as being responsible for transport of large

boulders to the Colorado River, and recognized their system-wide potential for creating and

modifying rapids. However, her studies were focused on characteristics of mainstem flow

through rapids rather than on the role of tributary debris flows in controlling the

geomorphology and sediment transport of the river. In her 1985 paper, Kieffer also

presents the first schematic time series showing a conceptual model for debris-fan evolution

through time. This model is based on interactions between additions to alluvial fans from

tributary debris flows and reworking of such deposits by mainstem floods. Her conceptual

model also includes estimates for upper discharges required to produce fan and channel

morphology observed throughout the system. These estimates were produced using principles of theoretical hydraulics combined with particle-size distributions measured and

several coarsened alluvial fans (Kieffer 1985).

Webb et al. (1987; 1989) made new estimates for the 1966 peak discharges formerly reported by Cooley et al. (1977) at Crystal Creek using alternative methods based on superelevation and runup of flowing mud around bends and against obstacles in tributary channels (Webb et al. 1989). They also studied more recent debris flows that followed the 1966 events, and used repeat photography as a means of estimating 70 geomorphic changes at rapids (Webb et al. 1988). This second period of debris-flow research provided valuable new information on the impacts of debris flows on river resources, and the limiting effects of reduced flows on erosion of new alluvial fans. Webb et al. (1989) benefitted from having access to many tributaries through fieldwork conducted during several raft-supported river trips. By examining many tributaries from the river, they were able to recognize the overwhelming evidence of debris-flow impacts along the river. One of the conclusions made by Webb et al. (1989) was that debris flows were the key sediment-transporting mechanism for delivery of large boulders from hundreds of tributaries to the river. Their study was the first to recognize that debris-flows were mainly responsible for the alluvial fans and rapids, and that this hillslope process fundamentally controlled the geomorphic framework of the river (Webb et al. 1988; 1989). Investigations by Webb et al. (1989) led the way for ever-more detailed and system-wide studies of Grand Canyon debris flows during the 1990s.

GCES, Phase I studies of debris flows and their impacts to downstream resources below Glen Canyon Dam ended in 1988. However, information from Webb et al. (1987;

1988; 1989) debris-flow studies had important management implications related to operation of Glen Canyon Dam, and preservation of downstream resources. In summer,

1989, an environmental impact statement on the downstream impacts of operation of Glen

Canyon Dam was mandated by the Department of Interior (GCD-EIS). These studies included continued investigations on debris flows and their influence on the geomorphology of the Colorado River. It was at that time that the author became involved in the phase of investigations reported in this volume.

Studies carried out as part of the GCD-EIS are referred to as GCES, Phase II, and were completed in 1995 (DOI 1995). Debris-flow studies during that period expanded on 71 the previous work of Webb et al. (1989) and included investigation of ongoing debris-flow impacts, documentation of historic debris-flow frequency over the last century, climate characteristics associated with debris flows, a detailed assessment of the reach-varied geomorphology of the Colorado River between Glen Canyon Dam and Lake Mead, debris- fan reworking and the effects of managed high releases on aggrading debris fans, and modeling of debris-flow probabilities in all Grand Canyon tributaries.

During GCES, Phase II, Hereford et al. (1996) made detailed stratigraphic and sedimentologic studies of large alluvial fans preserved in the wider reaches of Grand

Canyon. The impetus for these studies was the potential impacts of dam operations on archeological sites preserved in flood terraces along the river, and on alluvial fans of the mainstem. Hereford et al. (1996) report studies of debris fans associated with the

Palisades Creek alluvial fan and vicinity (river mile 65.5-L; fig. 1.1), and provide weathering and archaeologically age constraints for fan surfaces.

Most narrow reaches of the Colorado River through Grand Canyon do not preserve large alluvial fans owing to reworking of large pre-dam floods. From their studies,

Hereford et al. (1996) concluded that most active-fan surfaces were mid-Holocene or younger in age. Their studies also provided a conceptual model for punctuated alluvial fan building based on alluvial chronology derived from associations with archeological remains. The conclusions of Hereford et al. (1996) provide a somewhat longer historical perspective on the role of debris flows in Grand Canyon during late Holocene time; infoimation that complements more recent information on events described in this in later chapters in this volume.

Presently, a program of long-term monitoring and research of resources of the

Colorado River below Glen Canyon Dam is being implemented through the Grand Canyon 72

Monitoring and Research Center, based in Flagstaff, Arizona. Establishment of the Center is called for in the Grand Canyon Protection Act of 1992, and as a component of Adaptive

Management of the Lower Colorado River (DOT 1995, p. 37). As part of this long-term effort, the database on debris flows and their impacts to the geomorphology of the river will continue to grow as additional events occur and their impacts are studied.

In addition, rephotography of images available in the existing photographic database of the mainstem will allow for ongoing documentation of geomorphic changes through time. Much of the information presented in the following chapters contributes to this long-term database on tributary debris flows in Grand Canyon. Hopefully, this growing database will be useful to future researchers and managers working to understand and preserve river processes and resources in Grand Canyon National Park.

In conclusion, the history of our understanding of the mass-wasting process of debris flow in Grand Canyon and elsewhere is notable for its slow development. The knowledge base on debris flows is one that started with chance observations, descriptions and often incorrect interpretations. In Grand Canyon, identification of debris flows as an important, system-wide geomorphic process has depended upon: the timing of an unusually rare storm (December, 1966), increased access to the tributaries of the Colorado

River, an accumulated body of published observations of debris flows worldwide, experimental research, and the construction of a large dam on the mainstem -- Glen Canyon

Dam. Hopefully, continued interest and monitoring of mass-wasting processes in Grand

Canyon, like debris flows, adds to the growing knowledge base on this interesting sediment-transport process. 73

CHAPTER 3

FREQUENCY, MAGNITUDE, AND INITIATION OF HISTORIC

GRAND CANYON DEBRIS FLOWS

Abstract

Debris flows in 529 ungaged tributaries throughout Grand Canyon, Arizona create and maintain hundreds of alluvial fans and associated rapids that control the geomorphic framework of the Colorado River. Interpretation of 1,107 historical photographs spanning

125 years, yielded debris-flow frequency data in 169 of the 529 tributaries (32 percent) studied. Repeat photography reveal that of these 169 tributaries, 112 produced debris flows that reached the river since 1872. Combined data from photographic and non- photographic sources further indicate that a minimum total of 120 debris flows have impacted the Colorado River since 1872. If representative, these results suggest a debris- flow frequency ranging from one every 10 to 15 years in certain eastern tributaries, to less than one per century in other drainage basins. Overall, debris flows may recur approximately every 30 to 50 years in individual tributaries. Debris-flow peak discharges, estimated in 18 drainage areas for events that occurred between 1939 and 1997, typically range from about 100 to 300 m3/s and contain 15 to 30 percent sand, and 10 to 25 percent water by-weight. The largest Grand Canyon debris flow since 1872 occurred in 1939 and had a peak discharge of 1,000 m3/s. Grand Canyon debris flows are initiated by hillslope failures that occur during intense rainfall often associated with infrequent multi-day storms.

The most common initiation mechanism in Grand Canyon, termed the "fire-hose effect," occurs when runoff pours over cliffs onto unconsolidated colluvium. 74

INTRODUCTION

Debris flows are mixtures of clay to boulder-sized sediments and water; typically, the volumetric water concentration ranges from about 10 to 30 percent, making them an extremely significant sediment-transport process (Pierson and Costa 1987; Major and

Pierson 1992). Debris flows are characterized by cohesive properties that are probably related to clay content, sand content, grain-particle interactions which result in shear strength, and the ability to transport large boulders (Rodine and Johnson 1976; Johnson and Rodine 1984; Costa 1984). Many debris flows are thought to move as laminar flowing plugs (Enos 1977; Johnson and Rodine 1984), and under certain conditions (i.e., in steep channels) may be highly erosive (Pierson 1980). In low-gradient channels, debris flows typically are non-erosional and depositional; however, all of these types of floods can cause erosion in steep channels (Thomas Pierson, U.S. Geological Survey, oral commun.,

1991). Debris flows are also an important mass-wasting process in the formation of alluvial fans in a variety of geomorphic settings throughout the world (Blair and

McPherson 1994). Costa and Jarrett (1981) describe debris flows as water-based slurries of poorly sorted material ranging in size from clay to boulders that exhibit non-Newtonian flow behavior.

Debris flows occur in arid, semi-arid, tropical, montane and alpine environments

(Brunsden and Prior 1984; Blair and McPherson 1994). In these settings, debris flows are typically called mudflows, debris slides, earthflows and debris torrents, to name a few of the more common terms (Blackwelder 1928; Sharp and Nobles 1953; Johnson and Rodine 1984; Pierson 1984; Pierson and Costa 1987; Cooley et al. 1977). When this type of flash-flood results from a volcanic eruption, it is termed a lahar (Pierson and Scott 1985; 75

Scott 1988). Debris flows frequently have devastating effects on populated areas (Pierson et al. 1990), but damage can be severe even in sparsely populated areas of the southwestern

U.S. (e.g., Glancy and Harmsen 1975). Blair and McPherson (1994) provide a historical summary of debris-flow studies and processes related to alluvial-fan formation throughout the world.

The complex flow properties associated with debris flows, including their ability to transport large boulders long distances on relatively gentle gradients (Blackwelder 1928;

Rodine and Johnson 1976; Johnson and Rodine 1984; Beatty 1989), continue to intrigue

Earth scientists and challenge their understanding of channelized mass-wasting processes.

Although much remains to be learned about debris-flow behavior and their significance in the geologic record, their occurrence and importance along the Colorado River in Grand

Canyon National Park and vicinity is generally recognized (Péwé 1968; Ford et al. 1974;

Cooley et al. 1977; Howard and Dolan 1979; 1981; Kieffer 1985; 1990; Webb et al. 1987;

1988; 1989; Melis and Webb 1993; Melis et al. 1994; Webb and Melis 1995; Griffiths et al.

1996; Hereford et al. 1996; Melis et al. 1996; Webb 1996; Webb et al. 1996).

Debris flows are a key erosion and sediment-transport process in the ongoing excavation of Grand Canyon tributaries, and their resulting alluvial fans control the hydraulics of the Colorado River over significant reaches (Webb et al. 1989; Melis and

Webb 1993; Melis et al. 1994; DOI 1995; Schmidt and Rubin 1995; Hereford et al. 1996).

Debris flows occur in at least 529 geomorphically-significant tributaries of the Colorado

River in Grand Canyon between Lees Ferry and Diamond Creek, Arizona (Webb et al.

1989; Melis et al. 1994; Melis et al. 1996; Webb et al. 1996; fig. 1.1). Fine-grained river sediments below Glen Canyon Dam have been limited by the existence of Lake Powell (fig.

1.1) since 1963, but remain important as the foundation of the river's ecosystem (DOT 76

1995). Debris flows provide an important source of sediment to the regulated river

downstream from the dam. More importantly, debris flows transport large boulders from

tributaries to the mainstem. Boulders that exceed the competence of river flows create and

maintain stable alluvial fans and associated rapids at hundreds of sites along the river (Melis

and Webb 1993). These alluvial fans constrict mainstem flow and create zones of

recirculating flow, termed eddies complexes, in the river channel where fine sediments are

trapped and stored as sand bars (Hamblin and Rigby 1968; Schmidt and Graf 1990;

Schmidt and Rubin 1995; fig. 3.1). As a result, changes in the number, spacing and

geometry of debris fans, in combination with local channel characteristics and sediment

inputs, directly affects storage of fine sediment along the river (Hamblin and Rigby 1968;

Dolan et al. 1974; Howard and Dolan 1981; Kieffer 1985; Webb et al. 1989; Schmidt and

Graf 1990; Melis and Webb 1993; Schmidt and Rubin 1995; Webb 1996).

In Grand Canyon, debris flows can modify the river corridor in a number of ways.

Debris flows can cause: 1) aggradation of debris fans, 2) erosion of debris fans, 3)

increased constriction of the river channel, 4) an increase or decrease in the severity of

rapids, 5) disturbance of existing vegetation on debris fans and, 6) erosion or burial of existing sand bars, and 7) changed patterns of deposition and erosion of sand bars caused by altered flow patterns around recently-aggraded debris fans. By depositing boulders that exceed the competence of river discharges, debris flows also affect the navigability of rapids along the Colorado River that attract more than 20,000 whitewater enthusiasts annually (Stevens 1990). Predicting short and long-term impacts of debris flows on the river channel is difficult because individual eddies have complex characteristics related to both debris-fan geometries and local to reach varied channel conditions. In most cases, reduced annual peak discharges in the river since 1963 have allowed the more frequent, but 77

n \\... \ ,.....;. 17.... 7 ,. • 7•

"t 1

ri -

- I (Modified from Hamblin and Rigby, 1968)

Explanation

1. Tributary debris fan 2. Rapid controlled by large immobile boulders 3. Debris bar (synonymous with "island" or "rock garden") 4. Riffle or rapid caused by debris bar

FIGURE 3.1. The morphology of a typical debris fan, rapid and eddy complex found along the Colorado River in Grand Canyon, Arizona (from Melis et al. 1994). 78 smaller debris flows to aggrade alluvial fans. Such aggradation increases the severity of rapids and alters sediment-storage characteristics of low velocity pools located upstream (Melis et al. 1994; Melis et al. 1995).

The numerous debris fans in Grand Canyon create a relatively unique geomorphic framework for the Colorado River that has been termed "debris-fan dominated" by recent workers (Schmidt and Rubin 1995). Because constricting alluvial fans control flow and sediment storage in the river (Schmidt and Rubin 1995), it is important to understand the spatial and temporal patterns of the debris flows that create, modify, and maintain them over time. Accurate estimates of debris-flow frequency, magnitude, and initiation are important when estimating a sediment balance for the regulated mainstem. Because sediment inputs are now most limited in reaches closest to Glen Canyon Dam, accurate sediment budgeting is critical to the management of sediment and related resources below Lake Powell in Grand Canyon National Park.

Debris-flow producing tributaries throughout Grand Canyon National Park and vicinity are unique because they have remained largely unaffected by large-scale human impacts since modem development of the western U.S. As a result, changes in the frequency and magnitude of sediment-transport events in these drainage areas are almost exclusively related to climatic factors and sediment availability. Because most tributaries in

Grand Canyon National Park are likely to remain unaltered by humans, long -tenu i monitoring of sediment deliveries to the Colorado River by debris flows may provide an important source of information on geomorphic responses to high-frequency climate changes in the future.

Before regulation, infrequent debris flows that contributed silt to boulder-sized sediment to the Colorado River went mostly unnoticed. Sediment input to the mainstem 79

varied greatly, but was more than sufficient to sustain sediment resources in all years

during that period (Andrews 1991), and aggraded debris-fan constrictions were likely

reworked during annual flood peaks (O'Connor et al. 1994; Webb et al. 1996). Since

closure of Glen Canyon Dam in 1963, debris flows have played an increasingly important

geomorphic role relative to sediment inputs and direct impacts on downstream riverine

resources (Melis et al. 1994; DOI 1995). Since regulation, reduced annual peak flows have

allowed debris fans to aggrade, and these accumulated deposits are now rarely reworked

and eroded. As a result, ongoing debris flows have increased navigational hazards in the

river's rapids, changed patterns of sediment storage and deposition within eddy complexes,

and altered the riverine ecosystem in ways that are poorly understood (Melis et al. 1994;

Melis et al. 1995; Valdez and Ryel 1996).

Development of a sediment budget for the Colorado River through Grand Canyon

is highly desirable for river management purposes, but relies in part, on an estimate of

long-term sediment volumes yielded from hundreds of ungaged, mostly-ephemeral

tributaries during infrequent flooding "events" that include both debris flows and

streamflows. Estimating the sediment contribution from debris flows in these tributaries

depends on estimates of long-term event frequency and magnitude for as many tributaries

as possible. Estimating flood magnitude for debris flows is somewhat more complicated than it is for streamfloods. For streamflows, magnitude is usually reported in terms of peak discharge (stream power) and flow duration (net volume) of mostly water. However, debris-flow magnitude is a combination of peak discharge and duration for flows that are mostly sediment. Thus, parameters of particle-size and water content must also be considered, and all such data are inherently difficult to obtain directly in Grand Canyon tributaries. 80

While sediment inputs from streamflows can be measured by streamgages on major

tributaries, or be estimated for ungaged tributaries using empirical formulas, an accurate

estimate of inputs from debris flows in ungaged, bedrock tributaries is a more difficult

challenge. This is true mainly because debris flows correlate poorly with annual

precipitation data for records near Grand Canyon (Griffiths et al. 1996). Increased

information on debris-flow magnitude, frequency and initiation in Grand Canyon can

improve sediment-balance estimates, and contribute to preservation efforts aimed at

managing Lower Colorado River resources for long-term sustainability. In addition, future

trends of debris-flow occurrence can also provide valuable information on geomorphic

responses to long and short-term climate variability. Few other areas of the world contain

as many "pristine" drainage areas where the influence of human activities have been

minimized. Only recently, have the full implications of debris flows and their ability to

control flow and sediment storage of the Colorado River been recognized (Webb et al.

1988; Melis et al. 1994; Melis et al. 1995; Webb 1996).

Previous related work

Only three previous studies have attempted to document debris-flow frequency and

magnitude with any detail in Grand Canyon (fig. 1.1); only two of those studies report

details of initiation mechanisms for such events (Webb et al. 1989; Melis et al. 1994).

Cooley et al. (1977) report debris flows that occurred in December, 1966 in several tributaries of the Colorado River, including Lava Canyon and Bright Angel and Crystal

Creeks (river miles 65.5-R, 87.8-R and 98.2-R, downstream from Lees Ferry; fig. 1.1).

They made a crude estimate of the magnitude (peak discharge) of the debris flow in Dragon

Creek, a tributary of Crystal Creek (river mile 98.2-R) using the Manning equation, and 81 inferred minimal frequency information (recurrence interval) from damage to archaeological sites dated there. This work brought attention to the fact that debris flows occurred in

Grand Canyon, and that they could exert impacts on the Colorado River over short time periods.

Webb et al. (1987; 1988; 1989) report more detailed and accurate magnitude and frequency data for three Colorado River tributaries previously studied by Cooley et al.

(1977), and provide additional frequency data from radiometric and dendrochronologie dates associated with debris-flow stratigraphies and tree-ring scars in tributaries (river miles

65.5-R; 93.5-L and 98.2-R). Webb et al. (1989) were also first to use historic photographs to document historical debris flows along the Colorado River, and inferred debris-flow behavior from preserved alluvial-fan deposits. The initial use of historical photographs to document debris-flow induced changes along the Colorado River, and river-management concerns led to a more extensive program to document frequency using repeat photography and other methods beginning in 1989. Based on this later work, Melis et al. (1994) reported detailed information on the timing and magnitude of major debris flows that significantly impacted the Colorado River between 1872 and 1994. Melis et al.

(1994) also revised and updated the list of historic debris flows and streamfloods in Grand

Canyon originally published by Webb et al. in 1989 (Appendix A).

Historic Grand Canyon debris flows

I define historic Grand Canyon debris flows here, as those occurring after 1872 that reached the Colorado River and caused detectable geomorphic changes in the mainstem.

To be detected, this set of debris flows must have deposited sufficient sediment in the river

(in the range of 1,000 m3 or more), to have been: 1) recorded in replicates of historic 82 photographic views; 2) noticed and recorded in written or oral accounts (diaries, reports by eyewitnesses); or 3) have deposited sediments that were preserved with detectable levels of

137Cs (indicating a post-1952 origin; see methods section).

Historic data on debris flows in Grand Canyon are limited and difficult to obtain because of the isolated setting and inaccessibility of many of the river's tributaries; especially before about 1960. By evaluating hundreds of tributary confluences using a variety of field methods, including repeat photography, it was possible to assess the frequency, magnitude, initiation mechanisms and impacts of historic debris flows on the

Colorado River between 1872 and 1997. For the most recent debris flows (post-1966), it was often possible for workers, including the author, to examine debris-flow deposits within hours to weeks following events.

In summary, this chapter presents relatively high-resolution frequency

(occurrence(s) in the last century) and magnitude (peak discharge and minimum total sediment volume) data for historic debris flows (1872 to 1997) for Colorado River tributaries in Grand Canyon National Park and vicinity, Arizona. It also includes information on the documented initiation mechanisms responsible for recent debris flows from 1939 to 1997, related to the unique stratigraphy and source-area settings found in

Grand Canyon (fig. 3.2), as well as trends and discreet characteristics of regional and local precipitation. 83 Colorado River Mile Formation First Encountered

Chink Formation Moenkopi Formation Kaibab Formation Toroweap Formation Coconino Sandstone

Hennit Shale 4.9 Esplanade Sandstone 11.4 Wescogame Formation 15 0 Manakacha Formation Watahomigi Formation Surprise Canyon Formation Redwall Limestone

Temple Butte Limestone 37.7 Unclassified Dolomite 35.1 Muav Limestone

Bright Angel Shale

Tapeats Sandstone Great Unc-onformity

Kwagunt Formation

Galeros Formation Colorado River

Dox Sandstone Formation 63.0

74.8 76.2 77.0

Zoroaster Granite

FIGURE 3.2. A stratigraphie column showing rocks exposed in Grand Canyon, Arizona, and the distances in river miles downstream from Lees Ferry, Arizona where they first appear along the Colorado River (from Melis et al. 1994). 84

METHODS

Designating drainage areas

To assign frequency and magnitude data to specific drainage basins in Grand

Canyon National Park, I systematically identified and designated geomorphically significant tributaries between Lees Ferry and Diamond Creek (fig. 1.1; Appendix B). My definition of "geomorphically-significant" tributaries includes numerous small drainage areas that have the potential to produce debris flows that affect the geomorphology of the

Colorado River channel. The criteria for designating drainage areas were determined from analysis of 38 U.S. Geological Survey 7.5-minute quadrangle maps of the river corridor, and 106 maps of the Grand Canyon region. Included were all tributaries that: 1) have drainage areas larger than 0.01 km2; 2) are mapped as having perennial or ephemeral streams; 3) were previously designated with an official name; 4) clearly terminate at the

Colorado River in a single channel; and 5) contribute to formation of obvious debris fans and (or) rapids.

The Paria and Little Colorado Rivers, as well as Kanab and Havasu Creeks were excluded from the list of geomorphically-significant tributaries because they already have gaging stations with fairly lengthy flow records, and their sediment-transport histories and drainage characteristics have been previously studied by the USGS. While all four of these drainage areas exhibit evidence of past debris flows, only Kanab Creek (river mile 143.5-

R) is included in this assessment of debris-flow frequency. Appendix B lists the geomorphically-significant tributary canyons of the Colorado River in Grand Canyon

National Park and vicinity, including information on drainage areas, reference maps, rapid ratings (Stevens 1990), locations, as well as official and unofficial place names assigned. 85

Additional drainage-basin characteristics for these tributaries were also derived from maps for use in related studies of debris-flow potential by Griffiths et al. 1996 (Appendix C).

Many areas that could not be designated as geomorphically-significant using the criteria outlined above were designated as "E" or "extra" drainage areas. This type of drainage area also contributes sediment to the Colorado River and typically consists of steep slopes with no identifiable channel on 7.5-minute quadrangle maps. Based on slope and drainage characteristics of these areas, I conclude that sediment contributed from these drainage areas is usually delivered to the Colorado River by overland sheet-flow or is transported by small debris flows originating as slope failures near the river. In addition, these smaller areas do not form obvious basins with terminal flow channels. Drainage areas of geomorphically-significant tributaries were estimated by hand from hand-drawn outlines digitized on 7.5-minute quadrangle topographic maps as noted in Appendix B.

The Colorado River through Grand Canyon National Park has been previously subdivided into 11 study reaches between river miles 0 and 225 mainly on the basis of first-appearance of bedrock type at river level, and the general geomorphology of the river channel associated with given lithologies (DOI 1989; Schmidt and Graf 1990). Schmidt and Graf (1990) used first appearance of key strata at river level as the main criterion for designating geomorphic river reaches (fig. 3.2), but did not incorporate other related data on features of the channel (such as debris fan geometry) into their criteria for designating geomorphic reaches. Despite its limitations, for this report, I retain the previous geomorphic classification of the Colorado River established by Schmidt and Graf (1990) owing to the fact that it has been referred to extensively in previously published reports

(DOT 1995). 86

Evidence for Grand Canyon debris flows and associated flow regimes

Tributary streamflows, hyperconcentrated flows and debris flows are commonly associated with one-another in Grand Canyon. Normal streamflow (Pierson and Costa

1987) typically constitutes the initial and recessional phases of Grand Canyon debris flows, although streamflow floods may also precede debris flows in some cases (Cooley et al.

1977). Evidence of normal streamflow includes deposition of well-sorted, clast-supported sediment on channel margins and in the bed, and the presence of well-delineated zones of organic debris along high-water lines.

If the properties of debris flow and normal streamflow are thought of as endpoints of a spectrum, then these can be contrasted by the properties of hyperconcentrated flow, which also has been documented in Grand Canyon (Webb et al. 1989; Melis et al 1994).

Hyperconcentrated flows contain between 20 and 60 percent water by volume (Beverage and Culbertson 1964; Pierson and Costa 1987). These flows result in deposits with faint but distinctive sedimentary structures (mostly faint, laminar bedding), and unique sorting characteristics (coarsening-upward texture commonly containing erratic cobbles and boulders). Some researchers believe hyperconcentrated flows behave as quasi-Newtonian fluids (Pierson and Scott 1985; Pierson and Costa 1987).

Four types of debris-flow evidence are commonly preserved in tributaries of the

Colorado River. These types of evidence consist of: 1) debris levees composed of poorly- sorted sediments deposited along channel margins; 2) continuous or discontinuous mudlines preserved on overhanging walls; 3) damaged plants on the sides of the channel that are scarred, buried, or surrounded by poorly sorted sediment; and 4) continuous or discontinuous muddy scour lines preserved on hillslope channel margins, channel walls or terraces. Often, evidence of debris flows may be difficult to discern when larger-stage 87 streamfloods scour channel and rework debris-flow deposits. Intact debris-flow levees containing fine-grained matrix sediment or debris levees reworked into lines of boulders are the most reliable indication of a recent debris flow. Evidence of ancient debris flows typically consists of levee deposits from which the original matrix sediment has been removed by rainfall, fluvial reworking by streamflows, and (or) weathering of boulder surfaces (Bull 1991, p. 116-118). The presence of a bouldery, poorly-sorted debris fan is

a near-certain indicator that a debris flow has occurred in a tributary; deltaic deposits resulting from streamflow floods are typically finer and graded. In addition, badly-

damaged trees, or trees sheared-off at ground level in or near the tributary channel, are also

a likely sign of debris-flow disturbance.

Based on the examination of 78 drainage areas in the eleven Schmidt and Graf (1990) study reaches for recent or ancient evidence of debris flow processes, Melis et al. (1994) concluded that historic debris flows have occurred in all of these geomorphically-

derived reaches. Preserved ancient debris-flow sediments reported by Melis et al. (1994) in all 78 tributaries suggest that debris flows are pervasive throughout Grand Canyon

tributaries.

Documenting debris-flow frequency, magnitude, and initiation

Repeat photography has been used successfully in previous geological, biological,

and related environmental studies in Grand Canyon and throughout the western U.S.

(Leopold 1951; Hastings and Turner 1965; Turner and Karpiscak 1980; Stephens and

Shoemaker 1987; Webb et al. 1989; Webb et al. 1991a; Melis et al. 1994; Griffiths et al. 1996; Melis et al. 1996; Webb 1996). Most of the frequency information for historic

debris flows (1872 to 1997) reported here, were obtained from systematic, repeat 88 photography and interpretation of historical photographs between 1989 and 1997.

Examination of abundant replicated historical photographs of the Colorado River corridor, dating as far back as 1872, allowed me to study a large subset of debris fans and rapids for changes caused by debris flows, river-reworking associated with mainstem floods, and other geomorphic processes such as rockfall (Ford et al. 1974). During this study, 1,107 oblique, historic photographs of the river corridor between river mile 0 and 225 (fig. 1.1) were matched and interpreted to determine significant changes to tributary channels, debris fans, rapids, sand bars, and riparian vegetation (Appendixes D, E and F). These historical

photographs and their replicates were studied and detailed notes were recorded both in the

field and office in an attempt to extract the maximum amount of information available. By

using historical photographs showing specific debris fans at different times, I was often

able to bracket when debris flows occurred in selected tributaries. For some tributaries, the

dates of debris flows could be determined to within 1 year. Debris flows since 1984 could

usually be related to specific dates based on eyewitness accounts, precipitation data, and

comparisons of aerial photographs of the river.

By far, the most valuable collection of photographs used in this study was that of

Franklin A. Nims and Robert Brewster Stanton, taken between 1889 and 1890 and

consisting of 445 views of the river made at approximately 2-km intervals (Appendix D).

A second set of 171 photographs of the Colorado River were taken during the USGS dam-

site survey expedition of 1923, led by Claude Birdseye. These mostly panoramic views

were taken by Eugene C. LaRue, but from lower vantage points than the Nims/Stanton

views (Appendix E). Because of the introduction and spread of non-native tamarisk trees (Tamarix chinensis) along the shores of the Colorado River below Glen Canyon Dam,

these photographs were often difficult to locate and match, and were, therefore, often of 89 only limited use. A total of 491 other photographs taken by many photographers at various times were also examined to obtain temporal coverage of specific debris fans in Grand

Canyon (Appendix F).

Several sets of low-elevation aerial photographs taken between 1935 and 1997 were also examined for evidence of historic debris flows. In 1935, the Soil Conservation

Service took black and white aerial photographs of Marble Canyon (river miles 0 to 61) and

Diamond Creek to Lake Mead (river miles 225 to 280) at a scale of 1:31,800; these photographs are stored at the National Archives in Arlington, Virginia. Another set of photographs, taken in November, 1935, under the direction of John Maxon of the

California Institute of Technology, recorded parts of eastern Grand Canyon from the vicinity of Bright Angel Creek to Specter Chasm (river miles 87 to 129) and western Grand

Canyon from about river mile 211 to Lake Mead (river mile 280) in 1938. The scale of these photographs is unknown (probably about 1:20,000). The 1965 aerial photography is available from the EROS Data Center in Sioux Falls, South Dakota, and 1973 aerial photography is stored at the USGS in Tucson, Arizona. Aerial photography flown on many dates between 1980 and 1997 are available for viewing on request at the Grand

Canyon Monitoring and Research Center, located in Flagstaff, Arizona.

Frequency data for the more recent historic debris flows (post-1952) were also

obtained through analyses of debris-flow sediments for their 137Cs activity. A by-product

of fallout from above-ground thermonuclear testing, the detection of 137Cs activity in

debris-flow deposits should indicate a post-1952 deposition; this is most-probable in arid

environments where soil leaching is minimal. This technique is useful for delineating the

time range of historic floods (Ely et al. 1992). For example, in cases where change was

caused by a debris flow that occurred since 1890 (for example, based on a Stanton 90 photograph), 137Cs analysis was useful in establishing whether the debris flow occurred between 1890 and 1952 or between 1952 and the date of sampling. The interior sediments of pre-1952 debris-flow deposits should not give a detect for 137Cs if leaching is not a factor, however, surface samples collected from such deposits should give a positive detection result. Ely et al. (1992) found that leaching of radionuclides from the surface to interior sediments was not a significant contamination problem in fine-grained flood

deposits. Samples from recent-looking debris fans and debris levees were routinely

collected and analyzed for 137Cs activity in fine-grained sediments (matrix clay to silt) to

date debris flows. To estimate the accuracy of this relatively new technique (and

previously untested for use with debris-flow sediments), silt samples were extracted from

the surfaces and interiors of debris-flow deposits of known age (post-1952 as determined

by evaluation of aerial photography) and analyzed for 137Cs activity. Sediments analyzed

for 1 37Cs were sand-size-or finer; in several cases, I analyzed concentrated silt and clay

fractions separated by dry sieving.

The opportunity and likelihood of directly measuring the velocity, discharge or total

volume of a debris flow in Grand Canyon is very limited. During a unique field experience

in spring, 1995, I heard a large debris flow pass by near my camp in the middle of the

night (Webb and Melis 1995). At daybreak I observed a newly formed debris fan

constricting Lava Falls Rapid (river mile 179.4). However, the only flow that I witnessed

in Prospect Creek (river mile 179.3-L), was a recessional muddy flood that lasted nearly 12

hours.

Debris flows occur infrequently in most tributaries and are seldom witnessed (see

chapter 2). Because of this fact, peak discharges of debris flows were indirectly estimated 91

overhanging cliffs; the arid climate of Grand Canyon allows for long-term preservation of

these mudlines making the estimation of peak discharge for prehistoric debris flows

possible in some cases. Scour lines also provide evidence for high-water marks where

debris flows have occurred. Plants—particularly cacti—at the maximum-flow elevation

frequently preserve mud, gravel, and other types of debris that can be used to reconstruct

the maximum stage of the debris flow. Exposed roots and freshly abraded tree trunks are

commonly found in channels through which debris flows have passed. These types of

flow-elevation evidence can provide data for estimating debris-flow peak discharge.

Bedrock tributaries in Grand Canyon are characteristically well-suited to preserving

mudlines used for estimating peak discharges of debris flows. These bedrock tributaries

typically have more than one channel bend along their course where evidence of

superelevation and runup are preserved, making it possible to compare several sites for

consistency of estimates and (or) changes in flow characteristics. Ideal sites have had

minimal scour during debris flows, such as bedrock-controlled channels with relatively

uniform reaches in which continuous mudlines are preserved on both sides of the channel.

Because the mechanics of debris flows are complicated, standard methods used to

estimate peak discharge for streamflows are not applicable to debris flows. Manning's

equation is inappropriate for estimating peak discharges of debris flows because it is

developed only for Newtonian fluids (Webb et al. 1989); debris flows behave as non-

Newtonian fluids (Costa and Jarrett 1981). Manning's n (a frictional coefficient) does not

adequately reflect energy losses in debris flow associated with channel roughness

(Antonious Laenen, U.S. Geological Survey, written commun., 1986; Laenen et al. 1987).

Debris flows have complex interactions among particles that create internal shear-strength, cause energy losses, and support large particles in suspension (Rodine and Johnson 1976). 92

During flow through channel bends, the surface of a flowing fluid typically rises on the outside of the bend and drops on the inside of the bend (Apmann 1973; Costa 1984).

The difference in flow elevation between the inside and outside of bends is termed

"superelevation" (Apmann 1973; Costa 1984). Superelevated flow is caused by centripetal forces being exerted on the flowing fluid mass as it moves around the bend (fig. 3.3). The mean fluid-flow velocity determined using the superelevation method (Vs) is related to the difference between the raised elevation of flow on the outside of the bend and the lower elevation of flow on the inside of the bend by

V s = (g • R c AHslW)° . 5 , (3.1)

where g = acceleration due to gravity (9.8 m/s 2), R, = the radius of bend curvature along the channel's centerline, AHs = the difference in cross-channel surface elevation at the point

of maximum superelevation, and the channel's top width at the point of maximum

superelevation. Use of equation (3.1) requires the assumption be made that the flow is

steady, uniform, and that all flow lines are parallel to one another as flow proceeds through

the bend (plug flow). Because debris flows are viscous slurries, this assumption may be justified.

When a flowing fluid encounters a stationary obstruction in its flow path (typically

a tree trunk, vertical wall, bridge pier, etc.), the momentum of the fluid forces it to "runup"

the immovable object. The height of runup is related to the peak-flow velocity using a

method similar to that of superelevation. In both cases, it is assumed that the peak velocity

at a point on the bedrock wall is close to the mean velocity of the flow; often a tenuous 93

• • . • • . •

...FLOW :::LEVEE

0 20 FEET 1 O 10 METERS

EDGE OF WATERFALL

FIGURE 3.3. Plan view of a typical channel bend showing superelevated-flow evidence used to indirectly estimate peak discharges of Grand Canyon debris flows (from Melis et al. 1994). 94

assumption. In Grand Canyon, many debris-flow runups occur in tight corners (nearly 90-

degree bends) where the flow encounters the channel wall nearly straight-on.

The mean velocity of flow at a runup site (V,.) is estimated using

V r = (2 • g . AHr)).5, (3.2)

where Al-1,-= the difference between the maximum runup and unobstructed flow-surface

elevation (usually a continuous mudline up and downstream from the runup obstruction).

Evidence of runup typically is also found at a superelevation site, allowing for estimates of

debris flow velocity by both methods (eqs. 3.1 and 3.2).

In Grand Canyon tributaries, evidence of runup is usually preserved in tight corners

of bedrock channels, or on rock buttresses protruding into the channel. As a result, many

locations preserve evidence of superelevation and runup in close proximity to one another.

In many cases, the peak-flow velocities estimated using superelevation and runup evidence

were quite similar (Webb et al. 1989). Pierson (1985) compared measured surface

velocities with indirect estimates of mean velocity. These velocities presumably are similar

if plug flow exists; he found that mean velocities estimated using indirect techniques

averaged 15 percent lower than actually-measured flow-surface velocities. If the plug-flow

analogy (Johnson and Rodine 1984) is valid for Grand Canyon debris flows, then

Pierson's work suggests that the velocities reported here may be underestimated.

Iverson et al. (1994) showed experimentally that peak-flow velocities for debris slurries estimated from superelevated mudlines were within 30 percent of measured values.

Their experiments also showed that velocity estimates vary widely based on the value of channel width used in superelevation equations. Melis et al. (1994) and Webb et al. (1989) 95 also showed that peak-discharge estimates made at superelevation sites varied significantly depending on where cross-sectional areas were measured relative to maximum superelevation sites.

In this study, debris-flow elevations on both sides of the channel were surveyed to fully encompass channel bends. The profiles of flow-surface elevations were compared to determine superelevation and (or) runup through the channel reach (fig. 3.4). Indirect- discharge estimates were subjectively evaluated as being good, fair, or poor based on several site characteristics, such as whether continuous evidence of debris flow on both sides of the channel was uniform, and whether the indirect site was bedrock-controlled or not. Bedrock-controlled cross sections were used for peak discharge estimations whenever possible. The points of maximum superelevation and runup were identified and superelevation and runup equations were used to estimate peak velocities at these points.

At sites where velocity estimates could be determined by both superelevation and runup methods, both of the estimates were used to determine a range in velocity and discharge.

Peak discharges were estimated using

Qp = A - V , (3.3)

where Qp = instantaneous peak discharge, A = the cross-sectional area of the flow and V = the mean flow velocity determined using either eqs. 3.1 or 3.2. Cross sections were surveyed above, at, and below the site of maximum superelevation and (or) runup wherever possible. Normally, the area of the upstream cross section(s), where the flow- surface elevation was approximately equal on either side of the channel, were combined with the velocity calculated at the point of maximum superelevation or runup to calculate

o I I I I I I I 1.6 co 96

- 0 0 CO

o

Q I I I I I 1 I o o o o o o o o o 06 6 4 6 CV I— c; a; co" N: O o o o o o a) a) o) ,-- ,-. 7.4. T.' T- 'T. St:13.1.31A1 NI INniva A8V1:11191:IV 3A0EIV NOI1VA313

FIGURE 3.4. Longitudinal variation of the flow-surface profile of the 1990 debris flow

at "Site A" in "Crash Canyon" (river mile 62.6-R), a tributary of the Colorado River in

Grand Canyon (from Melis et al. 1994). 97 peak discharge. Upstream cross sections were preferred because they may better reflect the velocity of the debris flow immediately before encountering the channel bend, where it then expends energy in superelevation and (or) runup. However, each site was individually evaluated for accuracy, resulting in both mean and averaged peak-discharge values at some sites.

Webb et al. (1989) found that cross-sectional areas at sites of maximum- superelevation were commonly 1.5 to 3.5 times greater than areas determined in cross sections up or downstream, assuming a linear cross-channel flow-surface. This can be explained by the fact that the viscous debris slurry does not maintain a linear flow surface as it superelevates around the bend. Photographs of superelevated debris flows taken in central China (Kevin Scott, U.S. Geological Survey, written commun., 1991) show that the surface of superelevated debris slurries are often curved. Experimental results of

Iverson et al. (1994) also support such observations. Therefore, cross-sectional areas

determined at the point of superelevation may be much larger than the actual cross-sectional

area occupied by the slurry. In two cases (a slight constriction occurring in the channel

bend), I found that the cross-sectional area at the point of maximum superelevation was

less than the area measured at cross sections either up or downstream from the bend. I

concluded that the velocity estimated at the maximum superelevation reflected the effect of

the constricted channel rather than the average flow velocity upstream or downstream.

Therefore, I chose to report the peak discharge that occurred at that point in these

exceptional cases.

All measured cross-sectional areas were evaluated at the 31 indirect-discharge

estimate sites visited. The number of acceptable cross sections varied from one at 7 sites to

11 at one site. Commonly, only one cross section in the study site had definite flow- 98 surface evidence and a stable channel bed, therefore, site selection was limited. In the case of one set of measured cross sections, flow elevations were rarely equal on both sides of the channel, and a linear flow-surface had to be assumed. Of the 24 indirect peak discharge estimates where multiple cross sections were surveyed, peak-discharge estimates were calculated from the cross section with the minimal cross-channel elevation difference in 20 cases. At one study site with 11 cross sections, a mean discharge was determined; at two other sites, an averaged value derived from the upstream and downstream cross sections were chosen.

Besides peak discharge, other factors to consider when estimating the magnitude of debris flows are total sediment volume deposited at the Colorado River (alluvial-fan volumes), the percent ration of water to sediment contained in the debris slurry, and the particle-size distributions of slurries and their resulting deposits. In this study, fan volumes were calculated from topographic surveys of preserved deposits along the mainstem, and average fan thicknesses deteimined from test pits excavated into deposits.

The water content of recent Grand Canyon debris flows (1990 through 1995) at the time they occurred was estimated using dry-sieved samples weighing 5 to 8 kg. All particles with b-axis diameters >16 nun were removed using standard sieves. In a plastic mixing tray, water was added in small increments to the sample (typically less than 100 ml) and mixed well; between additions, the mixture was qualitatively examined for properties similar to those of debris flow. These properties include the ability to flow as a plug, to form lateral levees (indicating internal shear), and to support particles >16 mm. When the mixture first exhibited these properties, the water content was calculated to reflect the driest possible debris-flow mixture. I continued to add water until the mixture no longer had clearly observable debris-flow properties, particularly the ability to support large particles. 99

This water content, the probable upper limit of hyperconcentrated flow (Beverage and

Culbertson 1964; Pierson and Costa 1987), typically occurs at a water content of 40 to 70 percent by weight. I report the range in possible water content at which the debris-flow sediment samples first exhibited debris-flow characteristics.

With respect to riverine resources and channel features of the Colorado River, contributions to the fine-sediment balance of the river are of most interest, especially the percent of sand-sized particles deposited by debris flows. With respect to the whitewater features of the river, the percentages of large cobbles and boulders are significant to long- term trends in rapids. Based on the data of Melis et al. (1994), recent Grand Canyon

debris flows vary widely in their particle-size distributions, and this likely reflects the

variability of source sediments for debris flows in tributaries. In general, finer-grained

debris-flow deposits, containing larger percentages of sand and finer sediment, are more

easily reworked by river flows. Reworking of debris fans re-distributes sand in the

channel for deposition on sand bars. Debris-flow deposits contain varied percentages of

large boulders that tend to aggrade existing fans and make them more highly resistant to

river reworking, especially with respect to reduced annual peak discharges associated with

regulated river releases. Hence, variability in particle-size distributions of historic debris

flows in Grand Canyon influences the extent to which rapids change, the volume of fine-

sediment added to the mainstem, and the degree to which aggraded alluvial fans resist

erosion by the Colorado River.

Representative samples of Grand Canyon debris-flow deposits for laboratory

sieving cannot be easily collected because a prohibitively large sample weight is necessary

to accurately account for boulders. Collection of such large samples is normally impossible

owing to the isolated nature of the study areas and the limitations of transporting heavy 100 samples out of Grand Canyon. Therefore, I used several methods in combination with limited sample collection to estimate the particle-size distribution of Grand Canyon debris flow deposits.

Point counts were made in the field on debris flow deposits containing particles that

are too large to easily transport to the laboratory for standard dry-sieve analysis (Wolman,

1954; Rice and Church 1996). For deposits judged to be representative, measuring tapes

were stretched over the surface of the deposit at parallel intervals to form a transect

sampling grid. The length of the transect was variable but typically ranged from 50 to 100

m, and the distance between adjacent tapes was generally 0.25 to 0.50 m. At this interval,

any large particle directly beneath the tape mark (at 0.25-m intervals) was selected and

measured; particles smaller than gravel were selected by their closest proximity to the mark

when choice was required. The standard unit of measurement of particles is 0, defined as

D = 2 - 1), (3.3)

where D = the diameter, in millimeters, of the intermediate (b-axis) of the particle. During

point counts, the b-axes were measured and aggregated into one of the cp, categories for

particles >2 mm in diameter. For example, a fine-grained deposit might have particles in

the categories of 2 to 4 mm (-1 to -2 (I) ), 4 to 8 mm (-2 to -3 0 ), 8 to 16 mm (-3 to -4 0),

and larger (Folk, 1974). Typically, 100 to 400 particles were measured during each point

count. 101

The percentage of sand included in debris-flow deposits was determined using standard techniques (Wolman 1954; Kellerhals and Bray, 1971; Folk, 1974; Rice and

Church 1996). Representative debris-fan matrix samples were air dried, then dried again in laboratory ovens overnight at 50 to 100° C. All particles >64 mm were removed from the sample after carefully retaining finer particles. The b-axes of these particles were measured, their individual weights were recorded and segregated by phi class (0), and, in

most cases, their lithologies were also identified. The particle-size distribution for samples

with particles <64 mm and >0.063 mm in diameter were determined using standard, brass

sieves in sets with 1 0 intervals between sieves. Particles with b-axis diameters between 16

mm and 64 min were removed by hand sieving the sample through 16 and 32 mm sieves

and the retained particles were weighed. The total sample weight was calculated as the sum

of the weights of particles >64 mm, particles between 16 and 64 mm, and particles <16

mm. The sample with particles <16 mm was then split into manageable subsamples of a

maximum of 400 g for standard dry-sieve analysis. A mechanical shaker was used for 15

minutes to segregate particles on sets of sieves. The sets had sieves with meshes of 0.063,

0.125, 0.25, 0.5, 1.0, 2.0, 4, 8, and 16 mm. Particles retained on each sieve were

weighed and the percent of the subsample in each 0 class was determined.

The particle-size distribution was then determined by reconstructing the percentage

of particles in each 0 class on the basis of sample weight or by occurrence in point counts.

If particle diameters were measured in the field, the particle-size distribution determined

using sieve analysis was adjusted for these particles after the particle weight was calculated.

If point counts also were made on the surface of the deposit from which the sample was

collected, the two types of data were combined. Although point counts are made using 102 surface exposure and dry-sieve analyses are based on weight percent of a sample, the order of magnitude of the resulting percentages is similar (Kellerhals and Bray, 1971). I assumed that point counts accurately measure particles in excess of 64 mm; therefore, the particle-size distribution >64 mm was determined using point counts, whereas the particle- size distribution <64 mm was determined by combining point count and dry-sieve data.

The percentage of particles <64 mm determined by point count was adjusted by the particle- size distribution of the collected sample.

I also measured the ten largest boulders that were obviously transported to the mainstem by recent debris flows. I determined whether large boulders that could have been present in the channel before the debris flow were transported by the debris flow by examining aerial or oblique photography whenever possible. I also examined limestone boulders for fresh percussion marks on their downstream sides and recently killed organic debris pinned beneath them. The a- (long), b- (intermediate), and c- (short) axes of the boulders were measured and their lithologies and shapes were recorded. The boulder volumes were estimated using equations for variously shaped solids. Typically, boulders were either shaped as rectangular solids, right-triangular solids, cylinders or ellipsoids. In estimating the weight of measured boulders, I assumed an average density of 2.65 g/cm3 regardless of source lithology.

Following recent Grand Canyon debris flows (1939 through 1996), I examined the settings of initiation source areas in tributaries. Areas where obvious slope failures had occurred were studied in terms of sources for sediments and general topographic setting relative to the configuration of the tributary. Initiation sites and mechanisms were then grouped together and types of settings and mechanisms were segregated into four dominant categories based on reconstructed initiation scenarios. 103

The clay mineralogy of source sediments is likely an important contributor to initiation of debris flows in many settings (Beverage and Culbertson 1964; Hampton 1975;

Pierson and Costa 1987). Grand Canyon stratigraphy contains diverse and abundant shale- bearing lithologies that are also thought to be important in debris-flow initiation (Griffiths et al. 1996). Until recently, little information was available on the clay mineralogy of bedrock, particularly shales, exposed in Grand Canyon's walls (Blakey 1990, p. 161).

Layers with large amounts of smectites, particularly montmorillonite, may act as potential slip surfaces and promote slope failures (Melis et al. 1994; Griffith et al. 1996; Webb et al.

1996). Alternatively, other clays minerals, particularly kaolinite, are dispersive and may aid in the rapid disintegration of cohesive sediments during intense rainfall (Webb et al.

1996).

Clay particles finer-than 2 p,rn were analyzed for mineralogy using a common technique (D.M. Hendricks, University of Arizona, written commun. 1988).

Representative samples extracted using the pipet method were saturated with magnesium

(Mg) and potassium (K), respectively, and mounted on glass slides. X-ray diffraction scans were made of the Mg-saturated samples following equilibrium at 54 percent relative humidity and solvating with ethylene glycol. Similar scans were made of the K-saturated samples after heating to 300° and 500° C. The clay mineralogy was interpreted qualitatively and semi-quantitatively from the relative intensities and positions of the peaks obtained from X-ray diffraction.

Local and regional climate anomalies and extreme precipitation events have been shown to be important factors affecting the occurrence of streamflows and floods in

Arizona (Cayan and Webb 1992; Webb and Betancourt 1992) and in Grand Canyon tributaries (Webb et al. 199 lb; Melis et al. 1996). In other settings, debris-flows have also 104

been related to weather patterns, storm types and climate variability, and this information is

often used to design early-warning systems for debris flows (Campbell 1975; Cannon and

Ellen 1985; Wieczorek 1987; Ellen and Wieczorek 1988; Wilson et al. 1993).

For historic Grand Canyon debris flows where I knew the exact or approximate

date of the event from various sources, I used daily rainfall data from nine records near Grand Canyon to calculate storm precipitation by summing over consecutive days with

rainfall preceding the events. For as many events as possible, I estimated the recurrence

probabilities of precipitation events that occurred during the month preceding debris flows

(proxy for antecedent moisture conditions), and the daily and stoim (multi-day)

precipitation directly associated with debris flows (the triggering precipitation), using the modified Gringorten plotting position (U.S. Water Resources Council 1981),

p = ((m - 0.44)/(n + 0.12)) • d, (3.4)

where p = probability of the event, m = the ranking of the event, n = the number of days in

the record, and d = the number of days in the season per year. The recurrence interval, R (yrs), is

R = 1/p. (3.5)

Twentieth-century characteristics of daily precipitation records and stations in the vicinity of Grand Canyon National Park were also examined for information on debris- flow initiation. Monthly and daily precipitation totals associated with the timing of recent debris flows (1939-1996) were also studied to determine the range of precipitation typically 105 associated with events. In addition, the annual frequency of daily precipitation > 25 millimeters (mm) was tabulated for several stations near Grand Canyon to determine if patterns of intense daily precipitation could be identified during the twentieth century that might be related to the timing of recent debris flows, or a perceived increase in debris-flow frequency since about 1960. Trends in the frequency of precipitation >25 mm were tested statistically to determine significance and temporal relation to historic Grand Canyon debris flows.

The non-parametric Kendall Tau-b test was used (SAS, 1993; Gibbons, 1985;

Conover, 1980) to identify statistically significant precipitation trends within the periods

1900-1929, 1930-1959, and 1960-1993. Kendall Tau-b (T) was chosen over the rank correlation coefficient (R) because T approaches normality more rapidly than R (Gibbons,

1985, p. 296). Hence, Kendall Tau-b provided an advantage when analyzing precipitation over the relatively short periods available (<35 years). The largest daily precipitation amounts for nine climatic stations in the vicinity of Grand Canyon were ranked. A storm was defined as the total precipitation over consecutive days; "storms" in the Grand Canyon region can last as few as 2 days and as long as 2 weeks.

RESULTS

Geomorphically significant tributaries of the Colorado River

A total of 529 geomorphically significant tributaries of the Colorado River between

Lees Ferry and Diamond Creek (fig. 1.1; river miles 0 to 225; Plates 1-12) were identified as having the potential to produce debris flows (Appendix B). These tributaries produce relatively-more frequent streamflow floods that also deliver gravel and sand-and-finer 106 sediment to the mainstem. The total drainage area of tributaries of the Colorado River, excluding the Paria and Little Colorado Rivers and Kanab and Havasu Creeks, is 9,516 km2 . The average geomorphically-significant tributary in Grand Canyon has a drainage area of about 18.10 km2 ; the median drainage area is 1.79 km2, and the range is 0.01 to

934 km2 . The amount of drainage area contributing runoff, but without definite channels

("E-areas") is about 408 km2, or approximately 4 percent of the area of geomorphically- significant tributaries. The combined drainage area of geomorphically-significant tributaries is 13 percent of the total contributing drainage area of Grand Canyon, which includes the Paria and Little Colorado Rivers and Kanab and Havasu Creeks.

The drainage area of geomorphically-significant tributaries varies among the eleven study reaches previously delineated by Schmidt and Graf (1990; Table 3.1; Plates 1-12).

Reach-10 tributaries have more than 3,500 km2 of drainage area, or over an order-of- magnitude more than Reach-1 tributaries, which have 268 km2 of drainage area. Reaches

2, 10, and 11 have the largest tributaries, whereas Reaches 5 and 7 have the smallest (Table

3.1). The amount of contributing drainage area of tributaries varies between the left and right sides of the Colorado River through Grand Canyon, mostly reflecting its local to regional geology. In reaches 1, 2, 6, 8, and 11, the total of drainage areas from one side is more than double the area from the other side (Plates 1, 2, 6, 8, and 11; Appendix B).

This asymmetry in contributing area of tributaries results from the slight regional dip (2 to 3 degrees) of Grand Canyon strata to the south; hence, the drainage divides of many south- side tributaries coincide with the South Rim of Grand Canyon National Park and vicinity.

This geologic control on drainage sizes, in part, explains many differences in the frequency and magnitude of debris flows and their associated effects on the geomorphology of the

Colorado River between the eleven reaches (Griffiths et al. 1996). Admittedly, the 107

TABLE 3.1. Summary of study-reach statistics for drainage areas tributary to the Colorado

River within Grand Canyon (from Melis et al. 1994).

Mean Area Mean Area Total Mean of Left-Side of Right-Side Study River Number of Area, Area, Tributaries, Tributaries, Reach Mile Tributaries in (km2) in (km2) in (km2) in (km2)

1 0.0- 11.3 15 268 17.9 8.8 31.5

2 11.3 - 22.6 26 1,468 61.2 17.2 149.0

3 22.6 - 35.9 35 480 13.0 12.4 13.8

4 35.9 - 61.5 66 555 8.4 9.0 7.8

5 61.5 - 77.4 49 222 4.5 4.8 4.2

6 77.4- 117.8 100 1,238 12.4 6.1 19.2

7 117.8- 125.5 23 105 4.4 4.4 4.2

8 125.5 - 140.0 47 425 9.2 2.5 16.6

9 140.0 - 160.0 21 258 12.3 10.7 15.4

10 160.0 - 213.9 113 3,536 31.3 41.7 21.4

11 213.9 - 225.8 30 964 32.1 52.5 8.9

Study-reach designations are from Schmidt and Graf (1990), based on bedrock geology and generally-related morphology of the river channel. Number of tributaries derived from Appendix 2 of Melis et al. (1994). 108 variability of drainage area between study reaches designated by Schmidt and Graf (1990) is greatly influenced by the fact that the reaches also vary significantly in length. This fact, however, also reflects the variability of bedrock geology at river level; a parameter that changes gradually in some reaches while changing more abruptly in others.

The influences of regional geology (structures and basin characteristics, including exposure of varied lithologies), combined with bedrock characteristics at river level that influence channel width, depth, slope, and the frequency and geometry of constricting alluvial fans, are hypothesized as the dominant factors controlling the geomorphic framework of the Colorado River channel throughout Grand Canyon. Griffiths et al.

(1996), development of statistical models that correlate known debris-flow occurrence over the last century in eastern and western Grand Canyon with key basin-area attributes, supports at least part of this hypothesis (see chapter 5).

Debris-flow frequency

Debris fans at the mouths of 169 tributaries were analyzed for changes using 1,107 replicated photographs of the river corridor; an example of how replicate photographs revealed debris-flow evidence at Crystal Creek (river mile 98.2-R) is shown in figure 3.5.

All original views used to determine debris-flow frequency were taken between 1872 and

1980. These data, representing 32 percent of the geomorphically-significant tributaries of the Colorado River in Grand Canyon, provide the primary means for determining the frequency of historic debris flows. A total of 142 tributaries were recorded in replicate photographs spanning at least a century. The remaining 27 tributaries were recorded in photographs taken mostly in 1923. Sixteen small chutes in E-areas also were recorded in the photographs. A 109

B

FIGURE 3.5. Replicate photographs showing changes in Crystal Rapid caused by a large debris flow from Crystal Creek. A. Stanton's 1890 view (#491), of the confluence of

Crystal Creek (river mile 98.2-R), and B. the Colorado River, showing the condition of Crystal Rapid after it was aggraded by a debris flow in December, 1966. 110

TABLE 3.2. Sites and approximate dates of major historic debris flows that impacted the

Colorado River since 1872 in Grand Canyon (modified after Melis et al. 1994).

Name of Name of River Date(s) of Tributary Canyon Rapid Mile Side Debris Flow(s)

Jackass Canyon Badger Creek 7.9 L 1994 Badger Canyon Badger Creek 7.9- R 1897 to 1909 Soap Creek Soap Creek 11.2 R 1923 to 1934 House Rock Canyon House Rock 16.8 R 1966 to 1973 Unnamed canyon Unnamed riffle 18.0 L 1987 Unnamed canyon Unnamed riffle 21.5 L 1890 to 1990 Unnamed canyon 24-Mile 24.2 L 1989 Tiger Wash Tiger Wash 26.6 L 1890 to 1990 Unnamed canyon No rapid exists 30.2 R 1989 South Canyon Unnamed riffle 31.6 R 1940 to 1965 Unnamed canyon Unnamed riffle 42.9 L 1983 Tatahoysa Wash "Boulder" 43.2 L 1983 Unnamed canyon New rapid 62.5 R 1990 Unnamed canyon Unnamed riffle 63.3 R 1990 Lava Canyon Lava Canyon 65.5 R 1966 Palisades Creek Lava Canyon 65.5 L 1966, '84, '87, '90, '93 Unnamed canyon No rapid exists 71.2 R 1983 Unnamed canyon Unnamed riffle 72.1 R 1983 75-Mile Creek Nevills 75.5 L 1959, '87, '90 Hance Creek Sockdolager 78.7 L 1890 to 1990 Monument Creek Granite 93.5 L 1960s, '84, '96 Hermit Creek Hermit 95.0 L 1996 Boucher Creek Boucher 96.7 L 1935 to 1965 Crystal Creek Crystal 98.2 R 1966 1987 Waltenberg Canyon Waltenberg 112.2 R 1938 to Unnamed canyon New rapid 127.6 L 1989 to 1923 128-Mile Creek 128-Mile 128.5 R 1890 1989 Specter Chasm Specter 129.0 L 1989 Bedrock Canyon Bedrock 130.5 R 1890 to 1991 Unnamed canyon Unnamed riffle 130.9 L 1890 to 1923 Unnamed canyon Unnamed riffle 133.0 L 1923 to 1942 Kanab Canyon Kanab 143.5 R 1993 Unnamed canyon New rapid 160.8 R 1939, '54, '55, '63, '66, '95 Prospect Canyon Lava Falls 179.4 L 1937 to 1956 205-Mile Canyon 205-Mile 205.5 L 1890 to 1990 Unnamed canyon Unnamed riffle 222.6 L 1872 to 1885, 1984 Diamond Creek Diamond Creek 225.8 L 111

Based on the evaluation of 1,107 historical photographs, at least 37 tributaries have experienced major debris flows in Grand Canyon during 1872 and 1997 that significantly altered the mainstem (Table 3.2). At selected sites in Grand Canyon, other dating methods have provided frequency data for preserved alluvial deposits pre-dating 1872 (see Webb et al. 1989; Melis et al. 1994; Hereford et al. 1996; Webb 1996; Webb et al. 1996), however,

these data are considered pre-historic and are not included here.

In some cases, time series based on repeat photography were developed for

tributaries by collecting historical photographs taken at different times during the last

century. For example, I have examined over 100 historical photographs of Prospect

Canyon (river mile 179.4-L) and Lava Falls Rapid (river mile 179) that allow detailed

reconstruction of debris-flow magnitude and frequency there (Webb et al. 1996). Other

sites with considerable records of repeat photographic include Badger Canyon, Jackass

Creek, Soap Creek, Monument Creek, Palisades Creek, Waltenburg Canyon, Kanab

Creek, and Diamond Creek (river miles 7.9-R, 7.9-L, 11.2-R, 65.5-L, 93.5-L, 112.2-R,

143.5-R, and 225.8-L, respectively; see fig. 1.1).

By analyzing the overall appearances of debris fans seen in historical photographs,

and by comparing the positions of boulders on debris fans and in rapids with replicate

views, I determined whether or not tributaries had debris flows that reached the Colorado

River during the span of time between matched photographs. As a result, I concluded that

112 of the 142 tributaries recorded in photographs (78.9 percent of total photographed) had

experienced historic debris flows at least once between 1872 and 1996 (Plate 13). Since an

additional 8 tributaries are known to have had at least one debris flow based on non-

photographic records, at least 22.7 percent of all the geomorphically-significant tributaries

are known to have had at least one debris flow between 1872 and 1996. 112

Assuming that the subset of frequency data from repeat photography alone are representative of debris-flow occurrence for the entire study area over the time period 1872 through 1996, as many as 417 tributaries may have produced debris flows during that period. This would mean that about 21.2 percent of all geomorphically-significant tributaries in Grand Canyon have a frequency of less than one debris flow per century. In

addition, twelve steep-angle chutes in E-areas that had debris flows during the last century

indicate that these drainage areas are active producers of relatively-frequent, small debris

flows. Based on these results and using system-wide drainage area characteristics, for geomorphically- Griffiths et al. (1996) statistically predicted debris-flow probabilities

significant tributaries in Grand Canyon during the next century.

Analysis of 1 37Cs activity in fine-grained debris-flow sediments appears to (fig. 3.6). equivocally distinguish debris flows into the age ranges of pre- and post-1952 of the deposit Pre-1952 debris flows had no detectable activity of 137Cs in the interior

(depths > 0.20 m). In contrast, most debris flows that occurred after 1952 had detectable as the activities (fig. 3.6). Debris flows initiated in weathered bedrock failures, such al. 1977), may not have Crystal Creek debris flow of December, 1966 (see Cooley et was not in the bedrock source detectable 1 37Cs activities because the human-made isotope fine-grained sediments by fire-hose effects can material. However, debris flows initiated in

be classified as pre- or post-1952 using 137 Cs.

Debris-flow magnitude areas for debris flows that occurred Peak discharges were estimated in 18 drainage Typically, discharges of recent debris flows between 1939 and 1997 in Grand Canyon. Q 113 1--. o CV

---

1-

0 N- CD C)

1- t-e---1 u_ u_ 0 1411, CD cr _ 1-r) 0) >-

NOANVO 103c1S0Eld NI

AAO -U SIHEI3C] 6C6 L. CO CO CY) 30V NMON>INfl JO 1-'

SMOld S11:1830 0S6L - 31:1c1

11111111111 1

0 LC) 0 LC) 0 LC) 0 CO N N- 1"' 1""' d c:5

AlIALLOV LC 1. - O

FIGURE 3.6. The activities of 137Cs in debris-flow deposits in Grand Canyon of known or constrained ages (from Melis et al. 1994). 114

co 1 I i I I ,— 11 c

• a

•

1- *

- -

1 T- 0 0 0 0 Q 0 0 0 0 Q CV 0 CD

1— 'T

CIN003S }:13d SI:13131A1 °ono NI

`30E1VHOSICI -Wad /V1Old SI193

FIGURE 3.7. Comparison of peak discharges of debris flows, and the total drainage area of the tributary, for debris flows caused by different types of storms in Grand Canyon between 1939 and 1994 (from Melis et al. 1994). 115 ranged from about 100 to 300 m3/s (fig. 3.7). The largest historical debris flow to date in

Grand Canyon occurred in Prospect Creek (river mile 179.4-L), and had a peak discharge

of about 1,000 m3/s. Estimates for peak discharges made at superelevation sites can vary

significantly depending on where cross-sectional areas were measured within reaches

(Table 3.3). This finding is similar to that of Webb et al. (1989) and emphasizes the

importance of careful selection of cross sections when estimating peak discharges for

debris flows. However, the extent that estimates of velocity made using superelevation or

runup evidence accurately reflect the mechanics of flowing debris is an unresolved

problem. Based on recent Grand Canyon data, debris-flow velocity is not clearly related to

channel slope at either superelevation or runup sites (fig. 3.8).

Because they contain relatively high ratios of sediment to water, debris flows

deliver significant volumes of sediment to the Colorado River (Table 3.4). The total

volume of sediment delivered to the river channel appears to be related to drainage area and

storm type (fig. 3.9). Relatively few historic debris flows were initiated during the latter

half of the twentieth century during winter frontal storms and autumn dissipating tropical

cyclones. However, these types of storms deposited larger volumes of material on debris

fans than did the more frequent debris flows initiated by summer convective

thunderstorms. This relationship between storm type and debris-flow volume suggests that

storms driven by circulation patterns of the eastern Pacific Ocean may have exerted a more

significant long-term influence on the geomorphology of the Lower Colorado River than

did summer storms driven by summer circulation patterns from Mexico before 1963. However, because mainstem regulation now limits reworking of debris fans, the smaller,

but more frequent summer-storm driven debris flows may now be exerting as much or

more influence on the river's geomorphology as do larger winter driven events. Recent 116

TABLE 3.3. Peak-discharges from cross-sections at (or) near maximum superelevation sites, and from sections upstream and downstream (modified after Melis et al. 1994).

Area Peak Mean of Upstream Mean River Mean at or Near Dis- and/or Down- Peak Mile Velocity Super- charge, stream Cross- Dis- and in, elevation, in sectional Areas, charge,in Site' (mis) in (m2) (m3/s) in (m2) (m3/s)

62.6-R 3.8 42 160 30 110 (Site A) 62.6-R 5.3 81 430 77 400 (Site E) 68.5-L 4.5 110 490 61 270 (Site A) 70.9-L 9.4 21 200 30 280 (Site A) 71.2-R 3.9 111 430 130 500 (Site A) 72.1-R 6.1 16 98 6 37 (Site B) 72.1-R 5.7 86 490 41 230 (Site C) 75.5-L 6.1 116 710 43 260 (Site A) 75.5-L 4.8 31 150 20 100 (Site H) 75.5-L 6.5 30 190 39 250 (Site I) 122.7-L 6.1 80 490 60 370 (Site A) 126.9-L 5.2 27 140 35 180 (Site A) 127.6-L 6.4 180 1,100 65 420 (Site A) 127.6-L 6.9 81 560 48 330 (Site B)

1 See Melis et al. (1994) for specific study site locations within tributaries listed here. 117

CD •

0 e.• t • 0 sp a • op • • •

0 • • 0 •

c6

CIN003S 1:13d SE13131A1 NI `A.L10013A

FIGURE 3.8. Peak velocity of recent Grand Canyon debris flows as a function of local channel slope, derived from superelevation and runup evidence (from Melis et al. 1994). 118

TABLE 3.4. Volumes of recently-aggraded debris fans in Grand Canyon (modified after Melis et al 1994).

Location, Area Estimated Total Peak and of Thickness Volume Estimated Debris- Year that Debris- of Debris- of Debris Sand Flow Debris Flow Fan, Fan, Fan, Volume, Discharge, Occurred in (102 x m2) in (m) in (103 x m3 ) in (103 x m3 ) in (m3/s)

062.6-R ('90) 0.7 1.7 1.2 0.2 300 068.5-L (`93) 3.0 2.0 5.9 1.2 300 070.9-L (`93) 13.1 0.3 3.9 1.6 200 071.2-R (`83) 4.8 1.5 7.2 3.4 500 072.1-R (`83) 3.0 1.5 4.5 2.0 100-200 075.5-L ('90) 10.0 1.2 12.0 1.3 100-300 093.5-L (`84) 14.6 11.5 16.8 12,0.4 1110 098.2-R (`66) 225.0 23.0 275.0 1,211.0 1300 127.3-L (`89) 1.6 5.3 8.5 1.7 300 127.6-L (`89) 2.5 2.0 5.0 0.7 200-400 157.8-R (`93) 3.4 4.0 10.0 - 13.0 2.5 - 5.0 UK 160.8-R (`93) 4.1 3.0 8.0- 12.0 2.0 - 3.0 UK 179.4-L (19,`63, 11.0, 6.2 3.5, 1.9 38.2, 11.8, 4.6, 0.8, UK and '95) 5.6 1.7 9.5 0.8 207.8-L (`91) 0.4 1.0 0.4 0.1 200

'Data from Webb et al. 1989; 2Estimated, based on the reworked size of the debris-fan deposit as it existed in 1992, and typical relations between river-reworked debris fans and non-reworked debris-flow deposits; UK, Unknown. Areas and volumes of debris fan deposits estimated above the approximate stage of 142 m 3/s in the Colorado River. Depth of debris determined from excavated test pits; sand volumes, estimated from particle-size distributions; peak discharge, estimated from superelevation and (or) runup evidence. 119 CC) T- 1 I

• 0 0

o o o 0 0 o 0 o o o o 0

SI:13.131A1 oteno NI VV\01J SI883C1 JO 31/1M0A 1V101

FIGURE 3.9. The relation of total volume of debris-flow deposits on debris fans, to the total drainage area of the tributary, for debris flows initiated by different stoiiii types in

Grand Canyon between 1939 and 1994 (from Melis et al. 1994). 120

Grand Canyon data also indicate that the total volume of sediment-transported to the river is poorly related to the peak discharge of debris flows (fig. 3.10). This relationship might be explained by the fact that larger volume debris flows that occur during more regional autumn and winter storms can be related to multiple slope failures that would unlikely occur simultaneously. Webb et al. (1989) reported up to eleven separate slope failures that contributed to the 1966 Crystal Creek debris flow. Since the chance of sediments from multiple failures combining together at once in the tributary channel are small during a large volume debris flow, it is reasonable to believe that such an event would be characterized by a series of limited peak discharges associated with a large total sediment volume. Likewise, localized summer storms often result in one to a few slope failures that can result in deposition of small volume debris fans associated with peak discharges that are as large or larger than those associated with debris flows resulting from many separate failures.

Often, extended periods of recessional flow lasting up to several hours, follow debris flows in Grand Canyon tributaries and erode fresh looking debris-flow deposits, obscuring evidence of the debris-flow process. Reworking by recessional flows greatly complicates studies of debris flow processes, particularly interpretations of flow behavior based on fan-surface textures and morphologies (see chapter 4).

As stated earlier, particle-size and water content are important factors to consider when estimating and comparing relative debris-flow magnitudes. Debris flows in Grand

Canyon transport very-poorly sorted sediment. Recent deposits generally contain between

15 and 30 percent sand, but the variability of sand content is large among sites (Table 3.5), and the sand content of debris flows is poorly related to peak discharge (fig. 3.11). These relationships are more likely related to local conditions of source-sediment lithology where slope failures originate during debris flows in Grand Canyon than they are with any 121

o 0 0

O O CO

o

4

•

•

0 o o 0 o o o

S1:131301 oieno NI `M01J SIE18301V101. JO 3V\IMOA

FIGURE 3.10. The relation of sediment volume deposited on debris fans to peak discharge of the debris flow in Grand Canyon between 1939 and 1994 (from Melis et al.

1994). 122

TABLE 3.5. Percent sand content (by-weight) of some recent debris-flow deposits in

Grand Canyon (modified after Melis et al. 1994).

Percent Sand Content, Tributary Name (River Mile) Timing of Debris Flow (by-weight)

"18-Mile Wash" (18.0-L) 1987 15 Unnamed tributary at mile 19.9-L 1987 46 Unnamed tributary at mile 62.5-R 1990 13 "Crash Canyon" (mile 62.6-R) 1990 13 Unnamed tributary at mile 63.3-R 1990 20 25 Lava Canyon (65.5-R) 1966 30 35 Tanner Canyon (68.5-L) 1993 25 - 35 Cardenas Creek (70.9-L) 1993 13 41 Unnamed tributary at mile 71.2-R 1983 47 Unnamed tributary at mile 72.1-R 1983 45 75-Mile Creek (75.5-L) 1987 22 75-Mile Creek (75.5-L) 1990 4 Monument Creek (93.5-L) 1984 1 30 - 40 Crystal Creek (98.2-R) 1966 1 10- 15 Unnamed tributary at mile 126.9-L 1989 3 Unnamed tributary at mile 127.3-L 1989 27 "127.6-Mile Canyon" (mile 127.6-L) 1989 15 Unnamed tributary at mile 157.6-R 1993 25 - 35 Unnamed tributary at mile 160.8-R 1993 25 Prospect Canyon (179.4-L) 1939, 1963, 1995 12, 7, 8 Unnamed tributary at mile 207.8-L 1991 6 Unnamed tributary at mile 224.5-L 2 1890-1935 11

'Data from Webb and others, 1989; 2 Timing constrained by repeat of oblique photograph (1890), evaluation of aerial photography (1935), and non-detection of 137Cs from interior of debris-levee sediments (pre-1952). 123

•

• •

0 0 0 0

(ID •ct CO C\1

IN301=19ci IHOI3M NI iNaLNOO CINVS

FIGURE 3.11. The relation of peak discharge to percent sand content (by-weight) for debris flows in Grand Canyon between 1939 and 1994 (from Melis et al. 1994). 124 specific properties associated with debris-flow or other hillslope processes.

Reconstitution of debris-flow samples indicates that most debris flows have a water content of 10 to 25 percent by-weight. Owing to the infrequency of debris flows, sand input to the river from these floods is considered to be comparatively small relative to annual inputs from floods of the Paria and Little Colorado Rivers (Andrews 1991).

However, unlike annual floods from the larger tributaries where floods are predominantly streamflows, debris flows continue to transport large boulders weighing 100 to 300 Mg to the Colorado River, and introduction of boulders from several recent debris flows have created new rapids and (or) have enlarged existing ones in ways that may be relatively permanent (Table 3.2; Appendix A).

The introduction of new boulders can constrict the mainstem in ways that enhance the accumulation of sand and finer sediment in low-velocity pools above rapids that act as check dams. Large boulders that exceed the competence of mainstem flows can increase the elevation of upper pools in ways that change flow patterns upstream in eddy complexes

(Melis et al. 1995). Most recent debris flows have deposited numerous large boulders in the Colorado River that have impacted fans, rapids and upstream eddies (Appendix G).

Results from censusing of the ten largest boulders deposited in the mainstem by recent debris flows, indicates that such particles range in weight from about 1 to 300 Mg at various sites. The lithologies that supplied large boulders to the mainstem during recent debris flows varied depending on basin-varied exposures of strata, but came mostly from the dominant cliff-forming units of Grand Canyon stratigraphy (fig. 3.2; Appendix G).

Particularly important in regard to boulder-source lithology are the Redwall and Muav

Limestones, as well as the Esplanade and Tapeats Sandstones (fig. 3.2). 125

Source areas, initiation mechanisms, clay mineralogy, and precipitation

The occurrence of debris flows is related to the abundance and settings of source sediments in tributaries. Source sediments may either be weathered and jointed bedrock, colluvium, or sediment stored adjacent to channels (most commonly channel alluvium in the form of debris flow levees). The Paleozoic and Proterozoic strata exposed in Grand

Canyon (fig. 3.2) provide a wide variety of source lithologies in a setting of extremely-high topographic relief. As the size and relief of Grand Canyon has increased through time, so has the variety of available debris-flow source sediments exposed by on-going erosion. In addition to weathered and fractured bedrock sources, abundant source sediments are available from Quaternary and Tertiary colluvium (Cooley et al. 1977; Huntoon et al.

1986). Observations of failure scarps, in both weathered bedrock and colluvium, suggests that source areas for debris flows may encompass most of the drainage area in many cases.

Ford et al. (1974) report that slab-failures and rock avalanches occur commonly throughout the Grand Canyon, although the exact frequency of these processes is poorly known. Furthermore, Ford et al. (1974) report that rockfalls generally occur in all seasons, at all times of day, and under a wide range of weather conditions. During my winter field excursions in Grand Canyon, I have noted an increased likelihood for rockfalls to occur during periods of prolonged rain. I witnessed such a period in January, 1993, when several hundred small rockfalls occurred in Marble Canyon during a 12-day period of constant rainfall following several consecutive days of lighter precipitation.

The elevation and temperature gradients of Grand Canyon range widely owing to the extreme topography of the region. The area is also subject to a high degree of annual and inter-annual climatic variability, especially during the warm-season (Graf et al. 1991;

Hereford and Webb 1992). These conditions are conducive to precipitation-related 126 infiltration and frost action that affect large surface areas of bedrock. Generally speaking, prolonged winter precipitation and freezing temperatures favor rockfalls and the formation of talus and colluvium, while convective thunderstorms, driven by monsoonal conditions in summer, promote debris flows by creating short-duration, intense rainfall.

Regional and localized faults, folds, and joints probably play a large role in controlling conditions that promote mass wasting in the canyons of the Colorado River.

Therefore, the degree of structural deformation associated with a given drainage basin can also influence the abundance of source sediments, drainage-area characteristics, and the location of future debris flows. In general, drainage basins that contain extensive faults can be expected to produce larger volumes of sediment available for debris-flow initiation, simply owing to the fracturing and weathering of exposed bedrock. 75-Mile Creek (river mile 75.5-L), which formed along strike of the east-trending 75-Mile Fault, provides a good example of the relation between faulted bedrock and debris-flow frequency. The

asymmetry of this tributary, and its high frequency of debris flows, is related to preferential

formation of drainage areas along the highly-fractured, footwall-side of the 75-Mile Fault.

Three debris flows since 1959 have been initiated exclusively in colluvium accumulated in

these footwall sub-basins.

Debris flows are commonly initiated by slope failures that occur during intense

rainfall. The intensity of rainfall required to initiate debris flows in Grand Canyon is

unknown, but sparse data indicate that a sustained intensity of more than 25 mm/hr and a

total rainfall of 25 to 50 mrn may be a minimum requirement (Webb et al. 1989; Mel'is and

Webb 1993; Melis et al. 1994; Griffiths et al. 1996; Webb et al. 1996).

I have identified four general initiation scenarios for slope failures that triggered

historic Grand Canyon debris flows: 1) bedrock failures in weathered outcrops; 2) the 127

"fire-hose effect" (Johnson and Rodine 1984); and 3) failures in colluvial wedges during rainfall only; and 4) complex mechanisms in which combinations of the scenarios described above occur together. Bedrock failures typically occur in Hermit Shale, the Supai Group, and (or) the Muav Limestone. Failures may also occur in Redwall Limestone, but these are rare and form only small debris flows. Failures in bedrock occur along faces of weathered cliffs during intense precipitation, providing an excellent initiation mechanism for debris flows. Failures in colluvium caused by the fire-hose effect occur when floods pour over waterfalls and rapidly erode unconsolidated colluvial wedges mantling outcrops of Hermit

Shale or Muav Limestone and Bright Angel Shale. This last mechanism is the most common type that initiates Grand Canyon debris flows (fig. 3.12). Talus-slip failures of saturated hillslope sediments in colluvium or talus-covered slopes, the third type of initiation mechanism, normally occurs during very intense and (or) prolonged rainfall.

Colluvial-wedge failures generally initiate small debris flows. The high frequency of fire- hose triggered debris flows is not surprising in Grand Canyon. The stratigraphy of rocks exposed in Grand Canyon (fig. 3.2) consists mainly of alternating layers of shales, sandstones and limestones. This stratigraphy creates the classic stair-stepped canyon wall profiles that promote the fire-hose effect in Grand Canyon, with colluvial slopes set at the base of high vertical cliffs at many elevations in tributaries.

The role of clay minerals in debris-flow initiation has recently been examined throughout the Colorado Plateau by Griffiths et al. (1996). Historic debris-flow deposits in Grand Canyon were found to contain a variety of clays within the debris-flow matrix

(Table 3.6). Of the clay-producing minerals found in Grand Canyon debris-flow deposits, kaolinite and undifferentiated micas were most abundant. Quartz, feldspar and calcite minerals were found, but not in large quantities, and montmorillonite was rarely found. 128

.c E Q).=

o L.L.0

FIGURE 3.12. The percentage of each type of failure that has initiated debris flows in

Grand Canyon from 1939 to 1997 (modified after Melis et al. 1994). 129

TABLE 3.6. Summary of mineralogies of clay-sized particles less-than 2 ttm collected as subsamples from debris flow matrix sediments identified at selected sites where recent debris flows have occurred in Grand Canyon (from Melis et al. 1994).

INTERPRETED CLAY MINERAL ABUNDANCE

Year of Debris Montmor- Vermic- Chlo- Kao- Feld- Cal- Site Flow illonite ulite rite Micas linite Quartz spar cite

"18-Mile 1987 1 0 0 4 3 1 2 2 Wash" Mile 1987 0 0 0 4 2 0 2 0 19.9-L Lava 1966 0 0 0 4 3 1 2 2 Canyon 75-Mile 1987 0 0 0 3 3+ 1 2 2 Creek 75-Mile 1987 0 0 0 4 2 1 0 0 Creek 75-Mile 1987 0 0 0 4+ 2 0 0 0 Creek Monument 1984 0 0 0 4 3 1 2 0 Creek Crystal 1966 0 0 0 4 3 1 2 1 Creek

(0), indicates that the mineral was not detected; (1), indicates that trace quantities of the mineral were detected; (2), indicates that a small amount of the mineral was detected; (3), indicates that a moderate and (5), amount of the mineral was detected; (4), indicates that a large amount of the mineral was detected; indicates that the mineral was the dominant mineral detected. Muscovite and biotite were not differentiated. 130 chlorite and vermiculite were absent from all deposits sampled (Table 3.6).

Other than availability of source sediments, precipitation is likely the single most important factor in initiation of debris flows in Grand Canyon and elsewhere. Storms that cause debris flows can be either local or regional in scale, and rainfall on an existing snowpack situated along the high-elevation canyon rims can greatly increase runoff over cliffs. The intensity of storms over Grand Canyon tributaries is also greatly influenced by orographic effects, travel paths inland from the eastern Pacific Ocean, and seasonal to decade-scale variability in atmospheric circulation (MeRs et al. 1994).

The moisture in storms that cause debris flows in Grand Canyon comes mostly from the eastern Pacific Ocean and Gulf of California, but Mexico is an important source during summer (Carlton et al. 1990). Moist, northerly airflow from Mexico and the eastern

North Pacific Ocean creates Arizona's summer "monsoon," which often lasts from July through September (Andrade and Sellers 1988). In most years, spring climate is characterized by prolonged periods of drought preceding the summer monsoon; as a result, recent Grand Canyon debris flows have not occurred during the months of April through

June.

Floods in central Arizona are related to seasonally-varied stol ni types (Hirschboeck

1985; 1991), and floods in other tributaries of the Colorado River in Grand Canyon have been related to summer monsoon thunderstorms, dissipating tropical cyclones, and winter frontal storms (Webb et al. 1991b; Melis et al. 1996), including debris flows (Melis et al.

1994). Thomas et al. (1994; p. 8) report that flood regions in the southwestern U.S. have mixed populations of floods, creating "... an aggregation of floods that are caused by two or more distinct and generally independent hydrometeorologic conditions such as snow melt and rainfall." Most recent debris flows in Grand Canyon tributaries from 1939 131 through 1996 were triggered by summer thunderstorms; however, the largest recent debris flow during that period were caused by winter frontal storms and dissipating tropical cyclones (Cooley et al. 1977; Webb et al. 1989; Melis et al. 1994; Griffiths et al. 1996; Webb, 1996).

My examination of relations between the annual occurrence of daily precipitation >

25 mm and selected station characteristics for nine sites near Grand Canyon revealed that the annual frequency of daily intense precipitation (>25 mm) increases with elevation and record length, but is weakly inversely related to station latitude; no relation between intense daily precipitation and station longitude was seen (fig. 3.13). A positive relationship between station elevation and the annual frequency of intense precipitation was expected.

However, the positive relation between annual occurrence of intense daily precipitation and record length is was not anticipated. Because all of the station records examined are limited to the twentieth century, and all but one of the records continued through at least 1995 (the

Supai record ended in 1984), this trend suggests that occurrences of intense daily precipitation began to increase sometime during the last half of the century. It is assumed that if the frequency of intense precipitation increased through time that the likelihood of debris flows would also increase in the Grand Canyon region.

To test this hypothesis, trends in daily precipitation >25 min were evaluated for seven of the nine climate stations in northwestern Arizona near Grand Canyon -- Ashfork,

Grand Canyon, Lees Ferry, Peach Springs, Seligman, Supai and Williams — during the periods of 1900-1929, 1930-1959, and 1960-1993, following the general approach of

Webb and Betancourt (1992). Regression trends are shown for four of these records plotted within these time periods in figure 3.14. The annual frequencies of daily precipitation > 25 mrn were tabulated from these plots for each of the periods, and the ▪

1.2 132

1.0 O F-

E 2 0.8 0 2

>- — 0.6 o w u_ w O° 0.4 "11 Lu (.3 cc LU 0.2 CL

0.0

800 1,200 1,600 2,000 2,400 STATION ELEVATION, IN METERS

1.2

1.0 O

Q.. 0.8 0

cLei a. (..D z 0.6 Z--"t w LL w °x 0.4 L1 o cc ci 0.2

0.0

10,000 15,000 20,000 25,000 30,000 35,000 RECORD LENGTH, IN DAYS

FIGURE 3.13. Relations between daily precipitation >25 mm and selected station characteristics for 9 records in the vicinity of Grand Canyon. A. Elevation. B. Record length. (from Melis et al. 1996). 133 2.0

R2 = 0.40

•

1.0

• • • •

0.0

37.00 34.00 34.50 35.00 35.50 36.00 36.50

LATITUDE

2.0

R2 = 0.02

•

1.0

•

• • • 0.0 113.00 113.50 114.00 111.00 111.50 112,00 112.50

LONGITUDE > 25 mm and selected FIGURE 3.13. Continued. Relations between daily precipitation C. Latitude. D. station characteristics for 9 records in the vicinity of Grand Canyon.

Longitude (from Melis et al. 1996). 134

A R2=0.16 R2=0.12 R2=0.19

7

WILLIAMS, ARIZONA

1920 1940 1960 1980 2000 YEAR

R2 =0.16 R2=0.21 R2=0.48

ASHFORK, ARIZONA

0

1900 1920 1940 1960 1980 2000

YEAR

FIGURE 3.14. The number of days per year with precipitation > 25 mm from selected long-term records near Grand Canyon. Trend lines for the periods of early 1900s to 1930, 1930 to 1960, and post-1960 are determined from regression analysis. For the significance of trends, see Table 3.7. A. Williams. B. Ashfork (from Melis et al. 1996). 135

Cf) 1 2 CC Fe=0.26 R2.=0.01 R2=0 .02 10

O CV ›- CO SELIGMAN, ARIZONA Z Z WO D O I-la W IX LL Z < 0 Z Z < < EsLu J a_

0 A /

1900 1920 1940 1960 1980 2000 YEAR

12 CO cc R2=0.14 R2=0.13 R2=0.35

10 — >- 2 - —1 0 u_ O>- coz E GRAND CANYON, ARIZONA D 0-1 in 0 CC X in inz < 0 z if

0_ 0

1900 1920 1940 1960 1980 2000

YEAR

FIGURE 3.14. Continued. The number of days per year with precipitation > 25 mm from selected long-term records near Grand Canyon. Trend lines for the periods of early

1900s to 1930, 1930 to 1960, and post-1960 are determined from regression analysis. For the significance of trends, see Table 3.7. C. Seligman. D. Grand Canyon (from Melis et al.

1996). 136 apparent trends were tested for statistical significance (Table 3.7).

Most of the trends suggested by the regression lines from the first, middle, and latter thirds of the 20th century were not found to be statistically significant (fig. 3.14;

Table 3.7). However, the annual frequency of intense daily precipitation was found to increase significantly during the most recent period, 1960-1993, in two of the seven records analyzed (Grand Canyon and Ashfork, fig. 3.14, Table 3.7). Increasing trends of intense precipitation at these stations may partially explain the apparently-increasing debris- flow frequency documented in Grand Canyon since about 1983 (fig. 3.15); an increase that

also apparently corresponded with a overall reduction in debris-flow magnitude. It must be noted that the perceived change in frequency and magnitude could also reflect the fact that censusing of debris flows of all sizes has been much more intense during the latter half of

the study period than between 1872 and 1966.

The possibility of an increasing twentieth-century trend in intense daily precipitation

also suggests climatic nonstationarity; a condition that may be an important factor in understanding future patterns of debris flows in this region. Similar precipitation and

flood-frequency trends were recently reported by Webb and Betancourt (1992) in southern

Arizona, and were shown to significantly influence flood frequency of the Santa Cruz

River. More recently, Karl et al. (1995) found an increasing trend in daily precipitation >

50 mm throughout many areas of the U.S. and other areas throughout the world. In addition to the increasing trend, the data also suggest that precipitation between 1929 and

1960 was less intense (fig. 3.14). Although not a significant trend when tested using nonparametric statistics, that period of reduced, intense precipitation corresponded well with flood accounts and photographic evidence reported for large Grand Canyon tributaries between about 1940 and 1970 (Webb et al. 1992; Melis et al. 1996). Recent interviews 137

TABLE 3.7. Results of Kendall Tau-b (nonparametric) statistics on trends in the annual frequency of daily precipitation >25 ram in records near Grand Canyon (modified after

Melis et al. 1996).

Precipitation Period Years in Kendall Tau-b Probability for Record Analyzed Record Trend (T) Two-Sided Test

Ashfork 1900-1929 17 Slight Decrease -0.151 0.368 Ashfork 1930-1959 30 None -0.091 0.255 Ashfork 1960-1993 33 Increasing* +0.369 0.003

Grand Canyon 1900-1929 26 None +0.034 0.280 Grand Canyon 1930-1959 30 None -0.008 0.255 Grand Canyon 1960-1993 33 Increasing* +0.272 0.026

Lees Ferry 1900-1929 13 Slight Decrease -0.193 0.359 Lees Ferry 1930-1959 30 None 0.031 0.255 Lees Ferry 1960-1993 33 Decreasing -0.145 0.194

Peach Springs 1900-1929 n.d. n.d. n.d. n.d. Peach Springs 1949-1959 11 Decreasing -0.414 0.491 Peach Springs 1960-1993 33 Decreasing -0.159 0.194

Seligman 1900-1929 25 Decreasing -0.216 0.287 Seligman 1930-1959 30 None -0.038 0.255 Seligman 1960-1993 33 None +0.058 0.638 Supai 1900-1929 n . d . n.d. n. d. n . d . Supai 1930-1959 n . d. n.d. n . d . n . d . Supai 1957-1987 21 Decreasing -0.249 0.314

Williams 1900-1929 25 None -0.084 0.287 Williams 1930-1959 30 Decreasing -0.151 0.255 Williams 1960-1993 33 Increasing +0.110 0.368

*, indicates a statistically significant increasing trend in annual frequency of daily precipitation >25 mm; n.d., no data; <0.05 probability was used for significance test. 138 o o 11 I 11111 o • • c\J • • • • o • o-) •• • (3)

:* o CO CT)

N- CT) • CDC LL1 ›-

C5")

Lt) O)

• t- • cr)

o

111111111111 r 111111111 0 0 0 0 0 0 0 0 o o o o o o o o o 000 oo o o ô c5 ô d co Lo Lr) 7t- N

çw) `11A1M0A NVJ - Sft193C1 C131.VVILLS3

FIGURE 3.15. Relationship between timing and magnitude (volume of alluvial fans deposited at the Colorado River) for selected debris flows between 1939 and 1997. 139 with river runners who ran the river in the 1940s, 50s and 60s revealed very few accounts of tributary debris flows during that era.

Long-term precipitation patterns in the Grand Canyon region were also examined relative to known Grand Canyon debris flows by constructing two time series that were generated from standardized seasonal precipitation from nine northern Arizona stations with various length records: Ashfork, Bright Angel, Grand Canyon, Lees Ferry, Peach Springs,

Phantom Ranch, Seligman, Supai and Williams. The summer and winter seasons were defined as July through September, and November through March, respectively in these analyses, since most documented debris flows occurred from July through September and were caused by summer-season thunderstorms. The months of April through June were excluded from the winter-season time series since no historic debris flows have been reported in those months.

The standardized seasonal time series (fig. 3.16) provide a clearer perspective on twentieth-century relations between climate and debris-flow frequency in Grand Canyon.

The interannual variability of summer precipitation has generally decreased in northern

Arizona throughout the 20th century, and summer precipitation was below normal in most years between 1940 and about 1982; with the exception of 1968 and 1970 (fig. 3.16a).

Despite the fact that these two years were wetter than normal for that period, neither years

are associated with debris flows in Grand Canyon, although 1970 was characterized by widespread floods throughout the central highlands of Arizona. After 1982, summer indices remained highly variable, and three out of four wetter-than-normal summers produced debris flows in Grand Canyon (fig. 3.16a). However, several other debris flows also occurred in dryer-than-normal summers from 1982 through 1996. 140 2.0 o

-2.0 1111111111111111111111/1111111 11111111111111111,111111111111111111111111111111111111

n 7 (Definite Winter Events)

11111/1/1111111111111//111111111111111111111 1 1 1 1 111111,1111111111111111111111111111111111 o o CD CD r-- cs) C)

YEAR

FIGURE 3.16. Standardized time series for seasonal precipitation at nine northern Arizona stations from about 1900 through 1995. A. July through September rainfall. B. November through March precipitation. Note: numerals above bars indicate number of debris flows in that year. 141

Standardized winter precipitation between 1900 and 1996 clearly shows patterns of above-normal precipitation during the first and last thirds of the 20th century (fig. 3.16b).

In addition, below-normal winter rainfall occurred during the middle portion of the 20th century; a period that corresponds with very infrequent debris-flow occurrence in Grand

Canyon. The overall pattern of the winter standardized precipitation series generally agrees with the trends of annual frequency for daily precipitation > 25 mm. Unfortunately, there are too few data on winter debris flows in Grand Canyon to make any definite conclusions about the relation between annual precipitation and debris-flow initiation.

The trend of above normal winter precipitation after about 1978 corresponds well with an increased frequency of large floods throughout Arizona over the last two decades

(Webb and Betancourt 1992; Ely et al. 1994), as well as the return of large floods in

Havasu Creek in the 1990s (Melis et al. 1996); a major tributary of western Grand Canyon.

However, the relation between Grand Canyon debris flow occurrence and above-normal seasonal precipitation is poor for both summer and winter seasons within the time series

(fig. 3.16). This fact supports the premise that debris flows depend more on episodes of extremely intense precipitation rather than whether annual levels of precipitation are above or below normal. In their recent study, Karl et al. (1995) found increasing rates of intense precipitation globally, despite the fact that annual mean precipitation did not vary significantly in most records during the twentieth century.

As another way of looking at the relations between precipitation and debris flows in

Grand Canyon, I ranked the top twenty (by recurrence interval) single-day and multiple- day precipitation events by season for each of the nine stations used to construct the time series above (Appendix H). Comparisons between these precipitation events and known debris flows in Grand Canyon suggests that debris flows in the Canyon only occasionally 142 correspond with unusual single-day precipitation events. However, debris flows more commonly occurred in periods where unusual multiple-day precipitation occurred in conjunction with relatively unusual single-day precipitation events that likely acted to trigger debris-flow initiation mechanisms.

Overall, there is a poor relation between Grand Canyon debris flows for which the exact year and season are documented, and the annual precipitation trends shown in the seasonal time series generated here, but better relationships are found between debris-flow occurrence and combined multiple and single-day precipitation events that have relatively low recurrence intervals. Therefore, I conclude that annual precipitation alone is not a reliable predictor for Grand Canyon debris flows. However, future increasing trends in intense precipitation events, especially those that may occur as unusual "microbursts" within multi-day storms, may result in more frequent debris flows throughout Grand

Canyon and vicinity through time; only continued monitoring of debris flows will tell.

One Pacific Ocean temperature and atmospheric-pressure anomaly (referred to jointly as ENSO (Southern Oscillation); see Diaz and Markgraf 1992; Andrade and Sellers

1988) that is recognized as a reliable indicator of an increased probability of flooding in many regions of the western Americas is known as El Nirio (ENSO). Relations between

ENSO in the Pacific Ocean and streamflow in the southwestern U.S. suggest that streamflow is positively correlated with changing conditions of temperature and atmospheric pressure in the Pacific Ocean (Kahya and Dracup 1993; Cayan and Webb

1992). This correlation reflects the increased precipitation and delayed timing of winter storms from late fall to spring that is characteristic of ENSO years. Karl et al. (1995) recently suggested that changes in the delivery patterns of regional and global precipitation patterns are likely related to changes in atmospheric temperature extremes and may be 143 related to more frequent ENSO conditions recently observed between 1989 and 1995.

Changes in atmospheric circulation patterns from dominantly zonal to meridional

flow over the western U.S. are commonly observed during ENSO conditions (Webb and

Betancourt 1992). However, ENSO conditions have recently occurred at a more frequent

rate than during other periods of the twentieth century and correlates with increased

flooding in regions such as Arizona and California (Trenberth 1996; Webb and Betancourt

1992). Meridional circulation can increase the transport of warm, moist air from the

eastern Pacific Ocean into areas such as Arizona and result in increased frequency and

magnitude of floods. The recent persistence of ENSO and its identification as a flood-

forcing mechanism, along with the perceived increase in Grand Canyon debris-flow

frequency, warranted an examination of long-term ENSO records relative to historic Grand

Canyon debris flows as part of this study on initiation mechanisms.

In this study, ENSO events were designated as individual years or consecutive

years as described by Trenberth (1996), based on monthly sea-surface temperature

anomalies within the geographical area termed Nirio-3 for ±0.5 threshold using a 5-month

running mean. The anomalies were determined relative to a base period climatology of 1950-1979 (Trenberth 1996). From 1939 through 1996, debris flows occurred in at least

sixteen years (27 percent of the time); during ten of those sixteen years ENSO events

occurred in the eastern Pacific Ocean (62 percent of the time). Of the 59 debris flows

documented in Grand Canyon since 1939 (fig. 3.16) for which the year of the event was

documented, 27 occurred during an ENSO event; a 46 percent correspondence (Table 3.8).

Although the general occurrence of debris flows in Grand Canyon is not highly correlated

with ENSO, the largest recent debris flows have all occurred during El Nirio years (Table

3.8). This fact supports the idea that climate phenomena originating in the eastern Pacific 144

TABLE 3.8. Timing and debris-fan volumes of recent Grand Canyon debris flows (1939 through 1996) relative to El Nifio and La Nina events.

Year of Summer (S), El Nifio (E), Debris-Fan(s) Debris or or Volume(s), Flow(s) Site(s) Winter (W) La Nina (L) in (103 x m3)

1939 179.4-L S E 44 - 63 1952 96.7-L S E rid. 1954 179.4-L S L 3.2 - 8.4 1955 179.4-L S L 15 - 21 1963 179.4-L S E 12 - 14

1966 65.5-R, 87.8-R, 98.2-L W E n.d., n.d., 75, 179.4-L 3.8 - 6.1

1983 41.0-R, 41.3-R, 43.2-L, 43.7-L, S E n.d., n.d., n.d., n.d., 70.9-L, 71.2-R, 72.1-R, 78.7-L n.d., 7.2, 4.5, n.d.

1984 93.5-L, 225.8-L S L 6.8, n.d.

1987 5.7-R, 18.0-L, 19.9-L, S E n.d. 65.5-L, 75.5-L, 116.5-L

1989 24.0-L, 24.2-L, 24.4-L S L n.d., n.d., n.d., 30.2-R, 30.5-R, 125.5-L n.d., n.d., n.d., 126.9-L, 127.3-L, 127.6-L 8.5, n.d., 5.0, 129.0-L, 130.5-R n.d., n.d.

1990 62.2-R, 62.5-R, 62.6-R n.d., n.d., 1.2, 63.0-L, 63.3-R, 64.5-L n.d., n.d., n.d., 65.5-L, 75.5-L n.d., 12

1991 122.7-L, 207.8-L, 208.6-L S E n.d., 0.4, n.d. 1993 68.5-L, 70.9-L, 157.6-R, 160.8-R S E 5.9, n.d., 11.5, 10 1994 7.9-L S E 0.7 - 1.3 1995 87.8-R, 98.2-R, 179.4-L W E n.d., n.d., 9.5 1996 93.5-L, 95.0-L S L 4 - 8, 1.2 - 2.5

n.d., no data, or very small fan volume; volumes from Melis et al. (1994); ENSO events, after Trenberth 1996. 145

Ocean may exert great influence on the geomorphology of the Lower Colorado River over short time periods; especially under regulated river conditions. A better understanding of the relationships between Grand Canyon debris flows and atmospheric circulation anomalies, such as ENSO and others, will require additional collection of debris-flow data in the future, since the present dataset used to make the correlation is sparse.

DISCUSSION AND CONCLUSIONS

Debris flow is an important sediment-transport process in 529 tributaries of the

Colorado River between Lees Ferry and Diamond Creek (fig. 1.1; river miles 0 to 225) in

Grand Canyon, Arizona. An episodic type of flash flood, debris flow transports and deposits sediment ranging in size from clay to boulders into the Colorado River and are one source of sand downstream from Glen Canyon Dam. The frequency, magnitude, and particle-size distributions of historic debris flows are important data for developing a

sediment budget in Grand Canyon. Debris flows also create and maintain the debris fans

and rapids that control the hydraulics of the Colorado River; most of the contributing

drainage areas are small tributaries of low stream order. In addition, debris fans constrict

the river's channel, creating areas of flow expansion downstream from stable debris fans.

These expansions form recirculating eddies, or flow separation zones, that facilitate the

deposition and storage of fine sediment along the river in sand bars, also called beaches.

Most of the debris flows that reached the Colorado River in the last century

occurred in Marble Canyon and eastern Grand Canyon (fig. 3.11; Plate 13). Western

Grand Canyon had comparatively fewer debris flows during the last century; the notable

exception is Prospect Canyon, where 4 debris flows occurred between 1939 and 1963. 146

Numerous factors influence the probability that tributaries will produce debris flows.

Variability of regional and local climate, particularly in the warm season (between June and

October), affects the probability of high-intensity precipitation, one of two causes of debris-flow initiation. Furthermore, drainage areas of western Grand Canyon are generally drier than those in the vicinity of the Kaibab Plateau; therefore, the frequency of debris flows should be lower in western Grand Canyon. However, the apparent clustering of debris flows (fig. 3.11) could also be a consequence of a uniform and random process being censused over a relatively short period. Other factors that may influence the magnitude and frequency of debris flows are: lithology and geologic structures, the shape and area of drainage basins, total relief, the dominant aspect of the drainage, or local orographie effects that increase the intensity and duration of precipitation. Local and regional geologic structure, such as large faults and folds, increase the abundance and distribution of source sediment, such as colluvium; this factor increases westward.

Debris flows are initiated by bedrock and colluvial slope failures that occur during

intense rainfall. The intensity of rainfall required to initiate debris flows in Grand Canyon

is unknown, but sparse data suggest that a sustained intensity of over 25 mm/hr and a total

rainfall of 25 to 50 mm may be required. Three different types of storms have initiated

recent debris flows in Grand Canyon. Convective thunderstorms occur mainly in summer,

frontal storms occur mainly in winter, and dissipating tropical cyclones occur infrequently

during the summer or fall months. The peak discharges of debris flows are poorly related

to the triggering storm type (fig. 3.7). In general, convective summer thunderstorms

caused most recent, small-volume debris flows in Grand Canyon. In contrast, dissipating

tropical cyclones and frontal storms generated moderate to large volume debris flows. Few

data exist for debris flows triggered by tropical cyclones (two) and frontal storms (three). 147

Four scenarios for initiation of historic debris flows in Grand Canyon (fig. 3.12) exist. These scenarios include: 1) slope failures that occur in exposures of weathered bedrock, typically Hermit Shale and Supai Group Formations; 2) failures in colluvial wedges caused by water cascading over cliffs (the "fire-hose effect"); and 3) classic hillslope failures in colluvium caused by intense precipitation. In addition, combinations of the main three types of failures may also occur, and often complicate interpretations of debris-flow initiation scenarios. Multiple source areas combined with extreme topographic relief often result in complex scenarios of debris-flow initiation in Grand Canyon. The most common initiation mechanism in Grand Canyon is the fire-hose effect, occurring mainly on slopes termed colluvial wedges below cliffs of Redwall Limestone (fig. 3.2).

With the exception of debris flows in Prospect Canyon, where fire-hose effects are uniquely related to the basin's geology, the largest debris flows generally have been caused by combinations of failures in colluvium and weathered bedrock.

Typically, as is the case with most highly-sediment laden floods, debris flows in

Grand Canyon are very-poorly sorted with particles ranging in size from clay to boulders.

The most common clay minerals in debris flows are micas and kaolinite (Table 3.6);

kaolinite is a single-layer, non-swelling clay. Smectites, a group of multi-layer clays that

includes montmorillonite, have been detected in only one Grand Canyon debris flow (18-

Mile Wash). Kaolinite has properties that may facilitate the initiation of debris flows

(Pierson and Costa, 1987). The presence of kaolinite in Grand Canyon strata, as opposed

to smectites, may be a major reason why debris flows occur in Grand Canyon.

Previous research yielded a general estimate of 20 to 30 years for the average

recurrence interval of debris flows in Grand Canyon (Webb et al. 1989). Estimation of the

frequency of debris flows is difficult because of the physical nature of debris flow, the 148 isolated nature of the study area, and the lack of low-level aerial photography before 1935 and between 1935 and 1965. Interpretation of over 1,107 historical photographs spanning 125 years, supplemented with oblique and vertical aerial photography taken after 1935, indicates that the recurrence of debris flows ranges from one debris flow every 10 to 15 years in certain tributaries, to less than 1 debris flow per century in other drainage areas.

On average, debris flows may recur approximately every 30 to 50 years for any given tributary. Tributaries in eastern Grand Canyon appear to have a higher frequency of debris flows than those in western Grand Canyon, at least during the twentieth century.

Debris flows appear to cluster in time within tributaries having a high debris-flow activity, but this could be an artifact of the relatively short observation period (a century).

In Prospect Canyon, four debris flows and one hyperconcentrated flow occurred between

1939 and 1963; no debris flows occurred in 1872 to 1939 or 1963 to 1993. In 75-Mile

Canyon, no debris flows reached the Colorado River between 1935 and 1987, but two debris flows occurred between 1987 and 1990. If clustering is real, then it may suggest that the initial debris flow destabilizes the hillslopes and channels in the tributary, which then yields a higher frequency of debris flows until sediment sources are stabilized or exhausted.

Debris flows can directly and indirectly affect the stability of sand bars deposited in recirculation zones. Debris flows also directly affect sand bars (primarily separation bars) through burial and (or) erosion. Such effects have been observed in relation to several debris flows, including the debris flow of July 1987 in "18-Mile Wash" (river mile 18.0-

L), the debris flow of September 1989 in "127.6-Mile Canyon;" the debris flow of

September 1990 at "Crash Canyon" (river mile 62.6-R); and the debris flows of August

1993 at river miles 157.6-R and 160.8-R. Flow patterns in recirculation zones can change 149 as a result of increased constriction, as demonstrated by the effects of the debris flow of

September 1989 at "127.6-Mile Canyon." These changes may change the form of (or eliminate) reattachment bars. In addition, debris flows control geomorphic characteristics of the river corridor, influencing the preferred habitat of certain endangered fish species, such as the humpback chub (Gila cyphus) (Dr. Richard Valdez, oral commun., Bio/West

Inc., 1994).

Large debris flows that occurred before closure of Glen Canyon Dam in 1963 deposited large volumes of debris in the river, but most of that sediment finer than boulders was removed from debris fans by river floods. This ongoing interaction between tributaries and the river resulted in accumulations of boulders that form rapids. Since

1963, reworking by annual peak flows has been greatly limited by dam releases in most years. Hence, comparatively small debris flows now have a greater impact on the channel of the Colorado River, and as a result, post-dam debris flows are progressively constricting the river's channel at many locations by aggrading rapids with sediment finer than boulders. As a result, debris-flow deposits that have buried or eroded pre-existing sand bars cannot be eroded from debris fans. In order to redeposit sand bars in their former locations, higher dam releases are required to inundate and (or) rework aggraded debris fans.

Saturation of failure-prone talus may be intensified by concentrated sheetflow runoff from cliff faces. This situation is exemplified by talus at the base of prominent, cliff-forming strata, such as the Redwall Limestone. Precipitation intensities on a slope during a storm may be augmented with sheetflow running off the face of exposed bedrock, and this may greatly reduce resisting capillary forces in hillslopes. Accumulated sheetflow runoff is then concentrated at the intersecting line between the talus and bedrock. Under 150 these conditions, saturation of materials may occur more quickly; the importance of this phenomenon has been discussed in relation to extreme floods in other arid environments

(Schick 1988). Because of these conditions, failures on talus slopes may occur during relatively moderate precipitation conditions, compared to those commonly reported at precipitation stations, when storm cells encounter high-relief, vertical walls commonly found in Grand Canyon. Talus-covered slopes in Grand Canyon that have limited vegetal cover, and are commonly adjacent to extensive bedrock exposures, further increase the

likelihood of slope failures. Multiple source areas combined with extreme topographic

relief can result in combinations of the three basic initiation mechanisms.

Ultimately, it is both the supply of source sediment, and occurrence of intense

precipitation that controls the frequency and magnitude of debris flows (Bull 1991). If

production of colluvium is reduced, and the frequency and intensity of storms increases,

then debris flows will occur increasingly until source areas are scoured clean. Once this

occurs, debris flow frequency will decline until sediment production re-supplies favored

source areas for debris flows. Geomorphic responses to a climate change over the last five

to fifteen thousand years (last glacial to Holocene) have likely occurred in Grand Canyon.

Although the probability of occurrence of certain initiation mechanisms (i.e., fire-

hose effects) decrease as source sediments are removed, large volumes of sediment may

remain on adjacent hillslopes. These source areas are still prone to classic soil-slip failures,

but may not be subjected to rainfall intensities necessary to cause debris flows under

prevailing climatic conditions. Hence, variability in the climate of northern Arizona, as

well as changes in the locations of source sediments strongly affects debris-flow frequency

and magnitude through time periods ranging from decades to millennia. Ultimately, both

past and present climatic conditions related to debris flow probability are reflected in the 151 geomorphic framework of the Colorado River.

Hereford and Webb (1992) and Graf et al. (1991) suggest that fluvial erosion on

the Colorado Plateau is driven mostly by warm-season precipitation occurring from June to

November. The tendency for frequent summer floods in Grand Canyon tributaries agrees

well with seasonal patterns of flooding recorded throughout the western United States. On

the basis of 1,300 stream gage records in 10 states, most of the annual peak discharges

between 29° and 37° N latitude have occurred from July through September (Thomas and

others 1994).

The documented twentieth-century debris flows in Grand Canyon through 1996

were related to general hydroclimatic conditions in the southwestern U.S. to determine if

climatic anomalies, such as ENS O, could explain the perceived pattern of events. In

addition, the annual frequency of daily precipitation > 25 mm precipitation was tabulated

for records around Grand Canyon and related to latitude, longitude, elevation, and record

length. Elevation, and to a lesser extent record length, correlate positively with the annual

frequency of daily precipitation > 25 mm, whereas longitude and latitude showed no

correlation. Debris flows in Grand Canyon were found to correlate poorly with seasonal

time series during the twentieth century. However, debris flows were found to correlate

more closely with relatively infrequent periods of combined multiple-day and high

recurrence interval single-day precipitation events. In general, the largest debris flows

occurred following winter storms that occurred in years characterized as ENSO events, but

debris flows of all magnitudes correlated poorly with twentieth-century ENSO timing,

overall.

The highest number of debris flows occurred in summer, but had relatively less

impact on the geomorphology of the Colorado River. The documented relations of 152 frequency and magnitude of Grand Canyon debris flows and their impacts on the Colorado

River during the middle to late twentieth century suggest that atmospheric circulation in the

eastern Pacific Ocean likely has a great influence on river geomorphology. However, there

are too few data on debris flows produced in winter by frontal storms and cyclones to make

tight correlations between these events and known atmospheric circulation anomalies

related to the Pacific Ocean. 153

CHAPTER 4

DEBRIS FLOWS IN GRAND CANYON NATIONAL PARK: PEAK DISCHARGES, FLOW TRANSFORMATIONS, AND HYDROGRAPHS

Abstract

Quantitative data on hydrographs and flow behavior of debris flows in remote drainage areas are difficult to obtain directly. Hydrographs and flow behaviors are inferred for recent debris flows along the Colorado River in Grand Canyon from physical characteristics (stratigraphic relations, sedimentology and surface morphology) of debris fans, and from evidence of flow-surface elevations preserved in bedrock channels. From analyses of newly deposited sediment on debris fans, three types of debris-flow hydrographs likely occur in Grand Canyon. Type I flows have a single debris-flow peak followed by recessional "hyperconcentrated flow" or streamflow. This style of flow forms massive, lobate deposits with small channels incised during recessional streamflow. Type

II flows have multiple debris-flow peaks with intervening "hyperconcentrated flow" and

(or) streamflow phases, and result in deposits with complex physical characteristics that result from interlayered debris-flow lobes, streamflow gravel, and "hyperconcentrated- flow" deposits. Type III flows begin as Type I or II flows, but the recessional streamflow stage is higher than the debris-flow phases. Recessional streamflow reworks debris deposits or buries them beneath streamflow sand and gravel. Protracted runoff causes reworking of fan surfaces, making it difficult to interpret flow behavior on the basis of that physical characteristic. Peak flows and fan characteristics indicate that most debris flows persist for seconds to minutes, while recessional flows last for several hours to a day. 154

INTRODUCTION

Debris flows in Grand Canyon National Park, Arizona (fig. 4.1), transport poorly sorted sediment 1 to 20 km from initiation points to the Colorado River, forming debris fans containing large boulders that commonly exceed the entrainment threshold forces available during regulated Colorado River flows. Before river regulation in 1963, most aggraded debris fans were reworked relatively soon after deposition by the Colorado River

(Kieffer 1985; Melis and Webb 1993). Once reworked, these debris fans, which lack fine- grained debris-flow matrix, contain little evidence about debris-flow behavior. Analyses of fresh debris-flow deposits, however, reveal stratigraphic relations, sedimentology, surface morphology (physical characteristics), and flow-duration evidence that are important clues needed to reconstruct some aspects of debris-flow behavior in these semiarid tributaries.

In this paper, I propose hypothetical hydrographs of Types I, II, and III debris flows to explain the physical characteristics, hydrology, and geomorphic origin of several recent Grand Canyon debris-flow deposits. In the absence of first hand observations on debris-flow behavior, I propose process-based interpretations that combine hydrologic and sedimentologic data from preserved deposits to provide insight into the debris-flow process in remote semiarid tributaries.

METHODS

I used previously published (Meus et al. 1994) and unpublished sedimentologic and hydrologic data on recent Grand Canyon debris flows to design the hypothetical 155

FIGURE 4.1. Location map of the Colorado River through Grand Canyon National Park and vicinity, Arizona (from Melis et al. 1994). 156 hydrographs presented in this paper. In addition, I witnessed the recessional phase of a debris flow at Prospect Canyon (fig. 4.1) in March, 1995, (Webb and Melis 1995; Webb et al. 1996). These flow and sediment data are supplemented by stratigraphic descriptions of recent debris-flow deposits made at many of the study sites (fig. 4.1). Because the presence of large boulders makes excavation of long trenches through debris deposits inherently difficult, our stratigraphic descriptions were drawn from stratigraphy in 1-m square by 2-m deep pits, and natural exposures within channels incised into the deposits.

Some of the best stratigraphic profiles were exposed along the distal margins of new debris fans that were partially eroded by the Colorado River. Plan-view maps of debris-fan characteristics were surveyed to produce the illustrations of fan types presented here.

Peak discharges for debris flows that occurred between 1939 and 1993 were mainly estimated on the basis of flow superelevation evidence in channel bends (Apmann 1973;

Costa 1984) at 18 sites in Grand Canyon (Webb et al. 1989; Melis et al. 1994) using

V s= ( g • R, • 41-Is / W) 0.5, (4.1)

where g = acceleration due to gravity (9.8 m/s2), Rc.= the radius of bend curvature along the channel's centerline, All, = the difference in cross-channel surface elevation at the point of maximum superelevation (m), and W = the channel's top width at the point of maximum

superelevation. Use of equation (4.1) requires the assumption that the flow is steady and uniform, and that all flow lines are parallel as flow proceeds through the bend (Apmann

1973; Costa 1984). Because debris flows are somewhat similar to highly viscous fluids, these assumptions may be justified. 157

Iverson et al. (1994) showed experimentally that peak-flow velocities for debris

slurries estimated from superelevated mudlines were within 30 percent of measured values.

Their experiments also showed that velocity estimates vary widely based on the value of

channel width used in superelevation equations. Webb et al. (1989) and Melis et al. (1994)

also showed that peak-discharge estimates calculated from superelevation evidence varied

significantly depending on where cross-sectional areas were measured relative to peak-flow

velocities, and that discharges are not clearly related to local channel slope in Grand

Canyon tributaries (Melis et al. 1994). With these inherent uncertainties, I consider these

peak discharges for debris flows to be conservative estimates because they were determined

by combining peak-flow velocities with cross-sectional areas measured immediately

upstream or downstream from the superelevation bend.

PHYSICAL CHARACTERISTICS OF DEBRIS FLOWS AND FANS

Peak discharge and volume

Webb et al. (1989) and Melis et al. (1994) report peak discharges of debris flows ranging from about 100 to 1,000 m3/s (fig. 4.2a). The largest debris-flow peak discharge in Grand Canyon since 1890 — about 1,000 m3/s — was initiated by a tropical cyclone in

Prospect Creek (fig. 4.1) in September, 1939 (Webb et al. 1996).

Grand Canyon debris flows deliver significant volumes of sediment to the Colorado

River. The total volume of sediment appears to be related to drainage-basin area and storm type (fig. 4.2b). Historic debris flows initiated during winter frontal storms and autumn dissipating tropical cyclones deposited larger volumes of sediment on debris fans than did •

158

1,200 n = 18 • a • CONVECTIVE THUNDERSTORMS (SUMMER) • DISSIPATING TROPICAL CYCLONES (Al: - • z 1,000

CC2 600 a. • O 2 400 u_- 0 • t:T3 • rr 0 200 z • • • o o 01 1 10 100 1,000

DRAINAGE AREA, IN SQUARE KILOMETERS

100,000 n = 18

• CONVECTIVE TFIUNDERSTORMS (SUMMER) A • DISSIPATING TROPICAL CYCLONES (AUTUMN) 80,000 A FRONTAL STORMS (WINTER) cn c wc w 60,000

LL O 0 2 40,000 • 0 –J O Z 20,000 • 0 1— • • • • • O I • • 01 1 10 100 1,000

DRAINAGE AREA, IN SQUARE KILOMETERS

FIGURE 4.2. For recent debris flows from 1939 through 1995. A. Peak discharge, and

B. Minimum volume of fan deposit, versus drainage area (after Melis et al. 1994.) 159 debris flows triggered by summer convective thunderstorms. However, the total volume of sediment transported to the river is only weakly related to the peak discharge of debris flows (Meus et al. 1994). On most recently aggraded debris fans, the volumes of sediment

deposited during debris flows likely depend on factors such as single versus multiple slope

failures, tributary size, and storm duration and intensity.

Particle-size distributions and water content

Particle-size distributions for Grand Canyon debris flows are related, in part, to

hillslope conditions at initiation source areas in tributaries as well as on the parent bedrock

lithology (Meus et al. 1994). Typically, debris-flow deposits contain 15 to 30 percent sand

and often more than 30 percent boulders (Melis et al. 1994). However, the variability of

particle-size distributions among Grand Canyon debris flows is large, and recently

aggraded debris fans commonly have complex physical characteristics that reflect multiple

flow phases associated with both fluvial and hillslope processes. Many debris-fan deposits

have simple physical characteristics, and consist of single, lobate snouts with poorly sorted

particle-size distributions that are typical of debris flows (fig. 4.3a). Relatively greater

variability in particle-size distributions is found within debris fans that display complex

physical characteristics that result from multiple-phase sediment-transport processes (fig.

4.3b). Multi-phase flows are also characterized by general ranges of water contents

(Pierson and Costa 1987). Based on reconstitution of dried samples with water, Melis et

al. (1994) report that most Grand Canyon debris flows are 10 to 25 percent water by

weight. Grand Canyon fan deposits associated with hyperconcentrated flow and

streamflow are assumed to have had water contents greater than that,which influenced their

flow rheologies (Pierson and Costa 1987). Water content and percentages 160

CO 0 Co a) LC) CO C\J 0

80 SNOUT OF SEPTEMBER,1990 DEBRIS FLOW "CRASH CANYON" 60

I LiJ 40

20

o- w 0

< 100 D FIRST DEBRIS FLOW D SECOND DEBRIS FLOW 0 80 - - - HYPERCONCENTRATED FLOW - STREAMFLOW

60 1987 DEBRIS FLOW AT 18-MILE WASH

40

20

-6 -4 -2 0 2 4 PARTICLE SIZE (I) UNITS)

FIGURE 4.3. Particle-size data. A. Simple type, single-lobed debris fan. B. Complex fan

A es el 1-1 -‘ 1,`I TY% 1 t;T\11. f1 ru, rPC1r1ITIPC (frnrin Melis et al. 1994). 161 of fine sediments, such as clay, are important factors that enable sediment slurries to transport large boulders (Costa 1984) to the Colorado River (Melis et al. 1994).

Areas of similar surface particle-size distributions representing different flow processes can be mapped across the surfaces of complex fans if extensive reworking by recessional streamflow has not occurred. The different depositional surfaces of unreworked fans may be represented in vertically arranged stratigraphie units deposited near the fan apex. When fan surfaces have been extensively reworked by late stage recessional streamflow, or buried under streamflow gravel deposits, excavation is usually required to reveal stratigraphy that is diagnostic of debris flow, "hyperconcentrated-flow"

(Scott 1988) and streamflow transport and deposition.

Fan morphology and stratigraphy

In the most common type of debris flow, debris fans are aggraded by single-pulse debris flows followed by brief recessional streamflow. Between 1987 and 1997, this type of debris-flow sequence occurred almost annually in Grand Canyon and created simple debris fans characterized by massive debris-flow deposits (fig. 4.4a). Simple debris fans

(Type I; fig. 4.4a), have variable lobate shapes with low topographic relief, are truncated along their distal edges by the Colorado River, have poorly sorted surface particle-size distributions that include large boulders, and are often immediately incised by small tributary channels during post debris-flow recessional streamflow (fig. 4.4a).

The stratigraphy of many recently aggraded debris fans in Grand Canyon indicates a hydrograph of three sediment-transport processes: debris flow, "hyperconcentrated flow," and streamflow (fig. 4.4b). Multiple debris-flow pulses are separated by

162

DEBRIS FAN MORPHOLOGY STRATIGRAPHY SEDIMENTOLOGY o 50 FEET SIMPLE Tributary I---,--1 A DEBRIS FAN, 0 10 METERS • 00 m TYPE I Poorly sorted, "-'• Ob.: EVENT . • • 4 boulder-strewn :•:,. - - Poorly sorted, '- - - .- ' • - • debris fan .-0::.".:.-f• weakly imbricated • to non imbricat,ed, -. - - DF . iii ' —0.5m - no preferred .0F. ;...... • 40 - - li • '. : • • • • • .. . : • DF orientatiOn, cloy, • • . • '' • ' -411, —1.0m silt, sand, •• ." - _____••• "te• gravel, cobbles, • ve r ------...„,,_ RI ___r____------and boulders C o /0 ro do — I.5m COMPLEX Boulders 0 FEET Tributary I----T-0 ...... 0.0171 B . .• •-• 0 10 ME-TERS DEBRIS FAN, ,r• • ...::: ?..:: • :I:1C F : . : TYPE II :-..:-...-.10'. Repetitively bedded thin debris flow EVENT . .OPp . "... 0 .1::. h70, Ise deposits, open- • — 05m worked gravels, I-tC F :--;HFC:-::!'- -..-_-- 4 and relatively -,.. .i.-D.F..*. well-sorted,weakly ,IDAP Debris-Flow • " - • •• m imbricated fluvial Pulse ...... „,i4er Complex surface deposits . R 'o'er n morphology made of !:. -••1-: • Front R e word 13/, r od 0 1-3 DF pulses and HCF C°' °. _...------and stream gravels • —I.5m cii____,_sp FEET SIMPLE DEBRIS FAN, Tributary • . •. 00m C TYPE III EVENT 0 ND METERS ',;::• •::: Streamf low sand Smooth gravelly surface results Boulders are It.':'.. ,.::•.t and grovel, ".. • • . •.• . from recessional shallowly streamflow . massive ,imbricated, that covers initial DF covered . preferred deposit with sand ,,, -.- .. '• ' ' . . • and gravel orientation - Small ..- Incised '.:Channel — of particles •• * : • • 1.0m • • . ' Boulder; • -'! • - • • . . . . • . - -.Snout •,n11400- - ' • ' - p V i c"°Pcdo ------"—r---- - .-:-I.5m

FIGURE 4.4. Plan views showing three characteristic debris-fan types commonly

deposited by recent Grand Canyon debris flows. 163 flow transformations between debris flow and the other fluvial processes occurs over very short time intervals. Based on field observations of debris-flow deposits in central China and in a flume (J. Major, U.S. Geological Survey, pers. commun. 1995), such deposits can also result from multiple surges with no intervening flow transformations. However, the surge-behavior hypothesis does not explain the intervening "hyperconcentrated-flow" deposits.

Complex debris fans (Type II; fig. 4.4b) are similar in overall form to the simple fans but have unique physical characteristics, such as topographic features including: multiple debris-flow pushout lobes; streamflow runouts; incised channels and boulder levees or coarsening-upward stratigraphy that indicate transport and deposition by multiple flow regimes. Such physical characteristics are most commonly observed on recent Grand

Canyon debris fans that aggraded almost annually between 1986 and 1997. In some flows, debris-flow deposits are either buried by late-stage streamflow sand and gravel deposits, extensively reworked during recessional streamflow of a higher stage and (or) peak discharge than the initial debris flow, or are immediately incised by smaller-stage recessional streamflow that increase net topographic relief on fans. These sequences create complex reworked fans of Type III (fig. 4.4c); where reworking is associated with tributary rather than mainstem flows. Large tributary streamflows may also precede some debris flows (Cooley et al. 1977); evidence of these floods usually cannot be distinguished in stratigraphie profiles (Webb et al. 1989).

Occasionally, recessional streamflows with higher stages than the initial debris-flow pulse(s) may persist for several hours or up to a day. Such flows rework freshly aggraded simple or complex fans, or bury them under streamflow transported gravel deposits (fig. 164 by obscuring evidence of the debris-flow process.

The initial variability within and between these three types of debris fans results from the combined influences of: different types of hydrographs, local-scale mainstem and tributary channel geometries, bedrock lithologies in source areas, and variations in pulse magnitude during the flows related to storm types and hillslope failures. Debris flows produce fan deposits with widely varying particle-size distributions (fig. 4.5a). Regulated flows since 1963 have had only limited ability to rework recently aggraded fans (fig. 4.5 b and c). However, as a result of pre-dam reworking by the Colorado River floods, most debris fans throughout Grand Canyon developed remarkably similar particle-size distributions (fig. 4.5c; Webb et al. 1997) that mainly reflect debris-flow deposition of large boulders in the mainstem. However, the overall physical characteristics, including lithologie compositions, remain highly variable because of several factors such as those mentioned above, as well as diverse lithologies that occur in bedrock source areas.

HYPOTHETICAL HYDROGRAPHS OF DEBRIS FLOWS

I developed hypothetical hydrographs to explain the three types of debris-fan deposits observed in Grand Canyon (fig. 4.4). These hydrographs (fig. 4.6) derive from unique physical characteristics observed on fan surfaces, the duration of peak discharge, and the stratigraphic signature of multi-phase flow processes. Although the hydrograph axes are dimensionless, data reported by Melis et al. (1994) constrain the discharge from

100 to 1,000 m3/s, and the duration of peak discharge to a range of seconds to minutes

(fig. 4.6). 165 Particle Size, in mm

CV CO. CV CV Co CV 100

Initial- Debris Flow

80

60

Badger 40 18 Mile — — -Tanner 126.9 Mile - - + 127.3 Mile - 20 -127.6 Mile - -•- 160.8 Mile —0-• - Lava Falls

0 100

Fully-Reworked Fans Near Study Sites 80 (pre-dam reworking)

60

7.9-R 19.0-R 40 — — - 64.6-R - - - -. 126.9-R - - +- -166.4-L 20 - 176.4-R

o

-11 -9 -7 -5 -3 Particle Size, in 4) units

FIGURE 4.5. Particle-size distributions reflecting the end points of river reworking observed on recently-aggraded and pre-dam reworked debris fans in along the Colorado River in Grand Canyon. A. Initial textures along distal fan edges following several recent debris flows prior to river reworking. B. Textures of fully-reworked debris fans near the sites shown in part A reflecting the degree of river erosion that occurred during the pre- dam era (after Webb et al. 1997). 166

EVENT REPRESENTATIVE TYPE HYDROGRAPHY STRATIGRAPHY EXPLANATION TYPE I A Streomflow -- —"Hyperconcentroted flay» 7,n53-,c1e6,-7 t — Debris-flow - — • ' .• LA 0 (r < i 0 \ cn n DF E .

TIME—

TYPEZ B

, SF —HCF t Lu DF (..D cr < ....:. HCF = L.) \ \ . DF co \ \ . _ ..._ E .- :-... . HCF

TIME DF - TYPE Ut ...... 0 • . u_ . ., . Ec" SF Ô t • •- co U-I • 0 C' . . Cr • (.0 Q . LI- s H ., • EE T „ 2 i 1.,J ao (-) • . . • a u-/ —> Lu co . : LIJ (5 H • : . DF Ca . cc < __,a

• . CO U) CC • . TIME— •: •

FIGURE 4.6. Idealized Type I, II, and III hydrographs for debris flows in tributaries of the Colorado River in Grand Canyon. 167

Field evidence suggests that Type I tributary flows are single-peak debris flows with minimal recessional flow; an example of this type of debris flow is the 1990 flow at mile 62.6 ("Crash Canyon," Mefis et al. 1994). These debris flows typically occur in small drainage areas or follow brief but intense storms, and result from limitations in the supply of sediment and (or) water. Type I flows deposit simple fans composed of lobate, poorly sorted, structureless sediments (figs. 4.4a and 4.6a). Type I flows occur throughout

Grand Canyon, typically in summer; however, many Type I flows that occur in larger drainage areas may have insufficient magnitude to reach the Colorado River.

Type II flows consist of a peak debris-flow pulse followed by successively smaller debris-flow pulses. The pulses are separated by intervening "hyperconcentrated flow."

The last debris-flow pulse typically is followed by streamflow. These flows are most

con-m-ion in larger Grand Canyon tributaries and they deposit complex fans with well- defined areas of similar-size sediments that form distinct stratigraphie profiles (figs. 4.4b and 4.6b). Type II flows result from protracted summer thunderstorms or warm winter storms and may represent either non-synchronous timing of runoff in sub-basins or pulses that originate from flow transformations induced by formation and breaching of boulder and (or) debris obstructions in the upstream channel. Examples of Type II flows include the 1987 debris flow at 18-Mile Wash and the 1989 debris flow at mile 127.6 (Meus et al. 1994).

Type III flows usually result in complex , reworked deposits and begin with either a Type I or II flow. The debris-flow phase is followed by a streamflow flood of a stage sufficient to obliterate most of the evidence of the preceding flows (figs. 4.4c and 4.6c).

An example of this type of flow is the 1989 debris flow at Fossil Canyon (Meus et al.

1994). Surface particles on Type III debris fans are usually clast supported and composed 168 mainly sand and gravel. Type III flows deposit fans with surfaces that are interpreted as fluvial, but the surface features of the deposits represent only the recessional streamflow flood. Excavation of Type III fans usually reveals evidence of debris-flow deposition typical of Types I and II debris flows. On some debris fans, the only evidence of the early debris-flow phase may be deposition of large boulders in the Colorado River.

DISCUSSION AND CONCLUSIONS

In Grand Canyon National Park, debris flows initiated during intense summer thunderstorms or protracted winter storms travel between 1 and 20 km from source areas to the Colorado River. Large, well-preserved, mid to late-Holocene debris fans are found in the widest reaches of the Colorado River where debris-flow deposits accumulated above

the highest flood stage. Hereford et al. (1996) describe the debris-flow stratigraphies of

some of these fans in eastern Grand Canyon, but offer no interpretation of debris-flow behavior or hydrographs.

Analyses of mudlines and debris-flow deposits, preserved as levees in tributary canyons and debris fans along the Colorado River, provide evidence for the form of hydrographs of these flows. In some tributaries, variation in the stratigraphy of levees

suggests that debris flows start, stop, and transform into other fluvial processes with

distance from the initiation point. In some cases, evidence indicates that a debris flow

stopped and dewatered, and was subsequently remobilized when it gained sufficient water

and sediment from tributaries upstream (e.g., the 1987 debris flow in 75-Mile Canyon;

Melis et al. 1994). In other cases, the peak discharge of the debris flow increased with

distance from the source area owing to bulking up of a debris flows from bank failures and 169 entrainment of bed sediments (the 1990 debris flow at mile 62.6; Melis and Webb 1993).

Therefore, the sediment-transport processes, and thus the shapes of hydrographs, likely

vary with distance from source areas.

Peak discharge, reconstructed from superelevation evidence, suggests that debris

flows reaching the Colorado River have a minimum possible peak discharge and that the

maximum peak discharge may be a function of drainage area and (or) sediment supply in

source areas. Based on data from 18 debris flows that occurred from 1939 to 1993, the

range in peak discharge was 100 to 1,000 m 3/s. Debris flows smaller than 100 m3/s may

not have sufficient energy to continue flowing in the wide tributary channels; the upper

magnitude of debris flows is probably limited by the distribution of stored colluvium and

(or) regional flood frequency (Enzel et al. 1993). Comparison of the deposit volumes with

peak discharge suggests that the debris-flow phase of the runoff event is usually short,

lasting from about 30 seconds to several minutes.

The pattern of debris-fan deposition suggests a relatively consistent order of events

following the peak discharge of a debris flow. Secondary pulses of debris flow may

follow the initial pulse. Mudlines, levees, and the volume of sediment deposited on the

debris fan indicate that the peak discharge of secondary pulses is much lower than the initial

surge. Sediments identical to previously ascribed to "hyperconcentrated flow" are

deposited immediately after the peak discharge and between the secondary debris-flow

pulses. Rather than resulting from a distinct fluvial process, "hyperconcentrated-flow

deposits" may actually result from matrix dewatering, subaqueous debris flows under

streamflow, or reworking of runout from bank collapse of freshly incised debris-flow

deposits. Streamflow, typically the last phase of the runoff event, can cause extensive

reworking of channel deposits. Some storms of long duration may cause streamflow 170 discharges that exceed the maximum stage of the initial debris-flow pulse(s). In these situations the entire suite of physical characteristics associated with debris-flow deposition may be altered through reworking. 171

CHAPTER 5

DEBRIS FLOWS AND ALLUVIAL FANS: THEIR INFLUENCE ON THE REACH-

VARIED GEOMORPHOLOGY OF THE COLORADO RIVER IN

GRAND CANYON, ARIZONA

Abstract

The channel of the Colorado River through Grand Canyon is dominated by debris- flow processes over at least 225 miles. Local characteristics related to fan-eddy complexes vary by reach along the mainstem and control recirculating flow, sediment transport and sediment storage. Storage in the channel occurs as alluvial sand deposits in deep pools, as terrestrial sand bars along shorelines, and as coarse alluvial fans and channel bars that are relatively permanent channel features. Terrestrial sand deposits occur mostly within "fan- eddy complexes" on or near alluvial fans, and respond dynamically to sediment availability and flow magnitude over short time scales. In contrast, most reworked alluvial fans are stable boulder piles, while channel bars are comprised of cobble to gravel-sized particles.

These coarsened deposits are far less dynamic than sand deposits, clearly reflect local source lithologies, and are only modified by large mainstem floods like those that occurred before 1963, or by tributary debris flows that transport boulders from tributaries up to several times per century. Analyses of key attributes of all fan-eddy complexes along the river in Grand Canyon indicate that the eleven previously-designated geomorphic reaches of the channel, based on bedrock lithologies at river level, can be combined into six reaches that better-reflect local features of fan-eddy complexes known to control flow, sediment transport and sediment storage within this relatively unique debris-fan dominated river. 172

INTRODUCTION

Discerning geomorphically-distinct channel reaches from one another along the

Colorado River is vital to understanding process interactions between the mainstem and its bedrock tributaries in Grand Canyon. Recent restoration efforts aimed at preservation of river resources below Glen Canyon Dam (DOI 1995) have provided additional motivation for defining river reaches based on geomorphic features; especially those known to control flow and sediment transport. The need for clearly defined river reaches below Glen

Canyon Dam comes from the documented relations between river resources, such as sand bars, vegetation and fish habitats, and landforms and processes that influence those resources within dominant channel units termed "fan-eddy complexes" (Webb et al. 1989;

Schmidt and Graf 1990; Rubin et al. 1990; DOI 1995; Melis et al. 1994; Schmidt and

Rubin 1995). Similar needs for understanding the role of geomorphic processes in ecosystem dynamics have been identified for a variety of rivers that have suffered

"disturbances" from regulation (Stanford et al. 1996; Ward and Stanford 1996a; 1996b;

Gore and Shields 1995; Ligon et al. 1995; Junk et al. 1989). Morphological and hydrologic data are increasingly being used to develop physical-habitat models for rivers, especially where threatened and endangered fishes are a concern (Parasiewics 1996).

The Colorado River through Grand Canyon is affected by debris-flow processes that result in alluvial fans and rapids at 76 percent of tributary confluences over at least 225 miles (Melis et al. 1994; fig. 5.1; Plates 12 and 13). Local channel characteristics of fan- eddy complexes are controlled mostly by bedrock lithology and geologic structure (Howard and Dolan 1981), but are also greatly influenced by periodic tributary debris flows (Meus and Webb 1993; Melis et al. 1994; Webb and Melis 1995; Schmidt and Rubin 1995). Fan- 173

o o t- en ro

FIGURE 5.1. Location map of the Colorado River through Grand Canyon showing the locations of major debris flows during the last century (from Melis et al. 1994). 174

eddy complexes vary from reach to reach along the mainstem, and are the primary control

on eddy dynamics associated with recirculating flow, and sediment transport and storage

throughout the river system in Grand Canyon (Schmidt and Graf 1990; Schmidt and Rubin

1995). Throughout the system, sediment storage in the channel occurs as subaqueous

alluvial sand deposits in deep pools, as terrestrial sand bars along shorelines, and as

alluvial fans and channel bars. Sand deposits are dynamic, respond to sediment availability

and flow magnitude over short periods, and occur mainly within fan-eddy complexes

(Schmidt and Rubin 1995). In contrast, river-reworked alluvial fans are composed mostly

of boulders, while channel bars contain cobble to gravel-sized particles (Howard and Dolan

1981; Melis and Webb 1993; Melis et al. 1994; Webb 1996). The coarsest deposits are

much less dynamic than sand bars, and are only significantly modified by infrequent events

such as: large mainstem floods that occurred annually before 1963, occasional, high flows

in the regulated era, and deposition of boulders on alluvial fans during tributary debris

flows up to several times per century (Melis and Webb 1993; Melis et al. 1994).

Mainstem channel width through Grand Canyon is controlled mostly by bedrock

lithologies that occur at river level (Howard and Dolan 1981). The spacing of mainstem constrictions caused by alluvial fans and channel bars, and the severity of rapids is known to be closely related to regional geologic structures, such as faults (Dolan et al. 1978).

Other important controls on the spacing of fan-eddy complexes are: drainage-area characteristics, such as shales that promote hillslope failures and more resistant lithologies, such as limestones that produce large boulders (Melis et al. 1994; Griffiths et al. 1996;

Appendix G), and the combination of frequency and magnitude of tributary debris flows and mainstem floods (Graf 1980; Kieffer 1985; 1990; Melis and Webb 1993; Melis et al.

1994; Webb et al. 1996). 175

There is high variability among fan-eddy complexes throughout Grand Canyon

(Schmidt and Graf 1990; Schmidt and Rubin 1995). This is true because the mainstem and its tributaries are influenced downstream from Lees Ferry by varied bedrock types found in the Mesozoic, Paleozoic and Proterozoic sections of Marble and Grand Canyons, and because the potential for debris flows is spatially non-random across the 225-mile long study area (Griffiths et al. 1996). Owing to the diversity of lithologies that influence the mainstem and tributary debris flows, analyses of all fan-eddy complexes are needed to evaluate the eleven previously designated geomorphic reaches of Schmidt and Graf (1990).

Their description of the river's geomorphic framework was based mainly on where individual geologic formations occur at river level. However, it is unclear whether designations based on such criteria actually represent unique groupings of channel characteristics that control flow and sediment.

Sediment and hydrology are generally acknowledged to be the foundation of the river's ecosystem (DOT 1988; 1995). Because sediment inputs to the river occur mainly at river miles 1 and 61, clearly-defined geomorphic reaches are needed to assist management efforts aimed at sustaining sediment-related resources below Glen Canyon Dam. Such efforts include periodic controlled flooding (see chapter six of this volume; DOT 1995;

Webb et al. 1997). In order to be useful, geomorphic reaches must be clearly differentiated from one-another and be designated based on physical features of the channel that significantly control flow and sediment. In Grand Canyon, these features are mostly related to: channel geometry and width; the type of shoreline at river level (sloping-talus versus vertical bedrock; the size and shape of fans; and the spacing of fan-eddy complexes relative to each other and the related features mentioned above. Mainstem constrictions and downstream expansions associated with fan-eddy complexes create recirculating flow 176 within eddies, hydraulic features of the river that trap suspended load (Schmidt and Graf

(1990). Cobble bars are also important channel controls with respect to flow and sediment and are part of the fan-eddy complex. However, these features are limited throughout the system to wider channel reaches (Howard and Dolan 1981). Because channel width, depth

and shape are limiting factors in the stability and accumulation of alluvial fans and cobble bars in the mainstem, the size of these and other fan-eddy complex features correlates

positively with average channel width in Grand Canyon (Howard and Dolan 1981). It is

clear that the dominant features of the fan-eddy complex are the alluvial fan and

downstream controls, such as cobble bars or other fans (Schmidt and Rubin 1995).

However, I hypothesize that the fundamental controlling factors of bedrock lithology and

structure are the primary controls on fan size, shape and spacing, and therefore, ultimately

on flow and sediment transport/storage in Grand Canyon.

PREVIOUS WORK

Conceptual models of fan evolution in Grand Canyon

Kieffer (1985) provided the first conceptual model of debris-fan evolution along the

Colorado River in Grand Canyon. Her study relied heavily on hydraulic principles of flow

and sediment transport to estimate how high mainstem floods and tributary debris-flow

deposits occasionally interacted; and how river reworking results in stable, constricting

fans that influence river geomorphology. Kieffer's (1985; 1987; 1990) conceptual model

was the first to explain the general interactions between debris-flow deposits delivered to

the mainstem and fan-reworking processes of the Colorado River. Based on a sample of

confluences, Kieffer (1985; 1987; 1990) estimated that the average mainstem constriction 177 by reworked fans system-wide in Grand Canyon was about fifty percent. Kieffer studied particle-size and mainstem flow-velocity characteristics at a recently-aggraded rapid

(Crystal Creek; river mile 98.2; see Cooley et al. 1977) during a large release from Glen

Canyon Dam in 1983. Based on the effects at Crystal Creek and other fans, she concluded that main stem flood peaks as high as about 11,000 m3/s were required to completely rework alluvial fans to stable, coarsened conditions (Kieffer 1990).

Later, Melis and Webb (1993) described an alternative conceptual model of fan evolution in Grand Canyon by using a simplified sediment budget approach. Their approach was based on recent debris-flow frequency and magnitude data in combination with volume and particle-size data from several pre-dam, reworked fans. Their sediment- budget approach also relied on relations of particle-size distributions measured from: 1) samples retrieved from test pits excavated into uneroded debris-flow deposits; 2) nearby pre-dam, river-reworked fans; and 3) a stable channel bar located downstream from one of the reworked fans. Melis and Webb (1993) report that sediments transported during three debris flows near river miles 62-63 in 1990 were very poorly sorted with 90 percent of the sediment less than 0.5 m in diameter (b-axis). According to them, comparisons among the respective particle-size distributions above provide a reach-specific estimate for the competence of pre-dam mainstem floods necessary to modify and maintain fans and bars within the channel. Pre-dam channel maintenance by high-magnitude floods apparently occurred in their study reach concurrently with relatively-frequent debris-flow deposition

(Griffiths et al. 1996; figs. 5.2 and 5.3).

By comparing the reworked and unreworked particle-size distributions of fan deposits at miles 62.5-R, 62.6-R, and 63.3-R, they estimated that between 100 and 400 debris flows, ranging from 1,200 to 4,000 m 3 total volume per flow, would have been 178

; located - complex," FIGURE 5.2. Replicate aerial photographs showing a typical fan-eddy fan in 1987, before being at river mile 62.6-R. Left. Condition of the 62.6-R debris showing poorly-sorted impacted by a debris flow in 1990. Right. The 62.6-R fan in 1993,

debris deposited at the mouth of the tributary by the 1990 debris flow. 179 PARTICLE SIZE, IN MILLIMETERS co

o UI (ID 1.0 100

80

60

40

20

o -12 -10 -8 -6 -4 -2 0 2 4 PARTICLE SIZE, IN PHI UNITS 80

70

60

zw 50

cc 40

30

20

10 o

100 n - 400 RM 62.6-R, PRE-DAM RIVER REWORKED FAN • RM 62.8 DOWNSTREAM CHANNEL BAR 80

Ui Ui -J 60 - zI- a. o cc 0 cc U) -J ô_ UJ 0 CD -J Z

NEE

o Ell ..: WM ,, PARTICLE SOURCE

FIGURE 5.3. Lithology and particle-size characteristics of a fan-eddy complex at the

mouth of tributary 62.6-R. A. Particle-sizes. B. Lithologies. C. Comparison of pre-dam, river-reworked fan and the downstream channel bar (from Melis et al. 1994). 180 required to aggrade each of the three fans to their present sizes and surface textures (Meus and Webb 1993). Their approach assumed that Grand Canyon debris fans foi III gradually over long periods of time, in much the same way that most other alluvial fans accumulate in arid and semi-arid environments (Bull 1977; 1991). That is, they assumed that most reworked fans along the Colorado River result from the cumulative aggradation of poorly- sorted sediment delivered by many debris flows over time, each event contributing a small percentage of the overall fan volume (Melis and Webb 1993). However, alluvial-fan evolution occurs somewhat differently in Grand Canyon than it does in environments where reworking of fan deposits by a large river does not occur. Under natural conditions where river interactions do occur, each consecutive debris-flow deposit is likely to be mostly eroded by river floods on a periodic basis. The degree to which this reworking occurs depends on the flood frequency of the river system and the local characteristics of the channel, as well as the particle-size distribution of the initial fan deposits. On the basis of pre-dam streamflow and paleoflood data (O'Connor 1994), significant fan reworking apparently occurred on an almost-annual basis.

The conceptual model presented by Melis and Webb (1993), therefore, assumes that each successive tributary debris flow contributes to the maintenance and evolution of a given fan-eddy complex by depositing poorly-sorted debris at the confluence up to several times per century (Melis and Webb 1993; Melis et al. 1994). However, only a small percentage of each debris-flow deposit - the large boulders exceeding the pre-dam competence of the Colorado River - can resist river reworking and remain behind to form a debris fan. Such relatively stable fans are observed at 76 percent of all tributary confluences throughout Grand Canyon from Lees Ferry to Diamond Creek (fig. 5.1; Melis and Webb 1993; Melis et al. 1994). However, Melis and Webb (1993) also acknowledged 181 that debris-flow frequency and magnitude are unlikely to remain constant over extended periods of fan evolution, stating:

Larger magnitude debris flows, similar to the one at 5,400 years B.P. [in tributary 63.3-R], may have occurred more frequently during some earlier stage of the history of these debris fans [62.5-R, 62.6-R and 63.3-R], but the amount of colluvial source-sediment for such flow magnitudes may have become limiting as the debris fans formed. This would only be true, however, if production rates of source sediments fell behind rates of evacuation owing to debris flows (Meus and Webb 1993; p. 1274).

Since large debris flows could more-rapidly aggrade alluvial fans throughout Grand

Canyon, the model presented by Melis and Webb (1993) is considered to provide a conservative time/event estimate for fan evolution.

One recent study that supports a less gradualistie model for fan aggradation in

Grand Canyon was conducted by Hereford et al. (1996). They documented considerably fewer than 100-400 "fan-forming" debris flows that were responsible for deposition of a large, mostly uneroded fan located in a wide-valley setting at river mile 65.5-L. The geomorphic contrast between this large, uneroded fan and the reworked, bouldery fans observed in narrower reaches warrants consideration of a third model of fan evolution. In this model, Grand Canyon fans are formed cumulatively by larger debris flows that occur when hillslope sources are abundant. As primary source areas become depleted, progressively smaller debris flows occur. System-wide, the accumulation of closely- spaced, coarsened fans in medium-width reaches contribute most significantly to the current fan-eddy complexes (Howard and Dolan 1981).

Regardless of which model of Grand Canyon fan evolution is most accurate, it is clear that before regulation by Glen Canyon Dam, debris fans were significantly eroded by large floods of the Colorado River. Significant reworking of most alluvial fans resulted in 182 residual surfaces comprised mainly of boulders; while gravel and cobble eroded from reworked fans were deposited downstream as coarse bars in the wider channel below fan constrictions. These downstream channel bars are well-sorted and imbricated (Melis et al.

1994; fig. 5.3). Uneroded debris-flow deposits, reworked debris fans, and channel bars have distinctive particle-size, as well as lithologic, distributions (fig. 5.3) that are controlled by drainage-area and fan-eddy complex characteristics (Melis and Webb 1993; Melis et al. 1994; Webb et al. 1996).

Melis and Webb (1993) compared particle-size data from the reworked debris fan at tributary 62.6-R with the uneroded 1990 debris-flow deposit. They found that about 15 to

20 percent of the particles contained in the 1990 deposit would probably not have been transported by high-magnitude pre-dam floods, owing to their large size and the shielding pocket-effect of nearby clasts (Wiberg and Smith 1987). Particles in this size range are abundant on nearby reworked fans. Particle-size data for the downstream channel bar located at river mile 62.8 (figs. 5.2; 5.3) indicate that most particles reworked from the upstream fan were smaller than boulders, and that clasts 128 to 512 mm in diameter (b- axis) were likely transported only short distances, either into the pool below the fan or onto the channel bar. High-velocity flow through reaches constricted by debris fans quickly decelerates as flow enters the downstream expansion, or "flow-separation" (Schmidt and

Graf 1990), zone, but at extremely high discharge coarse particles may be scoured from pools to downstream channel bars by highly turbulent recirculating flow within the expanded eddy-pool complexes (Kieffer 1990). Because very large floods are thought to have occurred infrequently in the pre-dam era (O'Connor et al. 1994), the larger particles eroded away from debris fans often deposit less than 0.2 km downstream from constricting fans in wider reaches where channel bars can form and persist. 183

Depending on local river channel geometries, debris fans and channel bars probably attain quasi-equilibrium conditions over decades to centuries, while gravel and finer particles are transported downstream. Quasi-equilibrium conditions of debris fans and channel bars depend on flood frequency and local channel morphology (Leopold, 1969;

Howard and Dolan 1981; Schmidt and Graf 1990; Webb et al. 1991). However, in the highly debris-fan dominated river channel through Grand Canyon, fan-eddy complexes impose relatively-unique local controls on flow and sediment transport that influence quasi- equilibrium conditions over significant reaches. These local controls are not common to most river channels.

Re-examining previously-designated geomorphic reaches

During his 1869 exploration of the Colorado River through Grand Canyon, John

Wesley Powell noted that the river's channel and its flow characteristics, or its "mood," as he referred to it, varied dramatically as he and his crew moved downstream (Powell 1895).

Powell attributed the severity of the river's worst rapids and the "oppressive constriction" of its narrowest granite reaches to the relative hardness of the various bedrock lithologies that form the channel at river level (Powell 1895). Powell's explanation for what controlled the geomorphic framework of the river remained mostly unchallenged until

Leopold (1969) offered an alternative explanation for the spacing, occurrence and severity of rapids, referred to as "quasi-equilibrium theory." Leopold (1969) concluded that rapids form in bedrock-channeled rivers where packets of coarse debris transported by the river accumulate on the bed to form a pool-riffle

sequence (Langbein and Leopold 1964). These accumulations of coarse-sediment were periodically transported as distinct packets by occasional large floods. However, later 184 researchers reported that the majority of rapids in Grand Canyon resulted primarily from boulders being deposited in the mainstem by tributary debris flows (Webb et al. 1987;

1988; 1989; Melis and Webb 1993; Melis et al. 1994; Webb and Melis 1995; Webb et al.

1996). In fact, only about one percent of all rapids presently found in Grand Canyon were

directly created by rockfalls or other processes unrelated to debris flows (Melis et al.

1994). Secondary riffles and rapids form downstream from debris fans as a direct result of

debris-fan reworking within fan-eddy complexes. These features are usually caused by the

highly-graded outwash channel bars composed of gravel to cobble-sized particles eroded

from constricting fans immediately upstream (Melis et al. 1994). Leopold's (1969) quasi-

equilibrium theory was proven to be useful in explaining coarse-sediment transport in many

river systems. However, it failed to adequately explain sediment transport and storage

when applied to the Colorado River in Grand Canyon (Dolan et al. 1978; Howard and

Dolan 1981).

Howard and Dolan (1981) described reach-varied characteristics of the mainstem

through Grand Canyon in terms of flood-stage valley width and reach-averaged gradient,

as well as characteristics of cobble bars, deep pools, alluvial fans, tributary entrances, and

sand tenaces. They concluded that:

These measures indicate covariant response to downstream differences in lithology and structure. The effect of lithology is primarily reflected by the valley width; resistant rocks exposed near river level create a narrow valley. Valley width, in turn, is strongly related to the other variables..., and is the major cause of the variation pattern. Differences in bedrock resistance affect channel morphology not only through variations in channel width, but also through fracturing of the bedrock which affects the frequency of tributary fans and deep pools (Howard and Dolan 1981, p. 273).

Based on their work in the 1970s, Howard and Dolan (1981), were the first to designate 185 geomorphically-distinct reaches for the mainstem through Grand Canyon, referring to four types of reaches: 1) a wide valley with freely-meandering channel; 2) valleys of intermediate width; 3) narrow valleys in fractured igneous and metamorphic rock; and 4) narrow valleys in massive limestone. In describing these reaches, they noted that the infrequency of debris fans in narrow valley reaches was explained by the locally-increased competence (stream power) of floods to remove debris delivered by tributaries (Howard and Dolan 1981). They also refuted Leopold's (1969) claim that the spacing of rapids in

Grand Canyon was random and did not correlate closely with tributary confluences; noting,

"...the major rapids occur at tributary entrances, although in some places the narrow channel may limit the fan to a boulder accumulation on the channel bed," (Howard and

Dolan 1981, p. 274). Overall, Howard and Dolan (1981) provided many important insights on the genesis and influence of key morphologic features of the Colorado River channel, such as alluvial fans and cobble bars, and explained how they varied by reach mainly in response to channel width and geologic structure throughout the Canyon. Most of the river's fall through Grand Canyon occurs over short distances in the rapids and riffles that separate low-gradient pools (Leopold 1969). Therefore, reach-averaged gradients reported initially by Leopold (1969), Howard and Dolan (1981), and later, by

Schmidt and Graf (1990), more readily reflect the frequency of fan-eddy complexes among reaches that are more or less affected by debris flows and alluvial fans, rather than typical/modal channel gradients.

The general reach descriptions provided by Howard and Dolan (1981) are most important for the insight they provided to later workers, including this author. However, their reaches were broadly described, rather than designated in tenus of specific river miles. 186

During the 1980s, environmental concerns about downstream impacts to river resources from operations at Glen Canyon Dam (fig. 5.1; river mile -16, near Page, Arizona) resulted in additional geomorphic studies of the Colorado River in Grand Canyon. These studies were especially concerned with the changing conditions of sand bars along the river, the overall sediment balance of the system, and the relations between physical characteristics of the river and its biotic resources (DOI 1988; 1995).

As part of their study on the effects of dam releases on downstream sandbars,

Schmidt and Graf (1990) were the first to specifically designate geomorphic reaches by

river mile between Lees Ferry and Diamond Creek (river miles 0 through 225; Table 5.1).

Although they reported several average channel characteristics within their eleven

geomorphic reaches, the beginning and ending points of those reaches were determined

solely by the first appearance, or later reappearance(s), of individual geologic formations at

river level (Schmidt and Graf 1990; Table 5.1). While Howard and Dolan (1981)

identified the role of bedrock lithology at river level in controlling average channel width,

they did not emphasize that individual geologic foimations are important to channel

geometry, but referred generally to grouped lithologic characteristics that control valley

width. Howard and Dolan (1981) also mentioned several other channel features, such as

alluvial fans and cobble bars as important components in the geomorphic framework of the

river. Schmidt and Graf (1990) included many of these features in their analyses and

designation of the existing eleven reaches (Table 5.1). Although some averaged channel

characteristics besides geologic formations at river level were reported by Schmidt and Graf

(1990), those averages were based on limited subsets of data for fan-eddy complexes from

selected reaches of the river (Table 5.1).

• • g •

187

TABLE 5.1. Previously designated geomorphic reaches of the Colorado River through

Grand Canyon (river miles 0-225), based mainly on bedrock lithologies at river level (after

Schmidt and Graf 1990).

cc c,r3 8 8 8 cilç i V) V) V) Cl) U o

as 0 0 0 C 0 0 0 C C 0 ON 'I' v% 0 — tr, r- o r-.1r•-,. tO NNC — N o _co 0 = C o. o o o 0000 o o o. c? o d d d 6 6 6 6 ci 6 o 6

• 0, tr) •ar h C — 00 0 tr) •rer CO nCD siD th 0 r-

c-- r- — tf:) CrN Cfn r-: 06 sc; 06

3 3 E 32z3 3i iz z3z

v) (.5 > (• .5 c/i c4 c/5 v) cc • N >

cu O 0) .9 ô 2-4 4 O 1 ti) >, 'E 6. '-'. fi a E L') 1 8 CD i . Z E L, u° U 0 C., -,-, o.) 0 ,. ,,, .-, 0 y,c . ,-,02 a , t -a • 3 *3' El C. '' E k L''' Tnu o ° o 7) 0 co Oto. 0 0 0 0 CI) 04 ...) U 0 —.tE .:), .-J ..] 0

co 0 in ' Ii -0- r-: N v..; — r-: ,..0 r-- — — N. 0.• 'CI' '0,...; .30 4—0. ,-..1 , — N en ,C, r- - (-4 r,-. -0- .,-) .0 188

Later, researchers of alluvial sand deposits in Grand Canyon reported that:

The fundamental channel unit of rivers that flow through canyons that have abundant debris fans is a complex composed of (1) a backwater upstream from the debris fan, (2) a debris fan and channel constriction, (3) an eddy or eddies and associated bars in the expansion downstream from the fan, and (4) a downstream gravel bar. These fan-eddy complexes exist at the mouths of nearly all debris-flow generating tributaries... A large portion of fine-grained sediment in these canyons is deposited in eddies; this proportion is as large as 75 percent in Grand Canyon (Schmidt and Rubin 1995, p. 177).

The above statement generally supports earlier conclusions by several previous researchers about the relations between channel structure and sediment resources of the Colorado River in Grand Canyon (Howard and Dolan 1981; Schmidt and Graf 1990; Rubin et al. 1990;

Melis et al. 1994; Webb 1995). However, for a variety of reasons, these previous studies did not attempt to classify reach-varied geomorphology of the river channel based on extensive measurements of fan-eddy complexes throughout the system. According to

Parasiewics (1996) with respect to rivers in general: "The main difficulty with river morphology/hydraulic sampling is the high variability and dynamics of the system. This requires the collection of large data sets and subsequently a large expenditure of time."

These reasons, combined with the added complications of logistics and access, likely explain why system-wide analyses of Colorado River geomorphology in Grand Canyon have not occurred until now.

Objective

In response to the perceived need by scientists and managers for a better-defined, reach-varied geomorphic framework of the Colorado River in Grand Canyon, the objective of this study is to examine Schmidt and Graf's (1990) eleven reaches within the context of new and more inclusive data for fan-eddy complex characteristics between river miles 0 and 189

225. As part of this evaluation process, these geomorphic reaches are evaluated using measurements of certain key attributes for all fan-eddy complexes related to tributaries that were termed "geomorphically significant" by previous studies (Melis et al. 1994). Hence, the first goal of this study is to identify whether the existing reach designations proposed by Schmidt and Graf (1990) might be revised to provide a clearer picture of how sediment resources (DOI 1995) are related to channel geomorphology that is uniquely dominated by bedrock geology, structure and tributary debris flows.

METHODS

Key attributes described by Schmidt and Rubin (1995) for fan-eddy complexes were measured remotely from aerial photographs (1:4,800) taken at a constant discharge of

226 m3/s in March, 1996. The measurements were made from rectified scanned images of the photographs using an electronic planimeter within the Map and Image Processing

System package (MIPS). The error for linear measurements made from the scanned aerial images error was estimated at three percent or less, while the error associated with planar area measurements was estimated to be less than six percent. Planimetry of photographic images was calibrated using data from 140 alluvial fans where linear and plan-area measures were derived from topographic surveys. All remotely-derived length and area

measurements for fan-eddy complexes were made by the author using consistent methods.

Measurement error was generally small, and was not considered to be a significant factor

affecting relative comparisons of fan-eddy complexes made throughout the 225-mile long

study area. 190

Measurements of fan-eddy complexes were made at all 529 tributary-confluences

along the Colorado River through Grand Canyon designated to be "geomorphically

significant" by Melis et al. (1994; Appendix B). Attributes of fan-eddy complexes

measured included: 1) the percentage of the mainstem channel constricted at the tributary

confluence by an alluvial fan, C,„ which was defined by:

2 W Cw = 1 X 100, (5.1) (W u + vvu)

where Wr = the maximum constricted (narrowest) channel width occurring along the distal

edge of the alluvial fan, W, = the width upstream from the rapid, and Wd --= the width

downstream from the rapid below the expansion zone; 2) the spacing in river miles between

fan-eddy complexes, determined by the total number of constricting debris fans per river

mile; 3) the plan-view area of the alluvial fan; 4) the shape of the fan, determined as the maximum length dimension of the fan parallel to the channel divided by the maximum width of the fan perpendicular to the channel; 5) the unconstricted channel width up and downstream from the tributary confluence, determined by averaging the measurements Wu and Wd; and 6) the total length of a rapid, if one existed, in conjunction with the constriction measured from the beginning to the end of turbulent whitewater in the aerial image.

In addition, topographic survey, particle size and lithologie source data were collected on 147 pre-dam, river-reworked alluvial fans using standard field methods

(Appendix I; also, see Wolman 1954; Melis et al. 1994; Rice and Church 1996; and the methods section of chapter 3 in this volume). These data were examined to better- 191 characterize the form and composition of stable, constricting alluvial fans that resulted from

the interactions of debris flows from bedrock tributaries and mainstem floods that occurred

before regulation.

The topographic data from these fans was used to determine fan-height and the ratio

of fan-volume to plan-view area, from reaches with different channel widths and bedrock

geology (Appendix I). Fan heights and volumes were estimated using the SDR Map

software, by Sokkia Inc. Fan geometry has been identified by previous researchers as an

important factor in the stability of recirculating eddies and the deposition and erosion of

sand bars at varying flows (Melis et al. 1994; Schmidt and Rubin 1995). Relative

comparisons of fan geometry were made by determining the maximum heights and

volumes of sampled fans above a baseline elevation that can be recognized on nearly all

fans throughout the system (a distinct line of riparian vegetation associated with the

maximum peak discharge of the Glen Canyon Dam hydropower plant (about 850 m3/s) that

occurred almost daily between 1986 and 1991. Particle-size and lithologie source data were collected and analyzed primarily to:

determine what percentage of the sampled reworked fan surfaces were boulders (>256

mm), determine whether any of the bedrock lithologies dominated reworked fan surfaces, fans and determine the degree of heterogeneity of clast lithology comprising river-reworked The relative to increasing lithologie source availability from Lees Ferry to Diamond Creek.

particle-size data were analyzed to determine the average coarseness of most reworked within alluvial fans (as determined by the percentage of boulders on fan surfaces)

geomorphically distinct reaches. The lithology of clasts on reworked fans was also

analyzed to determine which, if any, of the bedrock lithologies found in Grand Canyon therefore imposing were responsible for significantly aggrading stable alluvial fans and 192 lithologie control on flow and sediment.

The data for measurements made on fan-eddy complexes from aerial photographs

(fan plan-area, channel constriction, fan shape, channel width, fan spacing and rapid length) were compiled in a spreadsheet and divided based on the beginning and ending points (in river miles) of the eleven previously-designated geomorphic reaches of Schmidt and Graf (1990; Table 5.1). Within these imposed geomorphic divisions, means, standard deviations and standard errors were calculated for the fan-eddy complex attributes and compared with one another to determine whether means in adjoining reaches differed significantly.

Next, the unsegregated data were used to produce scatter plots using a two-step process. Initially, raw data were plotted with a lowess (locally weighted scatterplot smoothing) curve overlaid on the scatterplot. Lowess is a robust curve-smoothing technique. The user determines how many points to estimate between the minimum and maximum value of the independent data (the horizontal span of the scatterplot), and what proportion of the data is to be used to determine the estimates. For each variable, I specified that 100 points be estimated, using 5% of the data for each estimate. For the 50th point, the 5% of the raw data that are centered on the midpoint of the scatterplot are used to produced a smoothed estimate. Data in the center of the selected 5% contribute more to the estimate than data at the edge of the 5% "window" - hence "locally weighted." Smoothing is related to the proportion of data used in the individual estimates - the larger the amount of data used, the more smoothing occurs. I experimented with several values and chose 5% because it allowed rapid change in values of the dependent variables over relatively short distances along the river while still smoothing enough to show large-scale patterns. 193

After the initial lowess curves were plotted, the variables were resealed (the y- values) so that the estimated values ranged between 0 and 100. This allowed me to plot the lowess curves for all variables with the range of estimates (the estimated y values) covering the same scale. The resulting graphs made maximum use of the y axis so that the finest possible visual distinctions among the estimates were possible, and permitted me to overlay variables that would otherwise have very different values. The overlays, in turn, allowed easy comparison of patterns of change in the fan-eddy complex attributes measured along the river corridor. To provide beginning and ending points for the lowess process that extended beyond the limits of the study area, channel attributes of Glen Canyon (river miles

-16 to 0) and river miles 226 through 236, were measured and included in the data set.

Incorporation of these additional data allowed lowess curve smoothing to establish stable trends before and after reaching the endpoints of the 225-mile long study area.

At the mid-points of abrupt changes in all possible combinations of conjointly- trending, smoothed-curve plots (eg. width versus fan size, width versus constriction, etc.),

I established new geomorphic river reaches. Once these divisions were made, mean values for five of the six fan-eddy complex attributes (area, shape, constriction, spacing, and unconstricted channel width) were calculated based on the newly segregated blocks of data, and then compared with means of the Schmidt and Graf (1990) reaches. Rapid length was not considered beyond these analyses because rapid-length plotting trends did not vary consistently with other key channel attributes; reasons for this are not clear.

In addition, values for debris-fan density (fan spacing) were determined by averaging the number of alluvial fans observed per river mile throughout the study area.

These data are important because net storage of sediment in the channel depends, in part, on how many fan-eddy complexes occur per river mile. The combined influences of fan 194 spacing and shoreline types (vertical cliffs, near-vertical cliffs, sloping-talus, etc.) are also important in overall storage of fine sediment (sandbars) throughout the river corridor.

Therefore, general shoreline classifications were established on the basis of how various shoreline types limit deposition of terrestrial sandbars. These types include vertical bedrock cliffs, near-vertical bedrock cliffs, and sloping-talus shorelines. These shoreline- type classifications were combined with reach characteristics of alluvial fans and rapids to derive five general shoreline types: 1) vertical cliffs with few large fans or rapids; 2)

sloping-talus with many large fans and rapids; 3) near-vertical cliffs with few fans, but many large rapids; 4) sloping-talus with many fans, but few large rapids; and 5) vertical

cliffs with many fans, but few large rapids. Finally, each river mile of the study area was

assigned to one of the five shoreline categories.

RESULTS

Evaluating the geomorphic reaches of Schmidt and Graf (1990)

On the basis of statistical analyses use of morphologic data (Table 5.2; and fig.

5.4), key features of fan-eddy complexes in Grand Canyon are similar among several

adjoining geomorphic reaches under Schmidt and Graf s (1990) classification. This is

particularly true for channel width, constriction, fan size and fan spacing. For example, in

Schmidt and Graf s reach classification, channel width does not vary significantly between in reaches four and five (river miles 36 through 77), and channel constrictions are similar

reaches one and two (miles 0 through 22), and in reaches three, four and five (miles 22

through 77). Fan spacing is similar throughout reaches two and three (miles 11 through

36), four and five (miles 36-77), seven and eight (117 through

195

TABLE 5.2. Reach-averaged channel characteristics for previously designated geomorphic

reaches of the Colorado River through Grand Canyon (Schmidt and Graf 1990), based on

measurements of fan-eddy complexes at 529 tributary confluences.

4 4 u 4 u.5• 8 2 g0 go 2 m œ.., c..) m . 4 w (..) ..: an 4 4 = 4 4 4 2._ g g v) gn 00 cn U cn cn (..) U v) ce)

sID 0 WI S N C, 00 6 ,6 r c7; — to •-n r--- oo N Ci 00 cr a. 0, h

nC) er M. Oi r N •—• +I -H -H -H -H +I +I +I -H +I — <1. —50 00 N •cr M •ec en cs) r N

— sO 50 Ci 0 41- — )r) re. co en E 8 8 s- E; 8 Z,E8E 5- 6666oOo666o

— — N =' d Q QCCQCQQ ti +1 -H +I +I +I -FI +I +1 oo <-4 r4

en 'I' •cr re. - N N $31 0 0, C,1 0, h Q 000 6 (-4

`41 7i 7i II T-1 +1 +I -71 .+1

s.0 V. 0, •"' 00 m 0 00 VZ ("4 0

CO VCi ,n — N — (+) N m s.0 •-n cr,

,0 ,O.

0 N m ,0 h Le-1 196

0 2 3 4 7 9 10 11 4.0 FAN SPAONG 3.5 3.0 2.5 2.0 1.5 1.0 0.5

0.0 4 $ 6 7 8 9 10 Ii 0.025 D DEBRIS-FAN AREA

'M 0.020

0.015

.4 0.010

0.005

0.000 3 4 6 8 9 10 1

MOM 4.0 ELONGATED FAN SHAPE, (LENGTH/WIDTH) ^ 4 " 3.5

3.0

2.5

u. 2.0

1.5 SEMI- COMMA 1.0 4 5 7 9 10 11 GEOMORPHIC REACHES (top) and FIGURE 5.4. Reach-averaged fan-eddy complex attributes, standard errors of Schmidt and Graf standard deviations (bottom) for the eleven geomorphic reaches Width. B. Local mainstem (1990) along the Colorado River in Grand Canyon. A. Channel E. Fan shapes. constrictions at tributaries. C. Fan spacing. D. Plan-view area of fans. 197

140), and ten and eleven (160 through 225). System-wide, fans are also of similar size

(plan-view area) in reaches two and three (miles 11 through 36), while fan shape

(length/width, the least variable parameter measured) is similar in reaches one and two

(miles 0-22), five and six (miles 61-118), and nine, ten and eleven (miles 140 through

225). Statistically similar averages of fan-eddy complex attributes in previously established adjoining reaches (Schmidt and Graf 1990) warrant a new examination of channel trends aimed at devising a geomorphic-reach classification that expresses greater variability based

on channel features that control flow and sediment.

The basis for newly-proposed geomorphic reaches

To determine new geomorphic break points between river reaches, smoothed-curve

plots were prepared and studied for obvious, large-scale trends in fan-eddy complex

features. Positive covarient trends are obvious in the lowess, curve-smoothed plots,

especially with respect to channel width, fan area and constriction (fig. 5.5a and c). In

contrast, channel width is inversely related to fan shape (fig. 5.5b). Mainstem channel

constriction caused by alluvial fans is positively related with fan size, but inversely related

with respect to fan shape (fig. 5.5 d and e). Overall, the smoothed-curve plots of channel

characteristics showed a remarkable degree of covarience.

The reasons for the relationships observed in the lowess plots between channel

width, fan area, fan shape and constriction are rather intuitive when considered in the

context of general principles of sediment transport and flow relative to stream power in a bedrock/alluvial river (Baker and Costa 1987). For example, a narrow channel results in through increased greater erosion of debris fans (reducing fan size and percent constriction)

competence related to locally-elevated stream power (fig. 5.5a and c). Also, narrowed —n421•=411,11840 --- - f-- 198

2G 40 50 50 100 120 160 160 160 2110 220 River mlle

O 20 100 120 40 160 150 200 220 Wet mile

0.9

OA -

0.7

0.6

OA

0.3 -

0.0 20 40 00 100 020 140 160 100 200 2e, River mile FIGURE 5.5. Lowess smoothed-curve plots of fan-eddy complex attributes measured between river miles 0 and 225 of the Colorado River in Grand Canyon. A. Channel width and fan area. B. Channel width and fan shape. C. Channel width and percent local channel constriction. 199

**;

40 la 220 140 101 144 XXI River ma*

•

124 144 ISO 1114 2192 220

Iver mile

la -

SJ

sa 144 149 on River aft FIGURE 5.5. Continued. Lowess smoothed-curve plots of fan-eddy complex attributes measured between river miles 0 and 225 of the Colorado River in Grand Canyon. D.

Percent local channel constriction and fan area. E. Percent local channel constriction and fan shape. F. Percent local channel constriction and rapid length. 200 channel conditions generally support smaller fans simply owing to space constraints.

Likewise, in narrow channels, where fan reworking is greatest, fans become relatively more elongated as they are graded, with fan shape (length/width) progressively increasing beyond a value of 1.0. In contrast, wider reaches tend to support larger and more semi- circular fans whose shape factors approach 1.0 (length equal to width; fig. 5.5b).

Concurrently, wider channels that support larger alluvial fans with shape factors

approaching 1.0 result in relatively higher-percentage channel constrictions (fig. 5.5d and

e). With respect to storage of fine-grained sediment, wider channel reaches supporting larger fans and cobble bars result in larger eddies, and in turn, do provide more space

(volume) for sand storage per river mile as terrestrial bars, subaqueous deposits in eddies

and pool deposits between fan-eddy complexes. Likewise, fan size, and hence, sand

storage also tend to increase in those reaches where the channel has sloping-talus

shorelines, as opposed to vertical bedrock cliffs.

Channel width and rapid length are also clearly covarient along the Colorado River

in Grand Canyon, but this relation occasionally reverses from inverse to direct for reasons

that are not clear (fig. 5.51). These reversals are likely related to space constraints on fan

evolution in the narrowest channel reaches, and perhaps to an upper threshold of fan reworking related to channel geometry (mainly width). In such reaches, boulders are often

strewn along the channel bottom forming deposits that cause severe rapids, rather than

accumulating as terrestrial fans (Howard and Dolan 1981). Rapid length can be viewed as

a proxy for the degree of separation between low-velocity pools where fine sediment may

be stored (in terms of local-scale channel gradient). However, since it remains unclear

whether significant storage of sand and finer sediment throughout the mainstem occurs

predominantly in mainstem pools or in eddies, rapid length was not used as a key 201 parameter of fan-eddy complexes during revision of geomorphic reaches.

On the basis of the abrupt changes observed in lowess smoothed plots of channel width, fan area, fan shape, and channel constriction, the eleven reaches of Schmidt and

Graf (1990) were combined into six main geomorphic reaches (Table 5.3) numbered and named: 1) Upper Wide I (miles 0-8); 2) Upper Narrow 1(8-38); 3) Middle Wide 11 (38-77);

4) Middle Narrow 11 (77-170 - with two distinct sub-reaches 4a) MN Ha [87-100] and 4b)

MN IIb [117-128]); 5) Lower Wide III (170-213); and 6) Lower Narrow III (213-225).

On the basis of conclusions by previous researchers as to which channel attributes are the most important geomorphic controls on flow and sediment (Howard and Dolan 1981;

Schmidt and Graf 1990; Schmidt and Rubin 1995), average channel width was ultimately chosen as the primary variable used to designate these six reaches. Fan spacing, area, shape and mainstem constriction were considered as secondary factors examined for covariance with channel width in assessing and establishing break points in the system- wide data set for fan-eddy complexes. Break points for the newly proposed reaches were selected with relative ease on the basis of visual examination of the lowess plots. Reach four (Table 5.3; Middle Narrow II) includes two relatively short widened-channel anomalies. These sub-reaches occur within what is otherwise a continuous, narrow reach, and are therefore not classified as major, stand-alone reaches. The causes of these widened sub-reaches are apparently related to: 1) large-scale, north-south trending faults that intersect the Colorado River deep in the interior of eastern Grand Canyon (sub-reach MN Sandstone downward to Ha, [miles 87-100]) and; 2) vertical displacement of the Tapeats see river level in the vicinity of Stephens Isle (sub-reach MN Jib; from miles [117-118];

1:62,500 scale geologic map by Huntoon et al. 1986). 202 TABLE 5.3. Revised geomorphic reaches and subreaches of the Colorado River through Grand Canyon, based on fan-eddy characteristics measured near 529 tributary confluences.

— tr)

srn .1- as cts

N r N 6 C•-• +I +I -H -H +1+1 +I +I ev cce s 'CI' NI N R _ 8 8 8 86 8 8 8

go r- E 5 g C''' g 6 6 6 6 6 d 6

R _.—, ..-..—... z...... 1.-. ....z. .••••n d 6 6 6 d 6 +1 $'1 +1 -H +I +.1 -H TI N .0 en s 0. t+'. .11 'I" N N N N N N N N

R Rt. (-7!‘

N 0' — 6 N

^ N +i +I +I -H 00 r•-, cr, -1. en 00 ‘,0

8 F:-.' — . c , ,,,,„ _ N 00 rs 00 ..- 1 1 CO en . , .—. ---k c:, n cl .0 co s cr ,ct s c, co c•-•,N — (-..4 •••••n N ..f) 203

Reach-averaged means for fan-eddy complex features for the revised classification

show higher variability among adjoining reaches (fig. 5.6) than seen in the Schmidt and

Graf (Tables 5.1 and 5.2) framework, particularly with respect to channel width, fan area

and fan spacing (fig. 5.6 a, c and d). Increased variability between adjoining reaches

indicates that the new reach classification (Table 5.3) is a more distinct representation of

river geomorphology than previously provided (Table 5.1). In the new reaches, variability

of channel constriction and fan shape were not as dramatic as in the case of the other three

attributes (fig. 5.6 b and e), but they become more distinct when MN Ha and MN Hb were

combined with the rest of the Middle Narrow II reach (combined measurements from river miles 77-170).

Several reach divisions designated by Schmidt and Graf (1990) based on the

appearance of individual geologic formations at river level generally carried through to the

six new reaches. This occurred despite the fact that appearance of individual geological

formations at river level was not considered as part of the revision process (Tables 5.1 and

5.3). Many of these new-reach break points occur within three miles or less of those

separating Schmidt and Graf's reaches: 1 and 2 (Upper Wide I and Upper Narrow I); 3

and 4 (Upper Narrow I and Middle Wide II); six and seven (Middle Narrow H and sub-

reach MN In) ; seven and eight (sub-reach MN Hb and Middle Narrow II) ; and ten and

eleven (Lower Wide III and Lower Narrow III). However, on the basis of evaluations of

Schmidt and Graf's (1990) geomorphic-reach classification, the influence of individual geologic formations on channel characteristics is less important than are stratigraphically- grouped formations of similar lithologic characteristics (fig. 3.2).

Despite the revision process in which eleven reaches were combined into six reaches, the geological influence of bedrock lithology and structure remained the single 160 A C1-1.0NWEL. WOTH 322 140 204 120 100 80 60 40 6 20 0 60 8 n.4111 CONSTRICTION AT TRIBUTARY MOUTHS 50

40

30

20

10

0 4. 40 5 4.0 FAN SPACING n-411 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.025 D DEBRIS-FAN AREA n-415 • 0.020

• 0.0 15

• 0.010

•:LC 0.005

0.000 4a 45

miety 4.0 n-41' • cotc..mo FAN SHAPE, (LENGTH/WIDTH) 3.5

3.0

2.5

• 2.0

1.5 SENI - CM 0.4A 1.0 40 5 REVISED GEOMORPHIC REACHES FIGURE 5.6. Reach-averaged fan-eddy complex attributes, standard errors (top) and standard deviations (bottom) for revised geomorphic reaches, based on system-wide data for the Colorado River in Grand Canyon. A. Channel Width. B. Local mainstem constrictions at tributaries. C. Fan spacing. D. Plan-view area of fans. E. Fan shapes. 205

most important factor in establishing breaks in reach-scale geomorphology of the Colorado

River in Grand Canyon. The fact that these new reach-breaks are highly related to

lithology and faulting supports the previous conclusion by Howard and Dolan (1981) that

the fundamental controls on sediment resources and navigability in Grand Canyon are

dependent on regional geology (Dolan et al. 1978). This conclusion is not fundamentally

different than that published by John Wesley Powell (1895) and Clarence Edward Dutton

(1882) over a century ago. Overall, these analyses support the idea that a close relationship

exists between the geomorphic processes of the Colorado River and Grand Canyon

geology, especially large-scale faulting. One other important factor that significantly

controls storage and deposition of coarse (stable fans) and fine sediment (sand bars) along

debris-fan dominated rivers, is the degree to which sloping-talus shorelines coincide with

closely spaced fan-eddy complexes, versus vertical or near-vertical bedrock shorelines that do not.

The influence of fan-eddy complexes on fine-sediment resources

How much flow, sediment transport, and sediment storage are influenced by debris flows along the Colorado River, depends on the percentage of tributaries that support stable alluvial fans (fig. 5.7). This too reflects the influence of lithology and structure on channel geometry, as well as Canyon-wide spatial patterns of debris-flow probability and distribution of source sediments (Griffiths et al. 1996). In Schmidt and Graf's (1990) framework, reach six (miles 77-117) supports the fewest number of fans relative to tributaries. This fact reflects the narrow channel geometry throughout the reach that results in relatively high unit stream-power during reworking floods (fig. 5.7a). The first 23, and the last 42 miles of reach four (Middle Narrow 11; miles 77-100 and 128- 206 1 00.0

80.0

zi— 60.0 w (..) cc 8-2 40.0

20.0

0.0 1 2 3 4 5 6 7 8 9 10 11 PREVIOUS GEOMORPHIC REACHES 100.0

80.0

zH 60.0 w (..) cc wo_ 40.0

20.0

0. 0 1 2 3 4 4a 4b 5 6 REVISED GEOMORPHIC REACHES

FIGURE 5.7. Percentages of tributaries with alluvial fans for previous and revised geomorphic reaches of the Colorado River in Grand Canyon, by reach. A. The eleven reaches of Schmidt and Graf (1990). B. Revised geomorphic reaches. 207

170) support the lowest percentage of fans relative to tributaries (fig. 5.7b). If storage and deposition of fine sediment is in fact controlled by fan-eddy complexes, then this type of reach should not be expected to support extensive sediment resources to the same degree as reaches with more densely spaced fans and sloping-talus shorelines. However, wider reaches with higher fan densities, large constrictions (eddy size) and sloping-talus shoreline types should support abundant terrestrial sand deposits, including separation, reattachment and channel-margin bars (see Schmidt and Graf 1990; Table 5.1).

The high frequency of tributaries that support alluvial fans in the Middle Wide II,

MN Ilb and Lower Wide HI reaches is reflected in the close spacing of fans characterizing lower Marble and eastern Grand Canyon, as well as in middle and western Grand Canyon

(fig. 5.8a). However, the relation between the spacing of fans and rapids throughout the system is mainly inverse (fig. 5.8a and b); that is, many large fans are not associated with rapids. This counter-intuitive relation is most likely another result of channel geometry and

variable fan reworking, relative to both reach-varied stream power, and the non-random

debris-flow potential recently identified by Griffiths et al. (1996; see Plate 14), throughout

Grand Canyon.

As shown by the lowess plots and patterns of fans and rapids, some relations

between the mainstem (constriction and width) and its numerous fan-eddy complexes (fan

spacing, as well as number and length of rapids) appear to be inconsistent throughout the

study area (figs. 5.5 and 5.8) for reasons not fully understood. Until recently, emphasis

by researchers on channel constriction (Kieffer 1990; Schmidt and Graf 1990), and channel

control by individual geologic formations at river level (Schmidt and Graf 1990), detracted

from previous efforts to identify other important factors that control flow and sediment

(Schmidt and Rubin 1995). In contrast to the results of Maier (1990), the average 208 cc a z 0.08 u_< o1-- 0.06 z co< 0.04 u.. o o 0.02 1=< cc

50 100 150 200 River Mile 8 B 7 - 6 - 5 4 3 2 1 0 0 25 50 75 100 125 150 175 200 225 RIVER MILE

4 C 3 _

3 --

2 -

2 -

1

1 o I I I 225 o h 75 1i 150 RIVER MILE

FIGURE 5.8. System-wide relations between tributary drainage areas and alluvial-fan characteristics associated with fan-eddy complexes along the Colorado River in Grand

Canyon. A. Basin-to-fan area ratios. B. Alluvial fans. C. Rapids. 209

percentage of the mainstem constricted by alluvial fans is 42.5 (±0.8) percent, a value

significantly lower than her reported estimate of about 50 percent (fig. 5.9). When zero

values are included for the 127 tributaries where there are no visible subaerial fans, the

mean constriction drops to 36.2 (±0.9) percent. Although mainstem channel constriction

was considered as one factor when revising geomorphic reaches, the degree-of-channel-

constriction may not be as important as the degree to which tributaries support fans from

one reach to another. Overall, fan spacing, channel width and fan size are likely more

important factors than the degree-of-constriction in controlling flow and sediment transport

along the Colorado River in Grand Canyon. However, the degree-of-constriction does

correlate positively with fan size, and therefore it is also likely to be a good indicator of

eddy size in wider reaches where constrictions are generally high.

In contrast to the conclusion of Schmidt and Graf (1990), the above results suggest

that grouped geologic formations having similar lithologie characteristics are a more

significant factor than are individual geologic formations in controlling channel-reach variability in Grand Canyon (figs. 3.2 and 5.6). On the basis of particle-size and lithologie source data from reworked fans, individual formations and groups of formations with similar lithologies may control both flow and sediment transport in the river even when they are high above river level (fig. 3.2). On the 117 alluvial fans sampled for particle-size and lithologic-source data, boulders derived from the Redwall and Muav Limestones dominate the reworked fan surfaces (fig. 5.10a). On most fans sampled, clasts from these two stratigraphically adjoining formations (fig. 3.2) occur twice as frequently as the next largest contributing lithology, the formations of the Supai Group (figs. 3.2 and 5.10a).

With the exception of the stratigraphically adjoining Kaibab and Toroweap Limestones, most other lithologies of the Paleozoic and Proterozoic sections are only minor Q 210 Q

,

o co

Q co

Q 71-

o c\J

Q Q Q Q 0 co .1. CV 1--

SNOLLOIELLSNO0 NVA-S11:1930 A01:1391MIN

FIGURE 5.9. Statistical distribution of local mainstem constrictions caused by alluvial fans along the Colorado River in Grand Canyon. 211 60.0 -r A FULLY REWORKED FANS 50.0

40.0

30.0

20.0

10.0

0.0 BEDROCK LITHOLOGIES 120.0 FULLY REWORKED FANS n = 117 100.0

80.0

60.0

40.0

20.0

0.0 1 2 3 4 5 6 REVISED GEOMORPHIC REACHES

FIGURE 5.10. Source lithologies of clasts comprising the surfaces of 117 pre-dam, reworked fans sampled from fan-eddy complexes of the Colorado River in Grand Canyon, by revised geomorphic reach. A. Percentages of clasts from each geologic formation found on fans. B. Percentages of dominant lithologic sources of clasts found on fans, by revised geomorphic reaches. 212

contributors to stable alluvial fans throughout Grand Canyon (fig. 5.10a).

Within the revised reach classification presented above, clasts from the Kaibab and

Toroweap Limestones dominate fans in reach one (Upper Wide I), while clasts from the

Supai Group Formations dominate fans in reach two (Upper Narrow I). However, clasts from the Redwall and Muav Limestones dominate the lithologie distributions of most other fans sampled throughout reaches four through six (fig. 5.10b; Middle Wide II, Middle Narrow II, Lower Wide III and Lower Narrow III).

On average, particle-size data from the reworked fans show that boulder-sized particles (>256 min, b-axis diameter) exceed fifty percent of particles counted in all six revised geomorphic reaches (fig. 5.11a). The average percentage of boulders on all reworked debris fans sampled varies little among all six reaches, and almost not at all among reaches two through four in the revised classification. This is interesting, considering that channel width varies significantly between all of the reaches (figs. 5.11a and 5.6a). A significant difference in the average boulder content of fans in adjoining reaches one and two (Upper Wide I and Upper Narrow I; fig. 5.11a) corresponds with a change in dominant clast lithology (Kaibab/Toroweap versus Supai Group). However, a similar difference in average boulder content between reaches five and six does not coincide with a change in dominant clast lithology (fig. 5.11a).

Average fan heights for 141 reworked fans show that this fan-eddy complex attribute also varies significantly among all adjoining reaches of the revised classification

(fig. 5.11b). Variability in reach-averaged fan height follows an increasing trend from reaches one through three (Upper Wide I, Upper Narrow I and Middle Wide II; Table 5.3), regardless of changes in channel width, but varies directly with channel width from reach three to four, four to five, and five to six (fig. 5.11b). Fan height is a general reflection of 213

cn 100.0 cc FULLY REWORKED DEBRIS FANS n . 117 - w 0 (PARTICLES WITH B-AXIS >256 MM) - 80.0 m 0 co L__

I-uiz- 60.0 0 ec a_w 40.0 w 0 =< 20.0 u..1 > < 0.0 2 3 5 6

20.0 n = 141 -

AVERAGE HEIGHT OF REWORKED DEBRIS FANS -

- (ABOVE —850 M 3/S DISCHARGE LEVEL) 15.0 -

-

10.0 -

5.0

0.0 2 3 4 5 6 REVISED GEOMORPHIC REACHES

FIGURE 5.11. The reach-averaged percentages of boulders (with dominant boulder lithologies) on alluvial fans and relative fan heights for 117 pre-dam, reworked fans sampled from fan-eddy complexes along the Colorado River in Grand Canyon, standard errors (top) and standard deviations (bottom) for revised geomorphic reaches. A. Percent boulders (with dominant lithologies). B. Fan heights (above —850 m 3/s discharge level). 214 fan volume, but may not parallel fan plan-view area except in wide reaches, such as three and five (Middle Wide II and Lower Wide III), where alluvial fans are largest (Table 5.3; figs. 5.6 and 5.11b). Melis et al. (1994) and Schmidt and Rubin (1995), have stated that fan height is an important controlling factor influencing the stability of recirculating eddies and deposition of sand bars during high river stages associated with large floods. Since recirculating eddies tend to enlarge and strengthen with increasing stage (Schmidt et al.

1993), eddies controlled by high-profile fans are more likely to retain sand and remain stable (such as those found in reaches three and five), than are eddies that are significantly overtopped during floods. Hence, fan height is another important factor that favors enhanced storage of fine sediment in reaches with tall fans (three and five) over others where low-profile fans are abundant (one, four and six).

On the basis of reworked fans sampled for lithologic-source data, most alluvial fans in Grand Canyon are mostly composed of clasts derived from two geologic formations

(fig. 5.12), although the two sources vary downstream from Lees Ferry. For example, between river miles 0 and 38, two to three lithologies comprise ten percent or more of fan clasts: the Kaibab and Toroweap Limestones, the Coconino Sandstone and the Esplanade

Sandstone of the Supai Group ((figs. 5.12a and 3.2). Below river mile 38, the remainder of the Paleozoic and Proterozoic sections rise above river level and the lithologie heterogeneity of reworked-fan clasts declines to mainly two lithologies. In western Grand

Canyon, the two most common lithologie groups that produce boulders found on fans are:

1) the Redwall/Muav Limestones, and 2) the sandstones of the Supai Group (fig. 5.11a).

Despite the high lithologie diversity found in the strata of Grand Canyon (fig. 3.2), pre- dam, river-reworked fans reflect a remarkably low degree of lithologie heterogeneity (fig.

5.12b). 215 A n 117

-a> 4235 03r> 0 0 0 Œ. 0 0

0 OD 43X> <11=00=13.0 0 0 033> 4:1103330 0 MX> eX> 0 000 CO C

0 a) 0 75 150 225 RIVER MILE 80 n- 117

70

60

50

40

30

20

10

0 2 3 4 5 NUMBER OF BEDROCK UTHOLOGIES CONTRIBUTING >10 PERCENT1 OF CLASTS ON FAN SURFACE FIGURE 5.12. Lithologic diversity of clasts comprising pre-dam, reworked alluvial fans of the Colorado River between Lees Ferry and Diamond Creek in Grand Canyon. A.

Number of lithologies contributing >10 percent of clasts by river mile. B. Statistical distribution of reworked fan-clast heterogeneity, with respect to source lithology. 216

Variability of shoreline types occurs both among and within the six revised reaches.

Only in sub-reaches MN-lia and MN-IIb is shoreline type continuous. As stated above, sloping-talus river banks that co-occur with tall and large, closely spaced alluvial fans, promote the greatest storage of fine sediment. Such geomorphic settings provide larger areas as foundations for terrestrial sand bars and are associated with wider channel reaches that support larger eddies and pools where subaqueous sand and finer sediment can accumulate. The five shoreline types designated in this study represent a spectrum in which vertical bedrock cliffs at river level promote deposition and stability of sand bars the least, while sloping-talus shores promote sand-bar deposition the most. The number and size of bars vary within this range of shoreline types depending on the additional fan-eddy complex characteristics already described.

The distribution of combined shoreline types and fan-eddy complex attributes provides the basis for a conceptual model of where fine-sediment storage is most likely to occur under given conditions of flow (mainly stage) and sediment availability at any one site (Table 5.4). Assuming that sand inputs are cumulative downstream from the Paria

River, and increase markedly below the confluence of the Little Colorado River (Andrews

1991; river miles 1 and 61 respectively), sand storage should occur most readily within fan-eddy complexes in reaches three, five and six, or other areas where sloping-talus shorelines co-occur with large, closely spaced fans (Table 5.4). In contrast, significantly less sand storage should occur within fan-eddy complexes in reaches one, two and three, or other areas where shorelines are vertical to near-vertical cliffs that co-occur with sparse fans with low-profile geometries (Table 5.4).

Of the five types of shorelines designated in this study, sixty-seven miles of the study area (29.7 percent) contain vertical cliffs that border relatively few large fans (fig. 217

TABLE 5.4. Shoreline characteristics, by river mile (0 to 225), associated with revised geomorphic reaches of the Colorado River through Grand Canyon.

hi !I. ON 0 N N 6 $ n o 0 h 0 I t..... N VI o N

C 00 N r•J

CO N $ 6

—

8 a'

rn N h N — N oo — oo Co er, m oo O 00 en h en 'd' 218

5.13). Areas within such reaches are unlikely to favor sand storage as terrestrial sand bars, but may store large volumes of sand in mainstem pools (fig. 5.13; reaches one, two and four). Throughout reaches three and four (Table 5.3), fine-sediment storage likely occurs as both subaqueous eddy and pool deposits. However, abundant terrestrial sand bars are found mainly in areas of reach three that have sloping-talus shorelines, and to a lesser extent in reach MN-IIb where many fan-eddy complexes co-occur with vertical cliffs (fig.

5.13). Regardless of sediment availability or flow regime, the greatest volumes of channel-stored sand are likely to occur in reaches two, three, five and six, because many miles of sloping-talus shoreline in those reaches co-occur with large, closely spaced fan- eddy complexes; some of which are the largest in the system (fig. 5.13). Despite variability in fine-sediment inputs through time, the accumulated channel length in these combined reaches that are bordered by sloping-talus shorelines (a total of 101 river miles), should retain, on average, the greatest volume of sand in the system (fig. 5.13). During pre-dam times, sediment availability was probably not a limiting factor on sand storage or on sand bar stability along the mainstem. The sole limiting factor in that era was most likely channel geomorphology. On the basis of the geomorphic framework of the river presented in this study, and the above conceptual model for geomorphic controls on flow and sediment storage, annual pre-dam floods probably deposited the largest sand deposits within and between fan-eddy complexes controlled by large, high-profile fans. On the basis of the above analyses, the large fans generally evolved in wide channel settings bordered by sloping-talus shorelines (Table 5.4 and fig. 5.13).

The conceptual model described here is generally supported by observations of pre- dam sand bars recorded on photographs made throughout the study area in the winter of

1889-90 by Robert Brewster Stanton (Appendix B), and from pre-dam aerial photographs 219

0 co (co

U3Al21 3Aliv1niA1flo

FIGURE 5.13. Relative percentages of shoreline-types found along the Colorado River between Lees Ferry and Diamond Creek in Grand Canyon. 220 taken in the 1930s. Sand inputs to the mainstem below Glen Canyon Dam declined by over 80 percent after dam closure in 1963 (Andrews 1991). Despite this fact, the general pattern of sand storage in the post-dam era generally parallels that seen before dam closure relative to the geomorphic framework described here (Kearsley et al. 1994; Schmidt et al.

1994a and 1994b), although most sand bars have decreased system-wide.

DISCUSSION AND CONCLUSIONS

Alluvial fans along the Colorado River in Grand Canyon result from the cumulative effects of the magnitude and frequency of debris flows from tributaries and the magnitude and frequency of floods on the Colorado River. On the basis of repeat photography over the last century and particle-size data, I conclude that, in the absence of debris flows from tributary canyons, alluvial fans in Grand Canyon were relatively stable features of the mainstem before regulation. During the pre-dam era, newly aggraded debris fans would be reworked over time periods of hours to days (Webb et al. 1996) by large floods that occurred almost annually in the mainstem. Although the total sediment flux to and away from fans may have varied greatly before 1963, total aggradation rates driving fan evolution were probably relatively low because only a relatively small percentage of boulders that exceeded the competence of pre-dam floods are transported to the river by most debris flows (fig. 5.3). During regulation, total sediment flux from tributary debris flows has been unaffected, whereas river reworking has decreased owing to drastically reduced annual peak discharges.

In the Upper Colorado River basin, Graf (1980) reports that regulation by Flaming

Gorge Dam increased the stability of 93 percent of the rapids of the Green River in Utah. 221

However, he determined that only about 62 percent of those rapids were probably stable before regulation (Graf 1980), a conclusion suggesting that aggradation of rapids on the

Green River may have started long before regulation (Graf 1980). A similar trend may have also existed in Grand Canyon before 1963 (see Melis and Webb 1993; and chapters three and six). Regardless of that possibility, there is no doubt that regulation has accelerated what may have been a long-term process that previously occurred at a much slower rate.

Results from this study suggest that future rates and patterns of alluvial-fan evolution and channel aggradation by both coarse and fine sediment will continue to be affected by the debris-fan dominated geomorphic framework of the river channel. Under regulated conditions, fan aggradation is accelerated, but rates remain variable downstream from Lees Ferry depending on reach-varied channel geometry, bedrock lithology at river level, geologic structure, varied debris-flow potential, and changes in available boulder- producing source lithologies in tributaries. However, fan erosion continues at greatly diminished rates owing to Glen Canyon Dam. Hence, regulation has fundamentally altered the geomorphic framework of the Colorado River through Grand Canyon, and it has done so in ways that promote tributary debris-flow processes over fluvial transport of sediment in the mainstem. The end result of this shift in process interactions is accelerated fan aggradation and the repercussions that follow changes in fan-eddy complex attributes.

Relations between pre-dam river-reworked alluvial fans and downstream cobble bars are illustrated by particle-size and lithologie data collected at river mile 62.6-R (fig.

5.3). These data show not only a downstream shift in the particle-size distribution from the reworked fan to the cobble bar, but also a shift in the lithologie sources of clasts. For instance, the Redwall and Muav Limestones produce abundant, erosion-resistant boulders 222 that commonly dominate reworked fans system-wide and contribute to the stability of fan- eddy complexes. In contrast, clasts from the less-resistant sandstones of the Supai Group

(fig 3.2) tend to dominate reworked fans in the first forty miles below Lees Ferry (fig.

5.11), but are more commonly found on highly graded cobble bars in central and western Grand Canyon.

This trend reflects the fact that softer Supai-derived clasts are transported from progressively higher elevations as the Colorado River progresses westward. Increased potential energy translates to greater kinetic energy dissipated over longer transport distances, resulting in more-highly eroded Supai clasts (small boulders and cobbles versus large boulders) at river level in the western Canyon. Hence, in terms of stratigraphie position, relatively greater numbers of Supai clasts deposit on fans in sizes well below the competence of pre-dam river floods (local unit stream power within constrictions) than do

Redwall and Muav Limestone clasts. These limestone clasts are harder than sandstone, are located at elevations nearer river level, and as a result, are subject to less mechanical erosion during transport to the mainstem (fig 3.2). Because of these transport-erosion relationships, Supai-derived clasts and other higher-elevation lithologies dominate pre-dam reworked fans in eastern Grand Canyon, but mostly dominate only cobble bars downstream from river mile 40 (fig. 3.2). In contrast, lithologies that are more erosion- resistant, such as the Redwall and Muav Limestones, Tapeats Sandstone, and various crystalline Proterozoic formations, dominate most reworked fans along the mainstem in central and western Grand Canyon.

Other erosion-resistant lithologies, such as the Kaibab and Toroweap Limestones and Coconino Sandstone, dominate reworked debris fans while they are at elevations close to river level. However, the percentages of these strata diminish on reworked fans as they 223 are shifted progressively upward from river level by large-scale faults that cross the

Canyon. Reworked fans reflect the combined influences of regional geology, lithologie characteristics, stratigraphie position, mechanical erosion, and transport by debris flows.

As a result, fan clasts from lithologies high above river level dominate cobble bars that occur in wide reaches immediately below constricting fans, while lower-elevation, more erosion-resistant strata control the stability of most alluvial fans by contributing greater shares of boulders by way of debris flows.

Particle-size data suggest that boulders with b-axis diameters of 0.5 to 1.0 meters caused net aggradation on debris fans before regulation by Glen Canyon Dam. Larger boulders also contribute to the size of debris fans, but these particles are transported more rarely in debris flows. Particle-size distributions and historic photographs reveal that many debris fans contain stable boulders greater than 1.0 m in diameter; such boulders probably remain on fans and in rapids for long periods. These particles provide an immobile foundation for additional fan aggradation. Regulated flows allow a larger percentage of finer-grained sediments deposited by debris flows to remain on top of what were formerly maintained solely as piles of boulders. While specific bedrock lithology is apparently critical in controlling channel morphology directly, it is probably more important to evaluate similar units of like lithologie characteristics, (for instance, shales, sandstones, limestones, etc.) than to devise

river-reach designations strictly based on specific formations. In addition, it is also

important to consider the channel-controlling influences of grouped or individual

formations and their lithologie characteristics relative to how they contribute boulders that

form stable fans in eddy-complexes. For instance, the direct influence of the Redwall and

Muav Limestone at river level on channel morphology is to create a narrow, flume-like 224 channel that supports only small, low-profile fans and, as a result, relatively small fan-eddy complexes (river miles 11-20 and 28-38 of reach two). However, once these lithologies are elevated to positions above river level, their role in controlling flow and sediment transport changes because they contribute large numbers of boulders to the mainstem that accumulate as large, high-profile fans in wider channel reaches. Transport and deposition of these coarse clasts during debris flows supports fans in most of the larger fan-eddy complexes found in reaches three, four-b, and five.

Implications of channel-reach variability on riverine resources

As tributary debris flows continue aggrading alluvial fans along the Colorado River in Grand Canyon, features of fan-eddy complexes will continue to be modified. Rates of fan aggradation will depend on the many basin characteristics identified by Griffiths et al.

(1996) that influence debris-flow frequency and magnitude (Melis et al. 1994), as well as climatic factors. One of the most immediate and obvious impacts of fan aggradation by debris flows is a progressive increase in the severity of rapids (Melis et al. 1994). On the basis of data presented in this study, many fans do not presently support rapids, but more rapids are likely to form as debris flows deposit boulders in the mainstem. In time, this process will result in a less-navigable river with more turbulent whitewater separated by more placid upper pools.

A less obvious, but perhaps more significant impact of aggrading fans will likely be the system-wide alteration of fan-eddy complexes in ways that increase storage of sand and silt along the mainstem. This type of long-term trend could benefit both abiotic, and biotic resources dependent on sand bars for habitat, such as native species that formerly relied on nutrient-rich mainstem silt deposits delivered by tributaries. Riparian plants and terrestrial 225 fauna alike might benefit from increased terrestrial sand bars that result from modified fan- eddy complexes. If progressive aggradation of fans results in increased storage of sand and silt along the mainstem, then controlled floods designed to redeposit sand bars may be prescribed more frequently in the future (see DOI 1995, p. 35-37).

The ongoing creation of rapids also relates to recent studies by Shannon et al.

(1996) who suggest that rapids play a key role in primary aquatic productivity in the mainstem. Rapids may increase productivity by pulverizing river-borne detritus, termed drift, that is added to the mainstem during tributary floods. This "ball mill" action by rapids turns coarse drift into finer particle sizes as the organic matter is transported downstream. Hence, this process apparently results in finer particulate organic matter

(nutrient sources) being released downstream that is more readily available for use by the aquatic ecosystem for primary and secondary productivity (Shannon et al. 1996). Hence, for any given rate of organic input to the mainstem, the existence of additional rapids may be effective in elevating aquatic productivity downstream until the process becomes limited by turbidity. In areas of reaches one, two and the upper part of reach three, above the Little

Colorado River, such a trend may have real benefits for native and non-native fishes, as well as waterfowl.

Additional studies by Melis et al. (1995), suggest that increased channel constriction caused by debris flows may also elevate aquatic benthos when debris-fan ponding occurs upstream from rapids. Increased ponding due to enhanced constriction by boulder deposition enhances sediment storage in the upper pool, and may also increase light penetration as suspended sediment deposits. Such physical effects have been recently documented at river mile 68 (Tanner Rapid) and were recently shown to be relative permanent local-scale changes in the mainstem (Webb et al. 1997; J.P. Shannon, 226

Biosciences Department, Northern Arizona University, pers. commun, 1996). Long-term

resource trends associated with such physical changes system-wide will likely alter this

desert-riparian river ecosystem in ways that were formerly transient in the pre-dam era, but are now relatively permanent under regulated conditions.

Valdez and Ryel (1996) report that the debris-flow dominated geomorphic

framework of the Colorado River in Grand Canyon may be a significant factor in the life

history of native, endemic fishes that evolved there; particularly the humpback chub (Gila

cypha). Based on studies conducted between 1990 through 1993 in Grand Canyon, they

documented distributions, as well as movement, feeding and spawning behaviors that were

interpreted as possibly being related to the closely spaced recirculating zones associated

with large fan-eddy complexes found in the vicinity of the LCR in reach 3, and 4b (Valdez

and Ryel 1996). The LCR is located in the middle of reach 3 and was identified as a key

factor in reproduction of humpback chub. However, Valdez and Ryel (1996) reported that

many fish who left the mainstem to spawn in the LCR showed remarkable fecundity for

specific fan-eddy complexes in the mainstem where they resided during the non-spawning season.

Another important factor recently documented in the life history of sub-adult humpback chub was related to various shoreline types found upstream and downstream of the LCR. Valdez and Ryel (1996) report that sub-adult chub preferentially choose irregular shorelines associated with talus and debris fans over bedrock-cliffs or smooth sand shorelines as habitat. The highest density of young chub were consistently found along vegetated shorelines which may provide higher levels of food and cover from predators

(Valdez and Ryel 1996). Vegetated shorelines are commonly associated with relatively stable terrestrial sand deposits within large fan-eddy complexes. Once mature, adult chub 227 are known to move from shoreline habitats to that of recirculating eddies. Once this shift in habitat use occurs, feeding habits of chub become progressively associated with dynamic recirculating vortices in eddies that effectively trap for drift and other detrital food sources

(Valdez and Ryel 1996).

In conclusion, ongoing biotic and abiotic changes that occur within the Grand

Canyon river ecosystem will continue as a result of introduction of non-native species and variations in sediment flux related to climate variability, but will mainly be effected by river regulation. One of the most fundamental effects of regulation below Glen Canyon Dam since 1963 has been a significant shift toward greater influence by debris flows on the mainstem channel and its resources. This shift in the geomorphic framework of the river system will continue to be expressed by changes in fan-eddy complexes, and these changes will likely alter the sediment balance of the river system (increased sediment storage in association with aggrading alluvial fans). Ongoing physical changes related to altered process interactions throughout the riverine ecosystem of the mainstem can only generally be predicted at present, since linkages between river geomorphology and biological activity are not fully understood. 228

CHAPTER 6

REWORKING OF AGGRADED COLORADO RIVER DEBRIS FANS BY THE 1996

CONTROLLED FLOOD BELOW GLEN CANYON DAM, ARIZONA

Abstract

From 1986 through 1995, debris flows increased debris-fan constrictions along the

Colorado River at the mouths of at least 25 tributary canyons in Grand Canyon, creating 2 new rapids and narrowing at least 9 existing riffles or rapids. The highest peak discharge between 1986 and 1996 was 960 m3/s in January, 1993, and in March-April, 1996, the reworking effects of a 7-day controlled flood with a peak of 1,370 m3/s were studied on 18 of these recently aggraded debris fans downstream from Glen Canyon Dam. Following the controlled flood, plan-view areas of recent debris-fan deposits had decreased by 2 to 42 percent, while volumes had decreased by 3 to 34 percent. The distal margins of most aggraded debris fans became armored during the flood with a lag of cobbles and boulders.

Observed debris-fan reworking during the flood indicates that such erosion occurs by two distinct mechanisms: lateral bank retreat of matrix-supported clasts; and entrainment of non- matrix-supported clasts as bedload, and these combined erosional mechanisms indicate that interactions between debris flows and large rivers are complex. The elapsed time between individual debris flows and the controlled flood was a key factor in the grading of recent deposits because coarser particles at older, partially reworked fans became interlocked, imbricated, and (or) sutured together during smaller hydropower plant releases. The degree to which similar-magnitude floods rework future debris-fan deposits depends on the magnitude of all lesser mainstem flows that inundate deposits prior to controlled flooding. 229

INTRODUCTION

Debris flows from unregulated tributaries between Lees Ferry and Lake Mead (river

miles 0 to 240) create and maintain rapids along the Colorado River in Grand Canyon

National Park, Arizona (fig. 6.1; Graf 1980; Kieffer 1985; Webb et al. 1988; Melis et al.

1994; Webb et al. 1996). Since 1986, the last year in which Glen Canyon Dam released at

least 1,270 m3/s, 25 debris flows have significantly increased debris-fan constrictions

along the Colorado River in Grand Canyon (Meus et al. 1994), creating 2 new rapids and

enlarging at least 9 existing riffles and rapids. The hydraulic settings of these affected sites

ranged from small riffles to Lava Falls Rapid, the largest rapid in Grand Canyon (fig. 6.1;

Webb and Melis 1995; Webb et al. 1996). In general, deposition of poorly sorted debris in

the mainstem controls the basic geomorphic unit characterizing the river channel in Grand

Canyon; the eddy complex (Melis et al. 1994; Schmidt and Rubin 1995). Periodic debris-

flow deposition in the channel increases local flow velocities and water-surface fall through

rapids (Melis et al. 1994; Webb 1996), changes the navigational severity of rapids, and

affects sand storage in low-velocity pools that exist between rapids (fig. 6.2; Melis et al.

1994; Melis et al. 1995).

In the unregulated river, all but the largest particles were swept downstream, and

cobbles and small boulders were redeposited on debris bars that constrained the

downstream extent of eddy complexes (Howard and Dolan 1981; Schmidt 1990; Schmidt

and Graf 1990; Schmidt and Rubin 1995) and controlled secondary rapids and riffles

(Webb et al. 1989; Webb et al. 1996). The residual deposits formed boulder-laden debris

fans that created eddy complexes and rapids (fig. 6.2). Before closure of Glen Canyon

Dam in 1963, the Colorado River removed most new debris-flow deposits that aggraded 230 debris fans during early summer floods that averaged 2,330 m3/s, and were occasionally as large as 6,230 m3/s (fig. 6.3). The interaction between the frequency and magnitude of tributary debris flows and mainstem floods resulted in debris fans and rapids that were relatively stable (Melis and Webb 1993). From 1963 through 1982, operations of Glen

Canyon Dam constrained discharges to an average annual peak of 870 m3/s, which allowed tributary debris-flow deposits to aggrade fans (Howard and Dolan 1979, 1981). The large dam release of 1983 (approximately 2,720 m3/s), had a magnitude similar to pre-dam floods (fig. 6.3) and partially reworked at least one highly aggraded debris fan at the confluence of Crystal Creek (Kieffer 1985).

A program to monitor the reworking of recently aggraded debris fans, which numbered 25 by 1996, was started in 1987 (Table 6.1; fig. 6.1; Melis et al. 1994; Webb et al. 1996), and 18 of these sites were studied in conjunction with the 1996 controlled flood.

Hydropower plant releases of up to 870 m 3/s significantly reworked the distal margins of most recently aggraded debris fans and armored older, partially reworked fans.

Hydropower plant releases combined with a flood in the Little Colorado River (fig. 6.1) in

January, 1993, resulted in boulders as large as 1 m in diameter being entrained at some fans (Melis et al. 1994), and increased armoring along the distal margins of many recent debris-fan deposits. Webb et al. (1996), document the complete removal of two historic debris fans at Lava Falls Rapid by dam releases in excess of hydropower plant capacity; the first in 1965 by a dam release of 1,640 m3/s, and the second in 1973 by a flow of 1,080 m3/s. Cobbles and boulders entrained from post-dam debris fans during those releases appeared to be redeposited in the pool immediately downstream from Lava Falls Rapid instead of on alternating bars farther downstream, as formerly occurred during the unregulated era (fig. 6.2). This altered pattern of grading reflects a fundamental change in 231

FIGURE 6.1. Map of the Colorado River in Grand Canyon showing the locations of debris fans monitored during the controlled flood (from Webb et al. 1997). 232

{•n••••••••• \,1 . ;.: 7: . ‘ \ \ji . • f \ ...4'..:

• \

(Modified from Hamblin and Rigby, 1968)

Explanation

1. Tributary debris fan 2. Rapid controlled by large immobile boulders 3. Debris bar (synonymous with "island" or "rock garden") 4. Riffle or rapid caused by debris bar

FIGURE 6.2. The morphology of a typical debris fan, rapid and eddy complex found along the Colorado River in Grand Canyon, Arizona (from Melis et al. 1994). 233

the geomorphic framework of the Colorado River resulting from mainstem regulation (Webb 1996).

The objective of this study is to quantify and record the effects of the 1996

controlled flood on the reworking of 18 recently aggraded debris fans and associated rapids

in Grand Canyon National Park. Such information is relevant to both regulated and

unregulated rivers that are impacted by tributary debris flows. With respect to debris-fan

reworking, two hypotheses were tested during the experiment: 1) that two types of

reworking occur on aggraded debris fans when poorly sorted debris encounters streamflow

in a large river, with lateral bank collapse being an important erosion mechanism initially on

matrix-supported deposits during rising flood stage, while entrainment of non-matrix

supported clasts is most important on deposits previously subjected to limited reworking;

2) that controlled releases of 1,370 m3/s from Glen Canyon Dam can periodically erode

newly aggraded debris fans to their former conditions. The results of the 1996 controlled

flood experiment demonstrate that sediment-transport interactions between debris flows and

large rivers are more complex than previously described, and provide system-wide

information that may be useful in designing future controlled floods intended as channel

maintenance tools.

Previous and related Work

Contrary to the conclusions of earlier Colorado River researchers (Leopold 1969),

most Colorado River rapids in Grand Canyon result from accumulations of large boulders

at tributary confluences that are deposited during tributary debris flows (Péwé 1968;

Hamblin and Rigby 1968; Simmons and Gaskill 1969; Graf 1979; Howard and Dolan

1981; Webb et al. 1988, 1989; Melis et al. 1994; Melis and Webb 1993). Boulders that 234

Q 0 o o Q o o o Q o o o N- (6 L6 ,71: 6 c.i. (sicw) e6Jetiosu

FIGURE 6.3. Annual flood series for the Colorado River near Grand Canyon, Arizona.

After closure of Glen Canyon Dam in 1963, the mean annual flood decreased from 2,330 to 959 m3/s. The 1996 controlled flood is about a 10-year flood in comparison with other annual maximum dam releases, but is only a 1.3-year flood on the unregulated Colorado

River. The 1996 controlled flood was about half the size of the 1983 dam release, but 20-

50 percent larger than dam releases between 1986 and 1995. 235 exceed the competence of the river's largest floods become stable features that form rapids.

Kieffer (1985) first studied the interactions between debris flows and mainstem fluvial processes on the Colorado River twenty years after closure of Glen Canyon Dam.

She concluded that floods on the unregulated Colorado River eventually widened debris- fan constrictions to an average constricted width of 50 percent by eroding boulders and debris-flow matrix from fans following debris flows. Based on a subset of measurements of channel and fan geometry throughout the mainstem, and using generally known hydraulic principles to study debris fan reworking at Crystal Creek (river mile 98.2-R),

Kieffer (1985) estimated that floods as large as 11,300 m3/s would have been required to produce the present channel/fan morphology seen throughout Grand Canyon at the time of her study.

Recent workers have documented that many rapids in the Colorado River system have become more severe since mainstem regulation because debris-fan constrictions and individual boulders cannot be totally removed by typical dam releases (Graf 1980; Howard and Dolan 1981; Melis et al. 1994; Webb 1996). Péwé (1968) speculated that regulated flows would have minimal effects on debris-fan reworking, and that debris fans would therefore aggrade throughout Grand Canyon. By analyzing 1965 and 1973 aerial photographs, Howard and Dolan (1979) estimated that 25 percent of the debris fans in

Grand Canyon had been impacted by tributary floods. However, they failed to document specific details of the geomorphic-process interactions that occur between tributary sediments deposited in the mainstem and river reworking. Graf (1980) hypothesized that debris fans on the Green River would aggrade or remain stable in response to operations of

Flaming Gorge Dam. Melis et al. (1994) also documented the aggradation of debris fans by debris flows and other floods in Grand Canyon, as well as the reworking of selected 236 fans by dam releases less than the maximum hydropower plant capacity of Glen Canyon Dam.

Using field observations and one-dimensional hydraulic modeling, Hammack

(1994), and Hammack and Wohl (1996), describe the reworking effects of an unregulated

Yampa River flood of 1,200 m3/s in 1984 on Warm Springs Rapid; a prominent geomorphic feature that was aggraded by a large tributary debris flow in 1965. Based on that study, they report that the duration of a flood is as important as its peak discharge in reworking aggraded debris fans.

Webb et al. (1996) documented several Grand Canyon debris flows at Lava Falls

Rapid that created constrictions similar to the 1966 event described by Kieffer (1985) and

Cooley et al. (1979) at Crystal Rapid. Webb et al. (1996) report that those constrictions were reworked by flows less than or equal to 3,530 m3/s. Webb et al. (1996) also suggest that dam releases above powerplant capacity, but less than the annual pre-dam flood, such as the 1996 controlled flood, might partially or completely remove aggraded debris fans, depending on local topography and initial condition of fan deposits.

METHODS

Debris-fan study sites

Working with a team of scientists and technicians from the U.S. Geological

Survey, Bureau of Reclamation and Universities of Minnesota and Delaware, I assessed the amount of reworking by the 1996 controlled flood on 18 recently aggraded debris fans

(Table 6.1). These fans are distributed along the mainstem downstream of Glen Canyon

Dam from Badger/Jackass to Prospect Creeks (river miles 7.9 to 179.4; fig. 6.1) and

•

237

TABLE 6.1. Characteristics of debris fans aggraded after 1986 that were monitored during the 1996 controlled flood in Grand Canyon.

x

X X X X X X X X

X X X X X X X X

5 • <95 10o MA F.'. § XX XXX X XXXX XXX XX XXX

OCCCOCCCDOC:)000=000 N.c ,C) N — ,0 ,0 ,0 00 0, C7, ‘.0 .11 C, 0, CS‘ sc0 ,0)

ur4uurzunuuuuuuuuunu

0, vet N en Oc-) en en V') CA 0, QN en en trl csec a a a a oc co a oc co oo co oo cn cr, C' a a, a a cr, a a a, a, a, a, a, a a, o, o >). — —

c7N v-n c•-• tr) Ch — va co co so 00 06 rsi r o-; o6 6 — N 06 r-: 6 h ci crn — sc) CI,NNN (-4 rn

Q co oQ .a -Ne c ...1e E ° >, co c 0.) ..x ‘4 c c c a., ,... 2 a) COm = CO U g u L) U •-- CO U U ., U 'A, U CO CO. - .-] ,-1 ,-.1 6. C4 C=4 7-1 a.) a.t.) g:4 -a t = g:4 cl , , , 0 , , c..) t) CO COU „ eZ a 0 . 1 z; L-:. s.c. cao, cl. CO CO .. 4 ‘r, f'-' • c.=' > a "2 ri —. , N N a -c N o • a a 75 --,) )).; Ni n-•:. CO CO CO — rq 6. r'4 C`l ("4 C,.' ,r) %I.) I.. U CO CCL> Z o •rc" .4 - ,a ,c a6, E- U 1.--- t--- U - - - (t) CI:1 - - a.., fn 238 represent a variety of geomorphic and hydraulic channel/alluvial fan configurations.

Debris-flow deposits at several sites, particularly river miles 62.5-R, 68.5-L (Tanner

Canyon), 127.6-L, 157.6-R and 179.3-L (Prospect Canyon), initially constricted the river significantly. At 18-Mile Wash, the 1987 debris flow covered both an existing debris fan and separation sandbar and significantly constricted the river. Some of the debris fans, particularly at river miles 65.5-L (Palisades Creek) and 160.8-R, represent debris flows that covered existing debris fans without significantly increasing the constricted width of the river (Table 6.1).

Most of these debris fans had experienced some reworking before the 1996 controlled flood by hydropower plant releases from Glen Canyon Dam that combined with tributary floods in January, 1993, and August, 1994 (Melis et al. 1994; Melis et al. 1995).

Two debris fans — at river miles 71.2-R and 72.1-R — were included despite reworking by discharges up to 1,510 m 3/s in 1986 (fig. 6.3), and were not expected to be eroded by the 1996 controlled flood. The debris fan at Tanner Canyon, was aggraded in August,

1993, and was initially reworked by a hydropower-plant peak of about 510 m3/s that was increased by tributary inflows of about 142 m3/s from Glen Canyon tributaries in August,

1994. The highest mainstem discharge to occur between 1986 and the time of the controlled flood was 960 m3/s, and affected 8 of the 18 debris fans (a combined peak resulting from floods on the Little Colorado River (river mile 61) and other tributaries and dam releases on January 13, 1993). The other ten debris fans were reworked by dam releases that ranged from 620 to 830 m3/s before the controlled flood (Table 6.1). Debris fans at Crystal and Lava Falls Rapids were aggraded in March, 1995, and because of their young age and poor sorting, were expected to erode significantly during the experimental flood. 239

Documenting debris-fan reworking

Photogrammetric analysis of aerial photographs, taken at steady river discharges of

230-260 m3/s before and after the controlled flood, was used to remotely assess changes in

river constriction, surface area, and the width of the reworked zone of the 18 debris fans.

Topographic control for debris fans between river miles 62 and 73 and for Crystal Rapid

(river mile 98.2) were obtained from digital orthophotographs (Glen Canyon

Environmental Studies Program, unpublished data). For all other debris fans, 9 to 12

control points on prominent rocks readily identified on low-level aerial photographs were

surveyed and used for ground control. Using image-processing software, which included

ARC/INFO and the Map and Image Processing System (MIPS), the plan areas of each of

the debris fans, the change in width of the reworked distal edge of the debris fan, and the

channel widths necessary to determine the constriction of the Colorado River were

measured. The reworked zone was observed as a band along the distal edge of a debris fan

where an increase in surface texture indicated an increase in particle size associated with

grading of debris-flow sediments.

The accuracy of image processing varied among the debris fans on the basis of the

amount of distortion in the aerial photographs, the accuracy of surveyed or

orthophotograph controls, the clarity of the image, and the choice of borders between

debris fans, sandbars, and water stage in the Colorado River. The river discharge at each

debris fan was not the same in the before- and after-flood aerial photography and is a

source of error. Although the pixel resolution for the rectified images ranged from 0.1 to

0.5 m, rectification involved the alteration of pixel locations by as many as several meters.

For consistency, all area measurements were rounded to the nearest 100 m2 (implied 240 accuracy of ±50 m2), and all distance measurements to the nearest 1 m (implied accuracy of

±0.5 m).

Ten of the eighteen debris fans included in this study were surveyed before and after the controlled flood to calculate changes in debris-flow deposit area, volume, and water-surface fall through the rapid/riffle. Recently aggraded fan sediment volumes were calculated above the minimum water-surface elevation measured at the most downstream point along the distal edge of the fan deposit during the pre-flood constant flow

(approximately 230 to 260 m3/s). The accuracy of surveyed ground points is on the order

of ±0.1 m horizontally and ±0.05 m vertically. Areas and volumes of debris-flow deposits

were calculated using the volume module within Sokkia Inc. MAP software. Although the

accuracy of surveying data is higher than image-processing data, for consistency all area

measurements were rounded to the nearest 100m2 (implied accuracy of ±50 m2), all

volume measurements to the nearest 100 m 3 (implied accuracy of ±50 m3), and all distance

measurements to the nearest 1 m (implied accuracy of ±0.5 m). The change in river-channel constriction at a debris fan caused by debris-flow

deposit reworking during the controlled flood was calculated as the percentage of channel

width constricted by the debris fan (Webb et al. 1996). The data for this technique were

obtained using image processing and verified using survey data. Representative channel

widths were measured upstream from the rapid and downstream from the pool below the

rapid (fig. 6.2), as well as 4-5 channel widths of the most constricted section adjacent to

the reworked distal margin of the debris deposit. The percent channel constriction, Cw,

was defined as 241

2W r(ave) c,, 1 X 100, (6.1) (W u + W d )

where Wr(ave) = the average constricted channel width, W, = the width upstream from the

rapid, and Wd = the width downstream from the rapid below the expansion zone.

Although C w varies with river stage, I report C w at discharges before and after the

controlled flood that varied between 230 and 260 m3/s. The difference in discharge before

and after the flood at each site was not enough to significantly affect the accuracy of the

constriction values, which was considered to be ±0.5 percent for C.

Unit stream power, co, was also calculated before and after the flood at each rapid

as

y Q S —

r(ave) (6.2)

where y = specific weight of water, Q = discharge, and S = water-surface slope through the rapid (Baker and Costa 1987). A value of y= 9,810 N/m 3 was used in all stream power calculations.

A variety of techniques were used to measure reworking on 10 of the 18 debris fans monitored before and after the controlled flood at discharges of about 250 m3/s (Table 6.1).

Point counts (Wolman 1954; Rice and Church 1996) made on the distal margins of debris fans were used to document changes in particle-size distribution and armoring (Meus et al.

1994). Surveyed water-surface profiles documented changes in the water-surface fall between pools upstream of debris fans and the start of runs below rapids/riffles; which is

242 dependent on both bed elevation and the constricted channel width. Overall changes in debris-flow deposit morphology were also recorded using oblique still photography at established monitoring stations.

On 7 of the 18 debris deposits (Table 6.1), selected boulders were marked with numbered bolts and their locations surveyed during low discharges between 1994 and

March, 1996. The three principal axes, as well as appropriate dimensions for calculation of volume, and lithology of each particle were recorded. Particle volume and weight were then estimated using appropriate mensuration formulae (Melis et al. 1994), and a particle density of 2,650 kg/m3 for limestone and sandstone and 2,700 kg/m3 for basalt. During the low constant flow that followed the recession of the controlled flood, the positions of marked boulders that remained above water level were resurveyed. At Palisades Creek

(river mile 65.5-L), 190 marked clasts ranging in size from small cobbles to boulders (b-

axis of 128-256 mm) were measured, placed, and surveyed along the coarsened, reworked

zone of the debris fan between the 230 and 1,270 m3/s discharge levels. Following the

flood, the reworked area was resurveyed to determine which particle sizes and lithologies

were removed from the reworked surface by the controlled flood. These data were

collected as a means of documenting debris-fan reworking solely by the mechanism of

bedload entrainment of non-matrix supported clasts smaller than boulders.

Surface-flow velocities were also measured before and after the controlled flood in

selected riffles and rapids (Table 6.1). For most of these measurements, tetherballs were

released from shore or from boats stationed at the top of rapids and their travel times

recorded along flowpaths down the center or along the sides of rapids. The configuration

of the tetherball float did not affect the velocity measurement; the velocities obtained with partially deflated balls, fully inflated balls, and balls with attached carabinier clips were 243 statistically identical. Velocity measurements were only made at times when wind velocity was minimal. The length of the velocity test was measured either with tapes stretched along the shore (an accuracy of ±1 m over a distance of 40-70m), or by survey (distances of 50-270 m) and timing was to the nearest 0.1 s. Therefore, the estimated accuracy of velocity measurements under ideal conditions was determined to be on the order of 0.1 mis. In some rapids, the velocity of the tetherballs was altered (or decreased to zero) by certain wave configurations, resulting in rejection of the measurement.

During the seven days of the controlled-flood peak, debris-fan reworking was measured at Palisades Creek, Tanner Canyon and Prospect Canyon/Lava Falls Rapid, river miles 65.5-L, 68.5-L and 179.3-L, respectively. At these selected sites, water-surface profiles were surveyed periodically to determine the rate of change during the rise of the controlled flood. Surface velocities were also measured occasionally during the rising limb of the hydrograph at Lava Falls Rapid and during the steady high discharge at all three rapids. At Tanner Rapid, driftwood was timed for streamflow velocity measurements during the peak of the controlled flood.

Before the flood at Lava Falls Rapid, radio transmitters (tags) were cemented into 9 selected particles on the distal fan margin that would be inundated by discharges between

250-280 m3/s.; a radio tag was attached to the outside of a 10th particle. Three of the 10 particles were cobble size (less than 256 mm); the rest were boulders (greater than 256 mm). Specially built radio tags transmitting in the 40-MHz waveband with a battery life of

30 days for maximum transmission power (Lee Carstenson, Smith-Root Incorporated, personal commun. 1996) were monitored to detect clast movements during the flood peak.

The time of initial motion during the flood was documented using both a scanning receiver attached to a data logger and an audible receiver. Radio signals from dislodged particles 244 were faint during the controlled flood because of signal attenuation in the deep water.

Weak to strong signals were detected during the steady low discharges after the flood. By triangulation with manual receivers, 8 of 10 radio-tagged particles were located after the flood's recession.

RESULTS

Hydrographs of the 1996 controlled flood

The 1996 controlled flood was designed to consist of 5 release periods from Glen

Canyon Dam: an initial low, steady flow of 226 m3/s for 4 days; a rapid rise at a rate of 113

(m3/s)/hr; a steady flow of 1,270 m 3/s for 7 days; a 3-stage downramping period (of no relevance to this study); and a concluding low, steady flow of 226 m3/s for 4 days (Glen

Canyon Environmental Studies Program, written commun. 1996). The hydrograph and discharges measured at 3 gaging stations below the dam differed only slightly from the intended design characteristics of the controlled flood. The initial low, steady flow period had discharge ranges of 231-238, 239-250, and 252-258 m 3/s at Lees Ferry, Grand

Canyon, and Diamond Creek, respectively. The ranges in peak discharges were 1,270-

1,300, 1,270-1,370, and 1,300-1,350 m3/s at these stations, respectively. Finally, the concluding low, steady flow period had discharge ranges of 229-234, 244-251, and 242-

253 m3/s at Lees Ferry, Grand Canyon, and Diamond Creek, respectively. The differences between the gaging stations are probably related to errors in the stage-rating curves of the gaging stations and to contributions of tributary inflows.

The discharge hydrograph for January-September, 1996, is considerably different from typical dam releases, recent dam releases with tributary inflows, and a comparable 245 pre-dam period (fig. 6.4). A 10-year flood on the unregulated Colorado River was approximately 3,390 m 3/s and had a considerably longer flow duration and higher flow volume (fig. 6.4a). The largest peak discharge on the mainstem since 1986 (966 m3/s) occurred in January, 1993 (fig. 6.4b), which is a 3-year flood in the post-dam period, but less than the mean annual peak flow (approximately 2,475 m3/s) before closure of Glen

Canyon Dam. In contrast, the peak discharge of 1,370 m3/s at the Colorado River near

Grand Canyon gaging station during the 1996 controlled flood was approximately a 10-

year flood during the regulated period, but only a 1.3-year flood relative to pre-dam annual

peaks (fig. 6.4c).

Relation between the hydrogra_ph and reworking at Lava Falls Rapid

In March, 1995, a debris flow from Prospect Canyon (river mile 179.3-L), the

sixth to occur since 1939 (Webb et al. 1996), constricted the Colorado River at Lava Falls

Rapid (fig. 6.1; Webb and Melis 1995). After a period of initial reworking in the days

immediately following this new debris-flow deposit, the aggraded debris fan had stabilized

to an area of 5,300 m2 and a volume of 32,900 m3 . During the 1996 controlled flood, the

partially reworked 1995 debris-flow fan underwent another significant episode of erosion

(fig. 6.5).

Fan reworking was observed at Lava Falls Rapid from the time of the initial rise of

the controlled flood until several days into the peak. Data were collected on the relation

between the period of active erosion and the hydrograph (fig. 6.6). As the discharge rose

on the morning of March 27th, the stage increased accordingly until a discharge of about

1,000 m3/s (about 11:30 hrs). The stage was approximately 0.2 to 0.5 m less than the top

of the aggraded debris fan, and lateral erosion of the 1995 deposit through bank collapse 246 f 3/22/96 4/6/96

0 50 100 150 200

5 I F 1 I - I I- 1 3/24/96 - 4 ------4/7/96 - -

3 - - _ ----- _ - , --- - 2 - - ,- - - 1 - - : B. Tanner Rapid 0 100 200 300 400 500 600 5 - 3/25/96 - 4 - 4/9/96

3 - -

2 - -

- 1 - C. Lava Falls Rapid 0 100 200 300 400 500 600 Thalweg Distance (m)

FIGURE 6.4. Hydrographs of the Colorado River near Grand Canyon showing the

January-September hydrographs of a pre-dam flood in 1957, dam releases augmented by a

10-year flood on the Little Colorado River in 1993, and the 1996 controlled flood. The peak discharges of the 1957 and the 1996 floods are 10-year events in the pre-dam and post-dam periods, respectively. 247

FIGURE 6.5. Aerial photographs that document reworking of the Prospect debris fan at

Lava Falls Rapid. Left. Lava Falls Rapid was constricted by a 1995 debris flow from

Prospect Canyon. Right. Reworking during the rising limb of the 1996 controlled flood removed 5,900 m3 of the edge of the debris fan, increasing the width of the rapid by an average of 6 m. 248

1,400 ca. 1300: Rate of Bank Collapse Decreases . 1215: Cutbank Collapses - 1,200 Continuously

- 1,000 1140: Sediment 1400: Sound of - Calving Begins N Rolling Boulders - Decreases - 800

V,

- 600 Discharge at Diamond Creek Stage at - 400 Lava Falls

1(11'111111 0 0 200 00 o 0 o 0 O o o o o o o (0 co o c., •1- co co O 0 — — r — —

Time on March 27, 1996

FIGURE 6.6. The relation between stage, discharge, and reworking of the debris fan at Lava Falls Rapid. The stage record at Lava Falls Rapid is for March 27, 1996, and the hydrograph for the Colorado River above Diamond Creek is adjusted for a discharge- independent travel time of 7.46 hrs. 249

(slab failures) was observed as early as 11:20 hrs. Bank collapse began on the downstream side of the debris fan. By 12:00, the cutbank of the aggraded debris fan was continuously collapsing, and the stage record flattened to nearly zero change with time, despite an increasing discharge (fig. 6.6). For about an hour, boulders were heard rolling along the bed of the river away from the debris fan. Bank collapse greatly decreased in frequency by 13:00 hrs, and boulders were heard rolling only occasionally by 14:00 hrs.

As reworking slowed, the stage began rising again, tracking the slow rise in discharge (fig. 6.6).

Areas and volumes of debris-flow deposits

The planview areas of recently aggraded debris fans in Grand Canyon generally decreased during the experimental controlled flood (fig. 6.7; Table 6.2). The largest changes were in the debris-fan deposits at Jackass Canyon (Badger Creek Rapid) and

Prospect Canyon (fig. 6.5), river miles 7.9-L and 179.3-L, respectively. At both locations, a debris flow had occurred less than two years before the controlled flood

(August, 1994 and March, 1995), and previous hydropower plant discharges (226 to 566 m3/s), were too low to significantly rework these deposits. Relative decreases in deposit area were observed at most of the aggraded debris fans studied, but the area of the 1989 fan deposit at Bedrock Canyon (river mile 130..5-R) increased slightly because reworked sediment was deposited at the downstream margin of the debris fan. As expected, the 1983 debris-flow deposit at river mile 71.2-R, which had been reworked at higher discharges between 1982 and 1987 (Table 6.1), remained unchanged. However, despite the fact that another 1983 fan deposit at river mile 72.1-R had also been partially reworked from 1983 through 1986, it lost 100 m2 of fan area during the controlled flood. These contrasting 250

FIGURE 6.7. Replicated photographs that document controlled-flood reworking of the

1989 debris fan at 127.6-L. Left. Aggraded condition of fan before controlled flood.

Right. Coarsened condition of fan following reworking by controlled flood. 251

TABLE 6.2. Changes in areas of debris-fan deposits reworked by the 1996 controlled flood in Grand Canyon.

IMAGE PROCESSING SURVEYED Pre-flood Post-flood Difference Pre-flood Post-flood Difference Study site area (m2) area (m2) (%) area (m2) area (m) (%)

Jackass Cyn." 1,900 1,100 -42 1,900 1,100 -42 18-Mile Wash 4,300 4,300 0 4,200 4,000 -5 62.5-R 1,600 1,500 -6 ns ns ils 63.3-Rt 1,400 1,300 -7 ils ns ns Palisades Crk.t 2,300 2,300 0 ils ns ns Tanner Cyn.# 2,100 2,100 0 2,000 1,900 -5 Cardenas Crk.# 7,800 7,900 1 ns ils ns 71.2-R§ 4,400 4,400 0 ils ils ils 72.1-R§ 1,500 1,400 -7 ns ns ns Crystal Crk.'* 5,300 5,200 -2 ns ns ns 126.9-L 3,100 3,100 0 3,200 3,100 -3 127.3-L 2,000 1,900 -5 2,100 2,100 0 127.6-L 3,300 2,800 -15 3,000 2,700 -10 Specter Chasm 3,800 3,400 -10 ils ils ils Bedrock Cyn. 3,000 3,200 +7 ris ns ns 157.6-R 3,400 3,300 -3 3,600 3,100 -14 160.8-R 4,100 4,000 -2 4,200 4,200 0 Prospect Cyn. - 5,000 4,000 -20 5,300 4,200 -21

Except as noted, all areas are of the entire debris fan. ns, not surveyed.*Only the area of 1994 deposition (Melis et et al. al 1.994). tOnly the area of 1990 deposition (MeHs et al. 1994). 00nly the area of 1993 deposition (Melis 1994). §This debris fan was not expected to change during the controlled flood (see Methods section). "Only the area of 1995 deposition (Webb et al. 1996). 252

results are difficult to explain in terms of river reworking, but could be related to measurement error and the fact that sandbar deposition near the sites made it difficult to duplicate pre- and post-flood measurements from aerial photographs.

Considerable agreement exists between the two methods used to measure debris-fan area changes in this study. The areas calculated using image processing differed only slightly from those measured by topographic survey for some debris fans (Table 6.2).

For example, at 18-Mile Wash, no changes were observed in the size of that 1987 aggraded

debris fan using image processing, but a decrease of 200 m2 in area was measured by

surveying before and after the experiment. The quality of the aerial photographs taken before and after the controlled flood, as well as delineation of the fan-water boundary, are

extremely important in the area calculations using image processing. Also, delineation of

the boundary between the new debris-flow deposit and pre-existing debris fan affects the

results obtained from both methods.

In general, changes in the volumes of recently debris-fan deposits mirrored

changes in planview areas (Table 6.3). The 1995 debris-fan deposit at Lava Falls Rapid

(river mile 179.3-L) lost the largest volume of sediment (5,900 m3), whereas the 1994

deposits on the Jackass Canyon debris fan (river mile 7.9-L) had the largest percentage

change (-34 percent). In most cases, the changes in area and volume accurately depict

flood-reworked conditions on the debris deposits except for the 1993 debris-flow deposit at

river mile 160.8-R, where a considerable volume of fine sediment, mostly debris-flow

matrix of particles less than 16 mm in diameter, was eroded from between boulders that

were present before deposition occurred. This type of erosion cannot readily be accounted

for using surveying or image-processing techniques. There were some small

inconsistencies between the changes in area and volume of debris-fan deposits. For 253

TABLE 6.3. Changes in volumes of debris fans reworked by the 1996 controlled flood in Grand Canyon.

Pre-flood Post-flood Volume Difference Debris fan volume (m3 ) volume (m3) eroded (m3 ) (%)

Jackass Canyon' 5,800 3,800 2,000 -34 18-Mile Wash 16,800 14,900 1,900 -11 Tanner Canyont 7,200 6,700 500 -7 126.9-L 15,700 15,700 0 0 127.3-L 8,800 8,500 300 -3 127.6-L 13,100 11,500 1,600 -12 157.6-R 9,800 7,900 1,900 -19 160.8-R 13,500 13,500 0 0 Prospect Canyon§ 32,900 27,000 5,900 -18

Except as noted, all volumes are of the entire debris fan. 'Only the volumes of 1994 deposition (Melis and others, 1994). tOnly the volume of 1993 deposition (Melis and others, 1994). §Only the volume of 1995 deposition (Webb and others, 1996). 254 example, on the 1989 deposit at river mile 126.9-L the survey data show no change in volume yet indicated a 200 m2 decrease in area (3 percent). Because debris-flow deposits are poorly sorted and contain large numbers of boulders with irregular boundaries it is often difficult to describe them accurately using standard field techniques described by

Wolman (1954).

Armoring of the distal margins of debris-flow deposits

Armoring of the distal margin of recent fan deposits was evaluated at 9 of the 18 debris fans studied (Table 6.4). Most debris fans showed significant increases in the sizes of particles in the reworked zones (figs. 6.7, and 6.8). The D50 of particles decreased, remained unchanged, or increased on some debris fans, such as the debris fans at 18-Mile

Wash, and river miles 126.9-L, 157.6-R and 160.8-R where river sand was deposited over

the reworked zone. However, D50 increased by up to a factor of 2.5 at other fans,

including Jackass Canyon and Prospect Canyon (river miles 7.9-L and 179.3-L). The

increases in D85 were generally greater; the typical D85 after the controlled flood was about

0.5 m, but was as high as 1.2 m at Prospect Canyon (Table 6.4).

Along the left side of Lava Falls Rapid, D50 in the area of the 1995 fan deposit

increased from 0.21 m to 0.53 m during the controlled flood (Table 6.4). As a result of

grading, the post-flood clasts on the distal margin of the debris fan were larger and better

sorted (fig. 6.9a). Reworking also preferentially removed limestones, leaving a deposit

dominated by basalt particles (fig. 6.9b), probably owing to differences in density and

shape factors. These results are in agreement with Melis and Webb (1993), who report

similar preferential lithologic sorting during reworking at most pre-dam coarsened debris

fans throughout Grand Canyon. 255

Particle Size, in mm Particle Size, in mm

CV CO cs Cs CV CI Cs 100 100

Initial Debris Flow After Controlled . Flood - o . • • April, 1996 ' • , • 80 80 Badger '- 18 Mile IL \ - - Tanner 126.9 Mile c) 60 - • 60 • • 127.3 Mile 2 - - 127.6 Mile - - 160.8 Mile —o Lava Falls Badger 40 18 Mile 40 - - - - Tanner 126.9 Mile - 127.3 Mile 20 - - - 127.6 Mile 20 - • • - 160.8 Mile —o• Lava Falls

0 o

100 100 Fully-Reworked Fans Before Controlled Flood - D. Near Study Sites March, 1996 80 (pre-darn reworking) 80

60 60 Vo\i — 7.9-R 19.0-R 40 40 — Badger 18 Mile \ 126.9-R - - -Tanner - 166.4-L 126.9 --\\ Mile - 176.4-R • • • • • 127.3 Mile 20 20 -127.6 Mile - • - 160.8 Mile —o - Lava Falls

0 -g -7 -5 -3 -11 -9 -7 -5 -3 Particle Size, in 4) units Particle Size, in 4) units

FIGURE 6.8. Particle-size distributions reflecting the variability of pre- versus post-flood

reworking on 8 recently-aggraded and pre-dam reworked debris fans in Grand Canyon. A.

Initial particle-size distributions of recent debris-flow deposits. B. Intermediate particle-

size distributions reflecting pre-controlled flood river reworking. C. Intermediate particle-

size distributions reflecting post-controlled flood river reworking. D. Particle-size

distributions for debris fans nearby controlled-flood study sites where debris flows have

not occurred in over a century. These debris fans were reworked by high magnitude, pre-

dam floods and approach the endpoint minoring of stable debris fans. 256 (D Particle Size, in mm O) o co Lo 71- (0 100 c ) (.0 — ,a- 1-

1995 Debris flow 1995 Reworked — — - 1996 Reworked - Fully reworked

o -12 -10 -8 -6 -4 2 0 Particle Size, in (1) units

100

li] 1995 Deposit 80 N=875 El 1996 Reworked N=100 111 Fully Reworked 60 N=100

40

20

0 C o C ... C D

FIGURE 6.9. Effects of the 1996 controlled flood on particles on the distal margin of the

Prospect Canyon debris fan at Lava Falls Rapid (modified from Webb et al. 1996), which was reworked by the 1983 flood. A. Particle-size distributions of the distal margins of the debris fan. The "fully reworked" curve is for the debris fan before the 1995 debris flow

(Webb et al. 1996). B. Lithologies of particles at various stages of reworking. 257

TABLE 6.4. Changes in particle-size distributions on the distal margins of debris fans reworked by the 1996 controlled flood in Grand Canyon.

Date of debris Particle Debris-flow Before flood After flood Debris fan flow (yr) diameter deposit (mm) (mm) (mm)

Jackass Canyon 1994 Dso 110 140 230 D85 480 470 540

18-Mile Wash 1987 Dso 11 180 170 D85 18 410 440

Tanner Canyon 1993 Dso 71 89 160 D85 290 550 540

126.9-L 1989 Dso 31 81 82 D85 110 290 310

127.3-L 1989 Dso 54' 54 89 D85 170* 170 240

127.6-L 1989 D50 10 170 260 D85 86 440 620

157.6-R 1993 Dso 120 180 180 D85 360 470 410

160.8-R 1993 Dso 150 140 190 D85 480 410 530

Prospect Canyon 1995 D50 140 210 530 D85 370 650 1170

Particle size measured in March, 1994 because a 1993 tributary flood altered the particle-size distribution of the distal margin of the debris fan. 258

Constrictions, flow velocities, water-surface falls, and stream powers

Of the 18 debris fans monitored as part of the controlled flood, the percent constricted channel width, Cw , decreased at 12 sites (Table 6.5), indicating a widening of the rapid and (or) changes in the stage-discharge relation. The greatest widening occurred at Prospect Canyon (Lava Falls Rapids) and at river mile 62.5-R, where changes in constricted channel width were greater than 10 percent of the pre-flood value. The 1989 debris-fan constriction at Bedrock Canyon was narrower after the flood because eroded sediment was redeposited on the downstream side of the debris fan across from the mid- channel bedrock island (see Stevens 1990). Surprisingly, the constriction widened at river mile 71.2-R despite its history of reworking by larger dam releases between 1982 and

1987. This widening may reflect decreases in sand-bar area mantling both shores of the river instead of real change in the debris fan.

The mean surface-flow velocity did not change consistently at the six rapids where velocities were measured (Table 6.6). At Lava Falls Rapid, flow velocities along the left and right shorelines were decreased by about half after the flood receded. This velocity decrease is consistent with the sizeable decrease in channel constriction at this site (Table

6.5). At Badger Creek Rapid, 18-Mile Wash, and Lava Canyon Rapid, on the other hand, surface-flow velocities increased after the flood (Table 6.6). These velocity increases may have resulted from decreased bed roughness in the rapid associated with deposition of cobbles and small boulders among larger particles on the bed of the river, since the water- surface fall remained unchanged or decreased slightly in these rapids.

The water-surface fall through six of ten rapids was not altered by the controlled flood (Table 6.7). The largest changes were in Tanner and Lava Falls Rapids, where the water-surface fall decreased by 0.2 m and increased by 0.3 m, respectively. The locations 259

TABLE 6.5. Changes in constriction ratios and percent constrictions of debris fans reworked by the 1996 controlled flood in Grand Canyon.

Pre-flood Post-flood Change Debris fan C, (%) Cw (%) (%)

Jackass Canyon 22 20 -10 18-Mile Wash 54 50 -7 62.5-R 55 45 -18 63.3-R 70 69 -1 Palisades Creek 55 56 2 Tanner Canyon 31 33 6 Cardenas Creek 56 55 -2 71.2-R 13 12 -8 72.1-R nd nd nd Crystal Creek 55 52 -5 126.9-L 54 53 -2 127.3-L 24 24 O 127.6-L 47 42 -11 Specter Chasm 32 27 -16 Bedrock Canyon 64 66 3 157.6-R 33 35 -6 160.8-R 47 48 2 Prospect Canyon 42 34 -19

nd, no data. 260

TABLE 6.6. Velocities measured in rapids and riffles before, during, and after the

controlled flood of March-April, 1996, in Grand Canyon.

CrN 0 71' 1 en 00 n0 1.0 ken en 7- 71- et 71- 7t N N NNNN

C CC o o o oc:r, on r- r- r- rn c.,1 ei.. N CV CA

oo o0 %.0 E 71- 71- 't kr) v") NNN NNNN

tn --. --. —. .--. N cr) 6 6 6 6 6 6, 6 -Id -F1 -I-1 E +! +I -id = -Id C en cz c en 00 N = Cn 4 C'; C'.1 G: esi '—' es'

rn N 'at N ^-4 6 6 c; 6 6 6 +1 -1-n +1 +I -id oo vL, C tri 4 4 LA

— cq — — cv rn c-.4 6 6 6 6 6 CEC +1 5 1 +1 +I +I c +1 c +1n tr) t--- Z, cn C — .D

t-,-; c.,t t-•-; 4 ri r,-; 4

t_ r c1) 4r 4r 4r 4r G Q G G 01.) G IL) II) • '-' o! [41 U U

CO

"C1 tu CO 261 of changes in water-surface profile varied among the rapids and debris fans (fig. 6.10). At

Tanner Rapid, most of the change occurred at the top of the rapid, whereas at Lava Falls

Rapid, the largest changes occurred at the bottom of the rapid. The change at Lava Falls

Rapid is probably related to the partial removal of a small, submerged debris bar that had deposited downstream from the rapid during initial fan reworking immediately after the

1995 debris flow (fig. 6.10; Webb and Melis 1995; Webb et al. 1996).

Changes in the percent constricted channel width reflect complex interactions between velocity, the stage-discharge relation, the width of the rapid, and changes in bed roughness. Stream power is one measure that reflects all of these variables. Unit stream power decreased at 7 of 10 debris fans by 4 to 8 percent (Table 6.7). The three exceptions are the 1989 debris-flow deposits at river miles 126.9-L, 127.3-L, and 127.6-L, where slight increases were calculated. The largest increase (4 percent) was at river mile 127.6-L, where the water-surface fall increased by 0.1 m as a result of the controlled flood (Table

6.7).

Transport of marked particles

The 1996 controlled flood removed large boulders from some debris fans, but did not noticeably shift the larger clasts on other fans. The number of marked cobbles and boulders that were transported downstream varied among 5 debris fans aggraded since

1989 (fig. 6.11). Only 4% of the marked clasts were entrained by the controlled flood at the 1993 Tanner Canyon debris fan, located in wide channel setting (Table 6.8; fig. 6.11b), while only 11% were moved at the 18-Mile Wash fan (1987), set in a narrow channel setting (fig. 6.11a). Fan size and the original position of the particles across the fan area were factors that also affected whether or not the particles were relocated once moved by 262 3/22/96 4/6/96

A. 18 Mile riffle

50 100 150 200

5' , 3/24/96 4 4/7/96

3 H

2 -

B. Tanner Rapid

100 200 300 400 500 600

3/25/96 -21

100 200 300 400 500 600

Thalweg Distance (m)

FIGURE 6.10. Water-surface profiles through selected rapids showing the effects of the

controlled flood of 1996. All water-surface profiles were surveyed at discharges between

230 and 248 m3/s. A. 18 Mile riffle, left side. B. Tanner Rapid, right side. C. Lava Falls

Rapid, right side. 263

TABLE 6.7. Changes in water-surface profiles and streampower around selected debris fans at a discharge of 230-254 m 3/s before and after the 1996 controlled flood in Grand Canyon.

a•-n U p L'• C.) • r7 °? ; 7+1-

00000000 • o r-- 00 h Crn C7rn N CO kn a N °

0000000000 ,C) \ CD 00 (0 kn oo oo oo m ‘,0 cr, N

oo oo ‘‘) • cn rn M Cn 'Tr

Cr, 00 V S tr) co CT, nC, 0 Lnentnommcv enrnm-

LL) — cr, r- o co r 00 r- d E 6666666666 0 — ce" fjj a — 6 6 6 6 6 6 6 6 6 c:> a. c.n 0

Lu v.) P ON on CO ,0 0 00 h N h — 8 6666666666 • = 6 6 c:; 6 6 6 6 6 6 6 —

Ln v.) 0 0 0 o tr) E m r- v-) h 00 z .-]

00 00 CC rn 6 6 6 6 6 6 6 6 6 6

71- 6 o — o —; 6 6

71- rn •er oo rn 6 6 —;

c ..,4 oa O 1.) 0a >, >, ..c a) c ,n çt.....) >,, e c ca a co a (...) U ,r) r,..., ) ...1 cn c.) a.) -./ ...] C.4 CL T.): cn 10 es upes aia CiL en ,6 ,r1, 0::, E. -6 c.:, 71.-- me vi) t--: r-: r.:. 6 cz N N N .11 ,..0 :c..1 C.. 264

Median Particle Diameter, in mm Median Particle Diameter, in mm CO 'el' N CO CO CO 0 N N (71- CO Cy) 16

- B. Tanner 12 - Canyon

8

4 0 ..- 1r711 I 111 I I .1.--1 I 40

30

20

10

0

40 Median Particle Diameter, in 4> Units

30

20

10

0 -11 -10 -9 -8 -7 -6 -5 Median Particle Diameter, in (I) Units

FIGURE 6.11. Transport of marked cobbles and boulders from five aggraded debris fans reworked by the 1996 controlled flood. 265

TABLE 6.8. Characteristics of marked boulders at five debris fans, and radio-tagged boulders at Lava Falls Rapid, transported during the 1996 controlled flood.

MARKED PARTICLE MOVEMENTS AT SIX DEBRIS FANS

18-Mile Palisades Tanner Wash Creek* Canyon 127.6-L 157.6-R 160.8-R

Length of fan (m) 110 85 55 86 96 140

Total Number of 27 190 24 83 50 50 marked particles

Percentage of 11% 94% 4% 45% 76% 40% particles removed from debris fan

TAGGED PARTICLES AT LAVA FALLS RAPID

B-axis Depar- Approx. Travel Particle diameter Volume Weight ture discharge distance number Lithology Type (m) (m3) (Mg) time (m3/s) (m)

3 Basalt Cobble 0.21 0.003 0.009 11:41 830 420 1 Sandstone Cobble 0.22 0.006 0.017 11:02 470 na 4 Sandstone Boulder 0.34 0.007 0.018 11:43 860 230 7 Basalt Boulder 0.28 0.011 0.029 12:02 940 310 2 Sandstone Cobble 0.23 0.015 0.040 11:12 560 110 6 Sandstone Boulder 0.31 0.021 0.055 12:14 1,020 50 8 Sandstone Boulder 0.38 0.061 0.16 11:45 870 250 5 Basalt Boulder 0.49 0.065 0.18 11:36 780 na 10 Sandstone Boulder 0.70 0.17 0.45 12:12 1,010 240 9 Sandstone Boulder 0.66 0.23 0.61 12:07 1,000 220

*, experimental plot where cobbles and small boulders were set out on a coarsened debris-fan surface. 266 the flood (Table 6.8). Greater numbers of coarse particles were entrained during the controlled flood at more recent fan deposits set in narrow channel settings, such as the

1989 fan at river mile 127.6-L, or at relatively unreworked fan deposits where debris-flow matrix remained a major component of the deposit supporting cobbles and boulders, such as the fan deposits at river miles 157.6-R and 160.8-R (Table 6.8). At river mile 127.6-L,

45 percent of the marked boulders were removed from the debris fan. At river miles

157.6-R and 160.8-R, 76 and 40 percent of the marked particles, respectively, were transported downstream; boulders with b-axis diameters up to 0.5 m were removed (Table

6.8; fig. 6.11). In contrast, only 11 percent of the marked particles, with b-axis diameters

up to only 0.26 m, were removed from the 18-Mile Wash debris fan from 1987. Based on

the resurveyed positions of marked particles after the flood, average travel distances ranged

from 0.1 to 32.3 m for cobbles, and 0.1 to 8.2 m for boulders.

At the experimental particle-entrainment plot established at Palisades Creek (river

mile 65.5-L), 94 percent of the marked cobbles placed on the coarsened, distal fan edge

before the flood could not be found after the flood receded (Table 6.8; fig. 6.12). Small

boulders were moved within the plot, but most were located within the reworked area after

the flood.

At Lava Falls Rapid, the 10 radio-tagged particles were dislodged at discharges

from 470 to 1,020 m3/s (Table 6.8) on the rising limb of the flood hydrograph (fig. 6.6).

Eight of the 10 particles were relocated downstream from Lava Falls Rapid after the flood.

The smallest particle, a cobble, travelled 420 m to another debris bar that forms the

secondary rapid (Lower Lava Falls Rapid). The other seven relocated particles were

deposited in the pool immediately downstream from the main rapid. The average travel

distance for the 8 relocated particles was 230 m. The discharge required to actively entrain 267

B-axis particle diameter (mm) c co 'uT) cJ c\J

C\1 C\J CO 120

100

80

60 • Pre-Flood 40 Post-Flood _

20

0 -8.0 -7.0 -6.0 -5.0 Particle Diameter (0)

coarsened FIGURE 6.12. Entrainment of cobbles and small boulders placed on the surface of the Palisades Creek debris fan. 268 particles has generally been related to the weight of the particle; and the timing of entrainment is related to the specific flow conditions around the particle and its topographic pocket (Wiberg and Smith 1987). However, the incipient motion of these transported clasts resulted from gravitational force down into the river channel as lateral bank collapse occurred during the rising limb of the flood. The timing of initial motions of these large particles demonstrates the importance of lateral bank retreat along the distal edges of matrix- supported deposits during abrupt rises in river stage. This mechanism of fan erosion likely facilitated removal of even large boulders from new debris fans during river floods that occurred in the pre-dam era.

Another important factor in large particle transport was the initial position of particles on the debris fan relative to the inundation stage. For some debris fans, such as at

Tanner Rapid, many large particles delivered to the mainstem by the 1993 Tanner Canyon debris flow remained above the stage of the controlled flood. Particle position, combined with the previous reworking of the debris fan by lower discharges in August, 1994, and the fact that the 1993 deposit was set in a wide channel setting, resulted in only one marked boulder being removed by the controlled flood (fig. 6.11b). In contrast, despite the fact that the 18-Mile Wash debris-flow deposit was 90 percent inundated during the controlled flood, very few marked cobbles and small boulders were removed (fig. 6.11a), because the debris deposit had been previously reworked to a limited extent by powerplant releases

(Melis et al. 1994). 269

DISCUSSION AND CONCLUSIONS

The 1996 controlled flood partially reworked many of the debris fans aggraded since 1983 that were monitored in Grand Canyon National Park. Using several measures of debris-fan reworking, including changes in fan area, fan volume, amount of river constriction, and sediment armoring, the controlled flood resulted in substantial changes to tributary debris-flow deposits that impacted the Colorado River below Glen Canyon Dam between 1986 and 1996. The peak discharge of the controlled flood, 1,370 m3/s, was a

10-year recurrence interval flood peak in the regulated Colorado River, but was smaller than the mean annual peak discharge in the pre-dam river (approximately 2,475 m3/s).

However, the duration of the controlled flood (seven days), was shorter than most unregulated spring floods on the mainstem that occurred annually in the pre-dam era, such as the spring flood of 1957 (fig. 6.4).

The results reported here do not support Kieffer's (1985) conclusion that extremely high peak discharges in the range of 11,000 m3/s were necessary during the pre-dam era to

significantly rework aggraded debris fans in Grand Canyon back to a mean constricted

width of 50 percent. Kieffer's study was limited with respect to fully understanding

debris-fan reworking, because much of that work was focused on a partially reworked fan deposit during 1983; seventeen years after it was deposited at the confluence of Crystal

Creek. Variable dam releases between 1966 and 1983 partially eroded and stabilized that

fan by progressively removing debris-flow matrix from the distal margin of the fan; a

process leading to particle interlocking, and clast imbrication that resulted in high friction

angles between large particles. However, the mainstem flows that affected the Crystal

Creek debris fan before 1983 were not comparable to the magnitude of pre-dam floods that 270 occurred in many years as a result of snowmelt runoff. The high dam release of summer,

1983, was comparable to the mean annual peak discharge of the pre-dam era, however, by

1983 its effect on the Crystal Creek fan was limited because movements of large boulders could only be initiated through forces of bedload entrainment, rather than by the combined

processes of entrainment driven by lateral bank erosion of matrix-supported sediments that

initially comprised the poorly sorted deposit. Had the December, 1966, Crystal Creek fan

been immediately hit by a mainstem flood like the 1983 high release, or the pre-dam flood

of 1957, it would likely have exhibited more erosion than it did during summer releases of

1983. Because Kieffer did not recognize that debris-fan - river interactions are a complex

process composed of a combination of at least two fluvial erosion mechanisms, her

conceptual model of debris-flow and river interactions was limited. As a result, Kieffer's

estimate of the flood magnitude needed to maintain mean debris-fan constrictions of 50

percent system-wide were related solely to reworking of coarse particles by bedload

entrainment.

In contrast, the general results reported here are in accord with the observations of

Hammack (1994) and Hammack and Wohl (1996), who report reworking of Warm

Springs Rapid on the Yampa River, and Webb et al. (1996), who documented the removal

of 2 historic debris fans at Lava Falls Rapid during the regulated era. At the Prospect

Canyon debris fan (Lava Falls Rapid), significant fan-deposit reworking was observed

during rising discharges between 1,000 and 1,200 m3/s. Two types of reworking or

particle entrainment on fans were observed during the controlled flood: (1) failure and

entrainment of matrix-supported fan clasts by lateral erosion (vertical bank collapse), and

(2) the entrainment of non matrix-supported coarse clasts from the bed of the river. Most

of the reworking of the 1995 Prospect Canyon debris-flow deposit, which had not been 271

subjected to river discharges greater than 670 m3/s before the 1996 controlled flood,

resulted from slab failures of unconsolidated debris-flow deposits that were laterally eroded

during the flood rise. It is important to note that the 1995 debris fan was at no time

overtopped during the flood, and that bedload transport of coarse particles contained in the

fan occurred initially in conjunction with slab failures associated with lateral retreat of the

distal fan edge in March, 1995, as well as during the rising stage of the controlled flood.

These multiple slab failures provided initial motion for large particles at discharges less than

what would normally be required to entrain boulders from a coarsened surface where

particle-to-particle friction angles dominate resistance. Other cobbles and boulders, particularly the ones embedded with radio transmitters, were entrained from the bed as individual particles despite the fact that they were not matrix supported (Table 6.8). This type of particle entrainment may be predicted using theoretical analysis (Wiberg and Smith 1987).

Reworking of the Prospect Canyon debris fan ended approximately 4 hours after the beginning of the rising limb of the controlled flood when large boulders armored the unconsolidated bank, preventing further bank failures. In this case, duration of the flood appeared to be unimportant to reworking, contrary to the observations of Hammack and

Wohl (1996) at Warm Springs Rapid. The two mechanisms of coarse-particle entrainment documented in this study have important implications for understanding the mobility and evolution of channel features, such as islands and channel bars downstream from large debris-flow affected alluvial fans and rapids, during both regulated and unregulated flows in the Colorado River, and the erosional mechanisms on fans observed during the controlled flood are likely very similar to those processes that drove fan reworking before regulation. Regulated flows have limited one of the mechanisms for debris-fan reworking 272

(entrainment of coarse bedload related to stream power), in ways that can only be partially restored by controlled floods in the range of 1,270 m 3/s. In principle, the lateral retreat, or bank failure mechanism of matrix-supported fan deposits has not been fundamentally

altered by regulation; movement of very large boulders can still be initiated by undercutting

of the fan's distal edge during rapid stage increases. However, the limited reach-averaged

stages associated with the 1996 controlled flood fell far below the maximum elevations of

several of the recently aggraded deposits, resulting in fan inundation percentages as low as

30 percent.

Significant reductions in the area and volume of recently aggraded debris fans were

measured at Prospect Canyon, Jackass Canyon, and at river miles 157.6-R, and 127.6-L.

The areas of other aggraded debris fans changed only slightly, and no change in area was

measured at 2 debris fans. Large volumes of sediment were removed from debris fans at

Prospect Canyon, Jackass Canyon, 18-Mile Wash, and at river miles 127.6-L, and 157.6-

R. Differences in stream power explain many of the differences in area and volume change

among the debris fans (fig. 6.13). Debris fans with high stream power before the

controlled flood, such as the Prospect Canyon and river mile 127.6-L debris fans, had Tanner large changes in area and volume, whereas sites with low stream power, such as at

Canyon and the aggraded fans at river miles 126.9-L and 127.3-L, changed relatively little.

The fan at river mile 160.8-L lost a large volume of fine sediment from between boulders,

but this type of change is not detectable from topographic surveys. The constrictions at many rapids, particularly at river mile 62.5-R and Lava Falls changed only slightly owing to Rapid, widened considerably, whereas other constrictions (fig. 6.14a). Initial and the stability of boulders at the distal edges of older fan deposits

reworked constrictions for the debris fans reported here fall within the range of 273

1,000 f ,m, 6,000 1995: 19951 - 0 5,000 1:1)=- 800 1994• CD 4,000 600 3 1993• 3,000 CD 0 400 1994 CD 2,000 cr 1987 • • •-1993 1989• • 1989 -n 200 1987 • • Volume 1,000 1993 11993 1989• • Area - 1989• 1989 1990 1993 1989 0 • I • " 0 0 500 1,000 1,500 2,000 2,500

Stream Power/Unit Width (W/m2)

FIGURE 6.13. Relations between stream power before the 1996 controlled flood

(discharges of 245-254 m3/s), the date of the debris flow that aggraded the debris fan, and changes in area and volume of selected debris-fan deposits on the Colorado River below

Glen Canyon Dam.

274 Prospect Canyon I I I 1 I A. 1995 7 160.8 Mile • 1993 157.6 Mile )-• 1993 Bedrock Canyon Specter Chasm • - 1989 1• -• 1989 127.6 Mile I • . 1989 127.3 Mile • 1989 126.9 Mile • 1989 Crystal Creek 1--• 1995 72.1 Mile • 1984 71.2 Mile • 1984 Cardenas Creek • 1993 Tanner Canyon • -1993 Palisades Creek • n 1990 63.3 Mile • 1990 62.5 Mile 1990 18 Mile 1987 Jackass Canyon ),-• 1994

60 I - I T - B. 50

40

30

20

10 0 111 h 0 10 20 30 40 50 60 70 80 90 100

Percent11111 Constriction, Cw

FIGURE 6.14. Percent constriction of the Colorado River by debris fans in Grand

Canyon. A. Changes in aggraded debris fans caused by the 1996 controlled flood. Solid circle, percentage of channel constricted before the flood. Vertical bar, percentage of the channel constricted after the flood. B. Percent channel constrictions (C,N) for 417 debris fans between Lees Ferry and Diamond Creek measured from March, 1996 aerial photographs (T.S. Melis, unpublished data). 275 constrictions measured throughout the river system at confluences where fans are present

(fig. 6.14b). Changes in percent channel constriction, which is a function of debris-flow deposit area, water-surface fall, and the stage-discharge relation, do not appear to be related to the elapsed time between the most-recent debris flow and the controlled flood. For some debris fans, relatively small changes in percent constrictions were measured, despite the fact that large particles were removed from the fan surface. The amount of reworking by the controlled flood suggests that the constriction ratio, which has been used previously to depict the reworking status of rapids (Kieffer 1985, 1987, 1990), is not a particularly sensitive measure of changes in debris fans and rapids caused by relatively small- magnitude debris flows or mainstem floods.

Because of reworking by the 1996 controlled flood, navigation of Lava Falls Rapid is essentially the same as it was before the 1995 debris flow (Webb et al. 1996). Velocities at about 250 m3/s on the left and right sides of the rapids decreased by half as a result of the controlled flood (Table 6.6). Although the total water-surface fall through Lava Falls

Rapid increased by 0.3 m (Table 6.7; fig. 6.10), the stage-discharge relation at the top of the rapid decreased by 0.4 m (fig. 6.10), causing exposure of several large boulders on the left side at low discharges. Velocities in other rapids, such as Badger Creek Rapid, increased; probably the result of partial removal of the aggraded debris fan, which had constricted the middle part of the rapid in 1994, decreasing the velocity at the head of the rapid. Increases in the velocity at 18-Mile Wash were probably caused by deposition of cobbles and boulders among the larger particles on the channel bed, decreasing overall bed roughness.

The relation between the amount of reworking, as measured by changes in area and volume, and the elapsed time between the debris flow and the controlled flood is largely 276 related to the history of reworking and armoring by lower discharges in the Colorado

River. The low-water controls at several debris fans that aggraded before 1993 and that were reworked by the January, 1993, flood were unchanged by the controlled flood.

Closest-packing imbrication and clast suturing (Webb 1996) were seen on parts of debris fans that had been submerged for long periods, typically longer than 3 years. For example, the 1993 debris flow that constricted Tanner Rapid (Melis et al. 1994) increased the stage- discharge relation by 1.2 m in the pool above the rapid. This change resulted in the increased storage of sand in the upper pool between October, 1993, and March, 1996.

Had the controlled flood restored the stage/discharge relationship in the pool above Tanner

Rapid to its pre-1993 conditions, this enhanced sediment storage condition would have likely reversed. However, large boulders deposited by the debris flow that control the elevation of the pool were unchanged by the flood, and the upper-pool elevation was only reduced by 0.25 m (a 20% decrease). If similar pool alterations occur because of debris flows throughout the river system through time, and these conditions are relatively permanent, then long-term benefits to sand retention in the mainstem may result from future constricting debris flows that deliver large boulders to the channel.

Debris fans aggraded after 1993 were more easily reworked; especially when they occuned in relatively narrow reaches. For example, the low-water control added by the

1995 debris flow at Lava Falls Rapid was almost completely removed (1 boulder remains), but the low-water controls at Bedrock and Specter Rapids, created by debris flows in 1989, remain. As the interval between the aggrading debris flow and the reworking flood increases, a flood will be less effective in clearing out the constriction, and larger peak discharges are required. If more significant reworking of aggraded debris fans is a desirable characteristic of future deliberate floods in Grand Canyon, the amount of elapsed 277 time between major debris flows and prescribed floods is one of the most significant criteria to consider in flood design.

Movements of cobbles and boulders, some with diameters as large as 1.5 m, were documented during the 1996 controlled flood. At 4 of 5 debris fans, 40 percent or more of the marked particles were removed from the debris fan, and others were transported as far as 38 m. As demonstrated with the radio-tagged particles at Lava Falls Rapid, most of the particles were deposited in the pool downstream from the rapid and not on the alternating debris bars farther downstream. The average travel distance by these particles of 230 m demonstrates the effectiveness of discharges as small as 1,300 m3/s in rearranging boulders on some debris fans in the Colorado River. These results, and the nearly complete removal of marked particles from the Palisades Creek debris fan, indicate that large cobbles and small boulders can be readily transported from previously reworked debris fans by floods half the size of the pre-dam mean annual peak discharge. These size classes are mobile at low discharges (relative to pre-dam conditions) in settings where particle-to-particle interactions and pocket topography are critical in controlling entrainment.

Based on observations that new-fan reworking occurred mostly on a time scale of hours during the rising limb of the flood hydrograph, the 7-day duration of the controlled flood was unnecessary for reworking of debris fans. If management of aggraded debris fans is one of the management criteria for design of future controlled floods, the duration of such a flood could probably be much shorter. However, the design discharge of 1,270 m3/s, which actually was as high as 1,370 m3/s in Grand Canyon, could be increased to have a greater effect on reworking debris fans, both in terms of bedload transport and increased stages that affect lateral bank erosion. Although some recently aggraded debris fans, such as the one at Prospect Canyon, were partially removed by the 1996 controlled 278 flood, others, such as the Tanner Canyon deposit, were not greatly affected because the

stage of the flood was not high enough to erode the side of the aggraded fan or overtop it.

Very short-duration and high-magnitude controlled floods would be highly effective in reworking aggraded debris fans. Such a flood could have a discharge as high as 2,800

m3/s, or twice the discharge of the 1996 controlled flood (less than a 5-year flood on the

unregulated river), but be designed with a duration of only 12 to 24 hours of peak

discharge at sites downstream from the Little Colorado River.

The conceptual model for interactions of debris flows and river floods on the

Colorado River, first presented by Kieffer (1985) and modified later by Webb et al.

(1996), is generally supported by the results of this study (fig. 6.15). A typical fan-rapid

system consists of a debris fan, a reach of whitewater, a downstream pool, and alternating

bars farther downstream. At stable rapids, the channel is usually constricted by 50 percent

owing to a reworked debris fan (fig. 6.15a; Kieffer 1985). Debris flows seldom dam the

Colorado River. Instead, deposition generally constricts the river by 60-80 percent. The

elevation of the bed rises, partly because boulders reworked from the debris fan are

deposited in the main channel or near the opposite bank (fig. 6.15b). Low discharges

(<1,000 m3/s) — whether small pre-dam floods or typical dam releases — may erode the

distal margins of newly aggraded debris fans and widen the constriction. The degree of

widening is dependent on local topographic conditions of the fan and mainstem channel, particularly the stage-discharge relation and water-surface fall, as well as the particle-size distribution of the debris-flow deposit. Debris eroded from the debris fan is deposited in the downstream pool (fig. 6.15c). High discharges (>1,500 m3/s) — large pre-dam floods or unusual dam releases such as in June, 1983, — typically remove most of a new debris- flow deposit, leaving a residual lag of boulders on the debris fan and in the widened rapid. 279

A. STABLE RAPID B DEBRIS FLOW

C w = [1 - 2Wr(ave)/( + Wd)] X 100 Cw 60-80% c w — 50 %

C. MINOR REWORKING, LOW RIVER DISCHARGES (<1,000 m 3/s)

D. MAJOR REWORKING, LARGE RIVER DISCHARGES (>2,000 m 3/S)

FIGURE 6.15. Schematic diagram showing a conceptual model of debris-flow aggradation and reworking of a typical debris fan by the Colorado River in Grand Canyon.

See text for description. This model is modified from Kieffer (1990) and Webb et al.

(1996). 280

At low discharges, the widened rapid may be more constricted than before the debris flow owing to the arrangement of immobile boulders and deposition on the downstream edge of the debris fan. Smaller boulders and cobbles are transported through the pool and deposited on the downstream debris bar (fig. 6.15d). Comparison of pre- and post-flood shifts in particle-size distributions measured on study fans with those of fully reworked fan surfaces observed throughout Grand Canyon (fig. 6.8b, c, and d), reveals the dramatic contrast in debris-fan reworking by the controlled flood and higher magnitude floods that occurred before regulation.

The most important advance in the original model of fan reworking proposed by

Kieffer was initially brought to light by Webb et al. (1996), and verified and quantified by the results of this study - that is - the recognition that debris-fan reworking by fluvial processes in a large river occur by a complex combination of erosional transport mechanisms. At least two separate mechanisms can be attributed to debris-fan reworking during the 1996 controlled flood, and both are limited by peak discharge, but in somewhat different ways. Entrainment of particles as bedload dominates in rivers where matrix- supported deposits are not present, while lateral retreat of unconsolidated, matrix-supported fan sediments during stage rise is a dominant factor in fan reworking where poorly sorted debris-flow deposits impinge on streamflow. While principles developed for estimating bedload transport in river are useful where particles are clast supported, understanding the interactions of streamflow on matrix-supported clasts transcends conventional methods. 281

CHAPTER 7

SUMMARY AND CONCLUSIONS

The present understanding of debris flows as an important geomorphic process affecting hillslopes and stream channels has derived from a long period of observations

(Blair and McPherson 1995a; 1995b), and only most recently, from controlled experimentation (Iverson et al. 1994). Debris flows are known to occur in nearly all types of geologic settings and occur under all types of climates when intense precipitation saturates weathered bedrock or unconsolidated sediments (Costa 1984; Johnson and

Rodine 1984). Debris flows may be the most important sediment transport agent in many arid to semi-arid settings where hillslope source materials are less subject to the stabilizing influence of vegetation.

Since the 1960s, studies on debris flows have increased markedly, and recent studies have advanced the understanding of debris-flow processes greatly. Observation and understanding of debris flows as a geomorphic-transport process has grown steadily in the U.S. as development and population growth has occurred throughout western regions where debris flows occur readily, such as the northwest. In arid to semi-arid regions of the southwest (including Grand Canyon), the relative infrequency of debris flows combined with low population densities meant that until recently very few direct observations of debris flows occurred.

During the twentieth century, only a few researchers had opportunities to directly observe debris flows in the southwest (Blackwelder 1928; Sharp and Nobles 1953). Direct observations of interactions between debris flows and streamflows in the Colorado River did not occur at all until very recently (Webb et al 1996; Webb and Melis 1995). 282

Recent environmentally motivated research along the Colorado River in Grand

Canyon National Park has provided a variety of opportunities to study how debris flows that are initiated in hundreds of bedrock canyons tributary to this large, regulated river.

Debris flows in Grand Canyon are the single most important factor effecting significant channel changes along the Colorado River over decade to millennial time scales. The co- occurrence of clay-bearing shales that promote debris-flow initiation, inter-layered with boulder-producing limestones and other erosion-resistant lithologies along the Colorado

River in Grand Canyon, allows debris flows to actively influence the geomorphic framework of this large, arid-land river.

Debris flows occur in at least 529 tributaries of the Colorado River in Grand

Canyon between Lees Ferry and Diamond Creek, Arizona (river miles 0 to 225). An episodic type of flash-flood, debris flows transport poorly sorted sediment ranging in size from clay to boulders from Grand Canyon tributaries into the Colorado River.

Impoundment of the river by Glen Canyon Dam in 1963 has greatly reduced the river's sediment load, promoted greater erosion of fine sediment stored in the river channel, and reduced annual flood frequency and magnitude relations that previously countered river- channel aggradation by debris-flow sediments. A management goal for operation of the dam now includes the restoration and protection of downstream river resources; including decreased erosion of sand bars (referred to as "beaches") located at hundreds of sites along the mainstem. Data related to the frequency, magnitude, and sand content of debris flows are important in accurately calculating the sediment mass-balance of sand transported by the

Colorado River through Grand Canyon. This information is required for future "Adaptive

Management" strategies related to preservation of sand bars and other associated river resources. 283

Debris flows in Grand Canyon tributaries are initiated by hillslope failures that

occur in three primary ways during intense rainfall. Failures in weathered bedrock,

particularly in the Permian Hermit Shale and Supai Group formations, have initiated many

historic debris flows in Grand Canyon. A second initiation mechanism, termed the "fire-

hose effect," occurs when runoff falls from cliffs onto unconsolidated colluvial wedges and

causes failures. A third initiation mechanism occurs when intense precipitation causes

failures (soil-slips), in colluvium overlying bedrock hillslopes. Multiple source areas in

combination with extreme topographic relief of tributaries in Grand Canyon often result in

complex types of debris-flow initiation comprised of combinations of these mechanisms in this unique geomorphic setting.

Interpretation of 1,107 historical photographs spanning 125 years, supplemented

with aerial photography made between 1935 and 1997, have provided frequency

information on 168 of the 529 tributaries (32 percent), designated in this study. Of these

tributaries (those shown in the historical photographs), a total of 112 contain evidence of

historic debris flow (je. debris flows that have occurred since 1872), or 67 percent, while 33 percent of them did not have a debris flow during the last century. Additionally, eight

debris flows were also documented from non-photographic sources, resulting in a total of

120 known debris flows that have reached the Colorado River in Grand Canyon since

1872. These frequency data suggests that debris flows occur, on average, from once every

10 to 15 years in certain eastern tributaries, to once in more than a century in other drainage

basins. On average Canyon-wide, debris flows may recur approximately every 30 to 50

years in individual tributaries, although adjacent tributaries may have considerably different debris flow histories. 284

Peak-discharges for debris flows were estimated in 18 drainage areas for debris flows that occurred between 1939 and 1994. Typically, peak-discharges for historic Grand

Canyon debris flows range from about 100 to 300 cubic meters per second (m3/s); the

largest peak-discharge for a Grand Canyon debris flow during the last century was about

1,000 m3/s. Debris-flow deposits generally contain 15 to 30 percent sand-or finer

sediment, however the variability of sand-or finer sediment contained by recent debris

flows is large. Recent debris-flow deposits also contained undifferentiated micas and

kaolinite in significant proportions relative to other clay-producing minerals. Water

reconstitution of Grand Canyon debris-flow sediment samples indicates that the water

content of most debris flows there ranges from 10 to 25 percent by-weight.

Debris flows are important in Grand Canyon, because they create and maintain

persistent, bouldery debris fans, and the hundreds of associated rapids controlling the

geomorphic framework of the Colorado River downstream from Glen Canyon Dam.

Debris flows commonly transport large boulders weighing between 100 and 400

megagrams into the river channel. Several recent debris flows have created new rapids or

have enlarged existing ones. Before flow regulation, alluvial fans aggraded by debris

flows were periodically reworked by large river floods that may have been as large as

11,000 m3/s, based on recent paleoflood studies. Flow releases from Glen Canyon Dam

since 1963 have partially reworked recently aggraded fans. Significant reworking of new

debris-flow deposits now occurs only during river discharges higher than typical hydro-

power plant releases, which presently range between 142 and 510 m3/s.

Debris flows have both direct and indirect effects on the formation and stability of

channel-stored fine sediment (mainly sand bars), along the Colorado River in Grand

Canyon. Debris flows often bury or erode existing sand bars when they reach the river 285 channel, and these effects are normally only mitigated by sediment-laden river flows which exceed typical power-plant generating releases from Glen Canyon Dam. Debris flows also indirectly affect sand bars by controlling the geometry of channel-constricting debris fans over short and long time periods. As debris flows modify alluvial fans through aggradation or erosion, they also affect flow-separation zones below those constrictions which trap and deposit sand to finer sediment. Finally, debris flows that increase the degree of river-channel constriction at debris fans can promote storage of fine sediment in the river channel's bed. This occurs as a result of increasing the size of low-velocity pools that form upstream of rapids. Ultimately, it is the three-dimensional geometry of debris fans (particularly fan height), that controls the size and shape of sand bars formed along the

Colorado River. Fan geometry and channel morphology restricts the range of discharges in the river at which sand bars may deposit and remain stable.

Quantitative data on the flow-behavior of infrequent debris flows in remote Grand

Canyon tributaries are difficult to obtain directly. However, I infer flow behaviors for several recent debris flows along the Colorado River from their sediment deposits and from flow evidence preserved in bedrock-channeled tributaries. Based on superelevation data, debris flows from 1939 through 1993 had peak discharges ranging from 100-1,000 m 3/s.

Based on four-dimensional analyses of uneroded alluvial fans, I propose three types of hydrographs to explain the flow behaviors of these extreme events.

Type I, has a single debris-flow peak followed by a decayed recessional streamflow that forms massive, lobate debris fans with small streamflow-incised channels. Type II, has multiple, decreasing debris-flow peaks with intervening flow transformations between debris flow and non-debris flow phases that form fans with complex morphologies, stratigraphies and textures. Type II fans are a composite of multiple debris-flow lobes, 286

streamflow gravel deposits I interpret as being associated with what is referred to as

"hyperconcentrated flow" in other parts of the world. Type II fan stratigraphies indicate

that hyperconcentrated sediments deposit between debris-flow phases and before late-stage

streamflow recession. Type III flows may have either a simple or complex debris-flow

phase (begin as either Type I or II), but are followed by a larger streamflow peak that

extensively reworks debris-flow deposits or buries them under a layer of streamflow

gravel. Protracted storms that cause extensive late-stage streamflow reworking of debris-

flow deposits make interpretations of flow behavior, based solely on fan morphology and texture, impossible.

The channel of the Colorado River through Grand Canyon is dominated by debris-

flow processes over at least 225 miles. Local characteristics related to fan-eddy complexes

vary by reach along the mainstem and control recirculating flow, sediment transport and

sediment storage. Storage in the channel occurs as alluvial sand deposits in deep pools, as

terrestrial sand bars along shorelines, and as coarse alluvial fans and channel bars that are

relatively permanent channel features. Terrestrial sand deposits occur mostly within "fan-

eddy complexes" on or near alluvial fans, and respond dynamically to sediment availability

and flow magnitude over short time scales.

In contrast, most reworked alluvial fans are stable boulder piles, while channel bars are comprised of cobble to gravel-sized particles. These coarsened deposits are far less dynamic than sand deposits, clearly reflect local source lithologies, and are only modified by large mainstem floods like those that occurred before 1963, or by tributary debris flows that transport boulders from tributaries up to several times per century.

Analyses of key attributes of all fan-eddy complexes along the river in Grand

Canyon indicate that the eleven previously designated geomorphic reaches of the channel, 287

based on bedrock lithologies at river level, can be combined into six reaches that better

reflect local features of fan-eddy complexes known to control flow, sediment transport and

sediment storage within this relatively unique debris-fan dominated river.

Ongoing biotic and abiotic changes that occur within the Grand Canyon river

ecosystem will continue as a result of introduction of non-native species, variations in

sediment flux related to climate variability, but will mainly be effected by river regulation.

One of the most fundamental effects of regulation below Glen Canyon Dam since 1963 has

been a significant shift toward greater influence by debris flows on the mainstem channel

and its resources. This shift in the geomorphic framework of the river system will continue

to be expressed by changes in fan-eddy complexes, and these changes will likely alter the

sediment balance of the river system (increased sediment storage in association with

aggrading alluvial fans). Ongoing physical changes related to altered process interactions

throughout the riverine ecosystem of the mainstem can only generally be predicted at

present, since linkages between river geomorphology and biological activity is not fully understood.

From 1987 through 1995, debris flows constricted the Colorado River at the mouths of at least 25 tributary canyons in Grand Canyon National Park, Arizona, creating

2 rapids and enlarging at least 9 riffles or rapids. Previous research has shown that modest powerplant releases from Glen Canyon Dam, particularly in combination with tributary floods, can significantly rework aggraded fans, entraining boulders up to 1 m in diameter, although discharges greatly exceeding the maximum powerplant release (946 m3/s) would be required to completely remove most aggraded fans. The highest peak discharge between

1986 and 1996 was 960 m3/s in January 1993. In March-April 1996, I studied the effects of a 7-day flood release that peaked at 1,300 m3/s on 18 recently aggraded debris fans 288

downstream from Glen Canyon Dam. I used a combination of pre- and post-flood surveys

at about 230 m3/s steady discharge, photogrammetric analyses of pre- and post-flood aerial

photographs, and direct measurements made during the flood.

The largest changes occurred at Lava Falls and Badger Rapids, 38 and 312 km

downstream from the dam, respectively; several other aggraded debris fans had little

change. Plan-areas of aggraded fans decreased by 2 to 42 percent; only the debris fan at

Bedrock Rapid, which is affected by a large bedrock boulder, increased in area. Volumes

decreased on 7 of 9 fans by 3 to 34 percent. The distal margins of most recent debris-fan

deposits became armored with a lag of cobbles and boulders, and the width of the

reworked zone on most fans increased by 4 to 30 m. At most recent fan deposits, the

width of the constriction increased, although some rapids, such as Tanner Rapid, are

slightly more constricted at low discharges owing to changes in stage-discharge relations.

Velocities on the left and right sides of Lava Falls Rapid decreased by about fifty percent,

but velocities increased in other rapids, such as Badger Creek. Streampower per unit width

decreased in 7 of 10 rapids owing to decreases in water-surface fall and widening of the

width of rapids. Changes in the sizes of separation bars downstream from the reworked

debris fan were inconsistent, although the upper pool deposit generally decreased in size.

The amount of unit streampower generated by the controlled flood greatly affected the variability of reworking among the 16 debris fans. Because of armoring of debris-fan margins by smaller dam releases combined with tributary floods, the elapsed time since the debris flow was also important because particles became interlocked, imbricated, and (or) sutured together and could not be dislodged by the controlled flood. The effectiveness of future floods of similar peak discharge in reworking new fan deposits at major rapids will depend on the release history and extent of armoring in the period between the debris flow 289

and the flood. If reworking of debris fans is a criterion for design of future controlled floods, I recommend a higher peak discharge of shorter duration because (1) some

aggraded fans in wide reaches were not overtopped by the 1996 controlled flood and (2) reworking mostly occurred during the rise in the flood hydrograph.

Grand Canyon debris flows occur infrequently, but provide a strong geomorphic

control on the Colorado River through Grand Canyon National Park, Arizona through

deposition and maintenance of alluvial fans. The high degree to which flow and sediment-

transport/storage are controlled along the mainstem by stable constricting fans reflects a

river geomorphic framework that may be termed "debris-fan dominated." Because river

regulation has upset the former relationship between tributary deliveries of sediment and

mainstem reworking by periodic floods, the reach below Glen Canyon Dam will likely

become more dominated by debris-flow processes in the future. Physical changes to the

river channel associated with this predicted trend will likely alter both the biotic and

physical character of the river there, and will do so in ways that are only poorly understood at present.

Summary of key results

Several important conclusions can be drawn from these recent studies on debris

flows and their impacts to the Colorado River in Grand Canyon National Park below Glen

Canyon Dam:

• Tributary debris flows occur infrequently in Grand Canyon National Park, but are important sediment-transport agents, particularly with respect to delivering large boulders to the Colorado River -- a process that contributes to the ongoing excavation of Grand

Canyon; 290

• Frequent, smaller-magnitude debris flows occur mainly during localized summer thunderstorms and are likely most important in fan evolution over long time periods.

However, data on twentieth-century debris flows in Grand Canyon indicate that less- frequent, larger-magnitude debris flows, initiated by atmospheric circulation patterns from the north and south Pacific Ocean, may impose the greatest influence on river

geomorphology in Grand Canyon over periods of decades to millennia under regulated

river conditions;

• During the Holocene, debris flows have deposited boulders in the Colorado River

at hundreds of tributary confluences. Debris flows that deliver these coarse clasts have

likely maintained rapids over millennia as relatively permanent features that armor the river

channel and impose a "stair-stepped" base level of erosion that temporarily inhibits river

downcutting into bedrock. While vertical incision of the river channel is slowed, lateral

retreat of bedrock tributaries is facilitated by ongoing weathering and transport by debris

flows. If correct, then such a scenario favors widening of Grand Canyon during relatively

drier climatic conditions associated with interglacial periods, and deepening during wetter,

colder glacial periods that would likely promote large Colorado River floods while limiting

tributary debris flows;

• The geomorphic framework of the Colorado River in Grand Canyon is dominated

by river-reworked alluvial fans that result from interactions between tributary debris flows

and mainstem fluvial processes that occur at over five hundred confluences throughout

Grand Canyon -- such reworking interactions are complex, with aggraded fans eroding

through a joint process of lateral bank failure/retreat of unconsolidated sediments, and

commonly described processes of bedload entrainment; 291

• The primary factors that directly control mainstem channel characteristics (mostly with respect to channel width) are: grouped bedrock formations with similar lithologies that occur at river level, and large-scale geologic structures (faults) that affect the positions of those grouped lithologies relative to river level;

• Secondary attributes of the mainstem that control flow and sediment transport are fan-eddy complexes, created and maintained by debris flows, key characteristics of these fundamental channel units are, in turn, influenced by system-wide variations in tributaries that influence debris-flow frequency and magnitude throughout Grand Canyon;

• Stable alluvial fans in fan-eddy complexes tend to be dominated by limestone boulders derived from formations in tributaries above river level -- the largest bouldery fans tend to occur in wide reaches where slope-forming lithologies occur at river level, but often these fans do not support rapids;

• During the pre-dam era, aggradation of alluvial fans by debris flows was limited by river-reworking during large mainstem floods occurring almost annually, but reduced annual peak flows since 1963 have accelerated rates of fan aggradation and rapid formation system-wide -- with implications on evolution of fan-eddy complexes system-wide;

• Relative to many other bedrock/alluvial river systems, the Colorado River in Grand

Canyon has a unique debris-flow dominated framework resulting from highly varied debris sources that occur as both inter-layered clay-bearing shales and other boulder-producing lithologies such as limestones and sandstones;

• Before and after regulation by Glen Canyon Dam, small and large debris flows exert(ed) geomorphic control on the Colorado River in Grand Canyon by contributing large 292 boulders to alluvial fans that exceeded the competence of large, fan-reworking mainstem floods -- these stable, coarsened fans constrict the mainstem, creating zones of recirculating flow that deposit fine sediment in hundreds of fan-eddy complexes as terrestrial sand bars and subaqueous channel deposits;

• Accelerated aggradation of fan-eddy complexes by debris flows in the regulated era may result in channel modifications that increase sediment storage in the Colorado River throughout Grand Canyon -- such physical changes may have ripple effects throughout the river ecosystem that can only generally be anticipated on the basis of present understanding of physical and biotic system interactions and linked processes.

Implications for management of river resources below Glen Canyon Dam

The dynamic nature of fan-eddy complexes of the Colorado River channel in Grand

Canyon is reflected in the numerous tributary-induced changes documented on fans and rapids below Glen Canyon Dam since its closure in 1963 (Cooley et al. 1977; Webb et al.

1989; Kieffer 1990; Melis et al. 1994). The dynamic interactions between tributary debris flows and large mainstem floods in the pre-dam era is also reflected in the characteristics of reworked fans observed throughout the river corridor, and in the pre-dam instrumental and reconstructed paleoflood record of flood peaks (O'Connor et al. 1994).

Because tributaries below Glen Canyon Dam remain unregulated while mainstem floods are drastically reduced, the ongoing evolution of fan-eddy complexes will continue at an accelerated rate. System-wide, this trend will likely result in a generally less- navigable river channel, as formation of new rapids and aggradation and steepening of existing rapids occurs. Fan aggradation will also result in increased local constrictions of the mainstem in fan-eddy complexes; such changes that will likely cause enhancement of 293 recirculating-flow within enlarged eddies.

Enlargement of eddies may allow for greater sand storage at hundreds of sites

impacted by debris flows. However, pools between fan-eddy complexes will also

eventually aggrade with coarse sediment because regulated flows can only transport coarse

fan-aggraded debris a short distance downstream during limited reworking. These pool-

aggrading sediments were formerly transported further downstream, and often were

deposited on active cobble bars during high flows. Over the lifetime of Glen Canyon Dam,

the shift from large-scale reworking of fans and cobble-bar mobility, to aggradation of

fans, pools and inactive cobble bars may reduce the net volume of fine-sediment that can

accumulate in storage within the system.

It is not clear how such a trend will affect other components of the aquatic

ecosystem, such as habitat related to fisheries, etc. However, on the basis of existing

information on habitat use by fish (Valdez and Ryel 1996), and aquatic productivity in the

mainstem (Shannon et al. 1996), eddy enhancement, rapid evolution, pool-aggradation and

stagnation of cobble bars may have long-term implications for the river's sediment balance

and ecosystem health. Only through better-integrated monitoring and research activities will scientists and resource managers eventually be able to anticipate and model future

changes in the ecosystem resulting from fundamental alteration of the river's

geomorphology caused by the combination of river regulation (reduced mainstem flows),

natural or man-induced climate variability (variations in flux of fine sediments), and ongoing natural tributary debris flows (local-scale alterations to eddies occurring system-

wide through time). On the basis of Griffiths et al. (1996) predictions of future tributary

debris flows, and the relatively close spacing of tributary confluences (fan-eddy

complexes) throughout Grand Canyon, the system-wide responses to local-scale channel 294

alterations should be detectable over the course of a few decades to a century.

Tracking geomorphically driven trends in river resources should be a primary goal

of any long-term river-ecosystem monitoring program below Glen Canyon Dam. It is

critical that the spatial and temporal patterns of geomorphic processes of the Colorado River

and its tributaries be viewed in terms of the spatial and temporal patterns of other resources

of concern, such as sediment inputs, fisheries habitat, primary productivity, camping site

availability, etc. (fig. 7.1). Only when all of the factors that influence the river system are

integrated into a conceptual model or a series of linked models, can stated management

objectives (DOI 1995), be actively pursued and achieved. A river restored?

Unfortunately, restoring the Colorado River to a condition that existed before about the mid-nineteenth century is probably no longer a realistic possibility (Schmidt et al. 1997;

Collier et al. 1995; Hirsch et al. 1990; Williams and Wolman 1984). Even without large dams, the river's ecosystem would remain altered by the existence of non-native species of plants and animals that were introduced before the first large dam was closed. Most significant among these introduced species are: predatory fishes such as trout, carp, large-mouth bass and catfish, as well as terrestrial plants such as tamarisk and camlethom. It is not clear as to how these exotic species could be extirpated from the river's ecosystem, even if dams no longer impounded reservoirs throughout the basin. Likewise, it is unclear as to how these introduced species would have impacted native fish and plants if the river had never been impounded by large dams.

The objective of restoring and preserving river resources below Glen Canyon Dam, as well as other large dams, remains a moving target in terms of management. This is true whenever non-regulated, geomorphic processes, like debris flows, continual alter river

295 oz O 0 -1 (D o Z HIGH 0 HUMPBACK CHUB 0i <1 CD /;' DISTRIBUTION CO CI_ (Gila cypho) O cr, = E LOW LOW PRIMARY AQUATIC PRODUCTIVITY (Decreases downstream owing to turbidity)

HIGH w w cc ce 1-- cc -- cc x cc cc cc cc a cc 4I Lu w LL, Lu z Lu w w 3 3uj 1--0 Li-itc = 3 Lu 3 = 3 = 3j 31 o 0 0 0 0 0 0 0 0 0 0 0 _I _..1 o E _J 0 _.1 __I "±- Z m

AVAILABLE SAND IN STORAGE BASED ON CUMULATIVE LOW DOWNSTREAM INPUT

.HIGHER POTENTIAL LOWER POTENTIAL HIGHER POTENTIAL

HIGH , I I I I I 0-1 w II _J4 n- wj.: a cc co -w w0 X i-- 0 CO 0 M W a_ 0 LOW 440 4 4b (.3 = w = CC z w = ct - Ix ô --.. 0 0 0 - cr 0 ...1 - • II Z»Œ w cr MIDDLE NARROW w w w im OUJ 0 . LUZ 0..(X Ct—.. 0 Lu 00 0— 04 Si-8;:.1 111 -- ..., ....IZ , , ,

-16 0 2 4 80 no 120 140 160 180 2 225 240 RIVER MILE

FIGURE 7.1. Conceptual model showing relative spatial and temporal relationships

between estimated future potential of tributary debris-flow impacts to the mainstem, varied

geomorphic river reaches, and the physical and biotic resources of the Colorado River

ecosystem in Grand Canyon. 296 geomorphology. The physical setting of the aquatic ecosystem in Grand Canyon will continue changing in distinct and predictable ways as long as tributary debris flows are not counteracted by natural-scale mainstem floods. As future conceptual models are developed for the Colorado River ecosystem, a key component in the ability to predict long-term impacts of regulation will rely on continued monitoring. Designing monitoring protocols and programs that establish linkages between geomorphic change in the river channel and resulting changes in its sediment balance are relatively straightforward. However, establishing cause-and-effect linkages between physical changes in the river channel and biotic-ecosystem processes presents a greater challenge. Overcoming this challenge is key to the long-term success of preserving the unique components of this large, desert-land river. ▪ ▪ ▪

<-1--1 "Cla) 0 -0 ct tb

- i•::,' 4F,7i s•-•

. a.) '."" u c7s 4E!) ... .c • . -rD 0 . 4 01-) a) -0 cA›..,..--. '-' ‘,,p

czt T't ,.._, 4,-A E ›, z ,,.., — O 0 a' co 0 4.-1 -5 (-) b.0 "0 0 ...5 .4 O U L.,'-7-'• 0 z;_•,, , cct ct o .,- t:IJ) a.) 0 ab , 71- 71- un 7t rn 2 ,---6 '4 ,.,11- ,2 • "CS.. oo 00 0o0 NN • s I .1:) I C4 r- (=> r- r, r- r- r- r- CN CN '' ;-, C+-," ,a) 00 r- c.\ cN 00 cs.•\ cp, 00 00 00 00 00 00 00 O • 0 ,-Z CN CN CN CN CN CN CN CN CN CN CN 0 (-A ,-. N N 4-1 1-41 1-1 4-41 cd U - (-) 77' cd QI., cd 1..) C....) 1_,cd 0" ,--, 77 cd O y b10 c) 173 os5 = • •-n 8 ,i-) acS, . 0 = (.D oo4c,,-)og . _ f• zi.,e, ›, g O cf) v) .g -.,.g g ,e • ai 00 g 0 ct ts,0 tso Of) !DO .. a) (...) aZ ep `3' 4 fM-1 0 . ""' • g g ,9 O ;,..• 1)$_, •-•-•• 1- p., • = a) çà cn • (14) . ..n • 5 '-(1.--') 5 .5 (:) O 4-. „0 0 174:1 0 "c ,C) c41:_,' 0 › • . 1.) , _ ct cu 1-4 g g g g -cs 4 4 .6 4 4 (/D (1) (0O0 y y (/) C/D c/1 Cf) C/D c.f) c/5 -ci › Ci)Y ô' Y • • ---, -c1 a.) 0 4-- 0 an .‘"F2 0 v)000000 -0 ccs --, aaa aa cp c t 0 . 000000 c..) CL) CA 0 0 77) o Z "Ci C78 •• ,5 -o o 7::, -0 -o -0 O s-, 0 ..= ''' 4-, 0 ,E, 0 .,,z, 3 -c) r•c• 3 0 3 3 3 3 3 3 3 O cri 0 6 ;4-1 Y, 0 1 0 -0 ,-, 4-41 0-., 4.-., i 0000 0 0 0 I 3-4 ;-1 ;.41 I :::41,1111 ''-- 0 V) 0 0 0 ..-, E CA 4.., . C/) . Ci.) . CA . CA . CA . C4 E 3 • — —,.,.0 t113-d• '-' C /D •s7; cl (1) 7," 7,4 r.." 72, as = al-, Z •- 0 -5 > > > -0 -0 -0 -0 ,.10 a) 0 . a) 0 u a) ai a) a) (1.) (71 ,:i..)- sin cf) , p4 a .62 c:2. o4 C:1 00 1.) s_, cfn- a) ---. —• ccs .,-0 1::: cu :- ct ,- 1-, 0 -,.. o . ç,.) o O = 4-. ‘--- "Tn -0 cn a) "0 = ri„) .--, c.,.., -O •-n t ,.. o a:1 „--, 0 .0 -0 -0 ›, a) c/D C-N1 0.. (71 = 1.) , , 1..) ,.....• 1..) › bf).t bi94-.n s.-n s-, 0,--n z00 0 /D:, -5 c.' c. b.0.-0 ra, ::)., 0 •,_--, o ;-, c• ,.,-., m •- c•-, o ;-n <:) ;-, 0 O E ô 0 ct ,..0 0 — aL, '' 0 0 ›,-E ,.,..,i —0 ..o bi:,, — a) ,-, a) a) ,.. z y .. 77»z: api) -. 0 0 • -• z 8 u ."'-' az$ p , .. ct,- ,4 4_,- si) 4 a4 a4 o4 4 •• ai E c.. 1..) .,=, " 0 0 > c/D azi g. (1) (L) 1-) If:) O 0 >, 0 0 , E-i u •,.,•5 r.,T,Z ,--.. (,)a) ,---, tko ,_, n C\! Ce:). . °°. C:). 77. (:) Cq, cl., an su t----- d = .,t4 r- C2) o CN °C r". --I r- 00 CN CN .7(1- ,f ,t -= r- ;-, ct PL, (cDg E. On 1.) ..= (-7-) < s • -

cq cl c‘.1

ce) 7t 71- 7t Ir) 71- 00 00 rr) 00 I 1 00 tr) CC) In 0.N (D ch ch \0 Cn Cr) C,I) r, r- en ro, re) ee) r- 0 C C r- 00 Ch 00 00 (N (A r- ch 00 ch 00 00 00 CO 00 ch lD ch ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN N

CI) (.1) C./) (.) (IDn c.) 1.) a) 1.) a) a.) +_ ,-, t10 00 b.1) b.0 0.0 . c",1 . 'Ci) 0 0 00 czi ct ct g g 0, al.n g: -= a.) a.) c.) a.) c.) 00 -ci -0 -al -1-.)." -a") O 0 0 00 00 0 g g g = 4 4 4 4 4 4 4) ") 4 4 4 4 4 4 (-) (4 (-) • c'7' Y ' c7) ' c2 (4 (4 (4 (-) Y . 7; • 7,' • c7) '7) ..,t1 1..,) 1.), o o .., i..1,-1) c.) o o o h h 't'il, -cii 0 o o o c.) ct ct 0. 0. - 0., sa,fa, = 0 0 ;-, çmL.,0,01,0-. O 1--, s„, a.) (D c4 a) a) cu cd ct ,,,1 to 0 ct 0 a) a) a) "j 71: -0 ,g , 5 "1:i o 0 .7 ,1.) c..) ci) ,c. () 71. ,C, ,t: 000c) ct ,4 ,-, C) C.) - • ^-, ,__, 0 0 11."" "C.J „17.; 0 ,-, 0 0 c...) o ,,$r_., 0000 C 0 0 ç s-, qc.54 s_, : yc...5, c..1 qc.5, -, u, t+-1 0 C.) " 0 -c't 0j) 0 (1.) r-, 0 1 ' " . 4 `• 1- . v) . 4 . ' 4 .4 .4 .` g O a) a) • ;.:1 $1.-.), ;-, ,.._ ,t), i:, ,1.-.5 5--, i72, 1.-. s, 'I:: 0 cd t '17-1 . :Il . $7:n . r---n O ta-. ta-,_ra _sa ›. -0 o -0 ›. „ta -0 -0 -0 -0 cd 1.) -.., ›-, ›-, a.) a) . a) 0 .,, a) •..., a) a.) a) a) a) ,.0 •. 0 a.) a.) 0 . r:' Z 62 a) o4 a4 o4 u edw)' 2

00 0 0 0 0 0 ',) :-. ' 4 2 5 5 "5 5 5 '-c, '-,, g '') 5 • 5' ct ct ct 5 5 -= 5 -0 -0, ,...9 4, m cz1 L) _CD 42, ,..C) C.-) c..) C.) 4, ,4 2,c, ,4• •-4Dn •-4nn • n..1 • 1.. `... • 1. ,,, ,---, • I. • ••••1 „o.', • Il n 0 Ail 4.'t .it 4'7j ›.' ,it Nb' Nit .n7.; " " " .;:j 1.1 = 41t M Ct Ct Ct C.71 Ct 7:1 "Ci 17J .7) /i ct C) 0 0 fl vi 7J 77 77 0.) a) a) '10 7:) r ) 7:) a) a) a) (1) a) u a) v) c/) coi 1.) a) a) a) a.) ,--, a) '/DoEctE000EE-=E EEEE c ct ct ,,,s czs m ,4 -0 -0 ,.s4 4 czt 0) ct .._ ,_ ,... cci ct l O 0 0 0 0 '5 0 ct et c.) c) ,2., 0 0 0 0 0 0 = = ,-, z O 0 0 0 0 0 0 ,,r ,t 0 ry, ct 0 , 0 ct czt ct 0 0 (...) 0 co0 H E-' P4 7' P* C.4 Z Z Z

a.) P4 0.4 o4 p4 o4 a4 o4 o4 1c4 o4 o4

n s-4 00 Cl Cl Cl V") 7t 0"). Cl r- lD 00 Cl Cl Cl Cl le) ID (:) (L) vj (D 6 6 6 tr; -. c\i ce.) fi rq rn rn rn rn ce) rn ce) cn 71- 7t 7t 71- tr) kr)

CSn C•1

1-1 14.1 1-1 4-.4 1-1 1-1 r. 1-4 1-4

m 7t - 4 71 Cn 71- 7t r "r4- 71- 7t 71- ,t r- oo oo oo r- oo oo oo oo III 00 00000000oo oo oo m m s.r.) 1.r) mmccim m 14-) m m o 0 CD i3 r- r- C C r- O D r- r- r- r- m m N D VD r- r- GN Ch CN CN CN 00 CN V:), ch CN CN CN CN CN (0 Ch 00 CN CN Ch CN Ch CN CN CN Ch CN Ch —. Ch CN N csi CANNNNNC,1^—I,--,C,INNNCV

cn cn ci) cL) cn o . " cn &D 0 c . 0 cci O-, ,-. fa, ,4 C.) 70 U 75 7:i 7.) 3 3 cn cn (::) 113_, ,--,5. cn 0 c9cn cn cn 4_, .4_, • ,..—, 4_, , .4_, 4_, ,4 ci) , , • ,—, ,....., • —, • ,—, v) U (ID cl) cn cn 77, ,7:::) cl) cf) (ID cl) cl) O 0 1) çl.,0 ,,,l' , 0V) 0Ld) 0V) 0V) a) , a.) a) a) a) 0 ,.., a) a) a) a) a) s:),o 00 b•A 00 00 ca.4 '—' 0.0 0.0 ID-0 b.I) ID-0 çl' P" C-'-':n 1-) L...' ' " 4C-4 2)"' '''' Z g Z n ,__,C g = Z CL) 1..) i•- 7:;) %, CL) S. CL) CL) Cd -6-. Cd Cd Cd — I4,-1 Cd Cd Cd Cd .4 ,4 .4 ,4 ,4 = ,4 = .4 ca.3g,-.2 8 0 ,ID 0 U 0 0 0000 0 O CI.,;ZIC.D0(:)= .111 (::)(:)== Cd ''' Cd Cd Cd Cd Cd al M ct czt = 0 I I. 0 = (.1. ,...,1. c+-1 4. 4. c+, , E 4. t...1 4. ,..1-, c+. 1 1 ocnoi Q I 1 ,n cn ,_, ., . ,,, ,n c/., cl) ,_, cn cn cn cl) cn cn cn cl) cn cn •—, • —, 4 , •-, 1_, • —, • . • . • .1 ;_, • 4. • . • . I. . IL.' _ci ,I) s-, a) s-. 1-, s-, „.) *::", ' s-, s-, s-, (i) 1-, 1. 1-, s.., , s: ,..0 0 ., (L) › _0 P-i,./3 -0 -0 › -0 -0 -0 0- -0 " 4) ,SZ) -0 -0 -0 a) a) >, e-N ." a) .--, a) a) a) . a) a) a) cu a) 4-' a.) a) a) a) a) Z t71 4 = a) a) ?? a) a)

5 5 5' "FI') 7'4) S' 5 5 5 z au a) co -n ,0 0 • '1"." \/—"r)'—'r);—" 0 a) •-.) — (I) y., 44_4 c..) g ?) 71?) 6 6 6 6 g "Zi e..") -F, -0 -0<,..)

'g T. c-) u ,E E (I ci) CIS (CI — CL) TEI CL) (1) CCS CL) CL) — , n ,c(ci 2 z ••— z ,) P. P. P. P. 4.1 (.) E-1 1:P 4.)1

a4 )-4 4 o4 .4 1-4 p4 4 04 0.4

m m oc) tic) oc) les m m oo ci oo c: V') \..0 cq cc) 4 ln

Cn 4 Ir) Ir; tr; (C-; tr) tr) Ir; O N N 0000 C c5 6 6 c5 6 VD VD VD VD VD VD n0 VD VD VD VD VD VD VD VD N N r- r- r- •

Cr)

• Cl cl cl cl ,.

../- 76 d- d- d- d. d- d- dr. d- d- 71- 00 0000 00 00 00 00 00 00 00 00 00 1 1 1 1 1 1 1 1 1 1 1 rn rn (4) vn rn rn rn un rn rn rn 71. n rn r- .0 7t r- 7f* n,c) r- n.0 r- r- r- C r- r- 0 r- r- rn \CD \ID Ch 00 Ch ch ch 00 ch 00 \ID ch ch ch ch ch ch ch 00 Ch 00 ch 00 \.0 ,.0 ,.--+ Ch Ch ,-, .--, Ch ,--1 Ch Ch ,-, ,- ,-, -- ,--( ,--d Ch Ch Ch ,- Ch Ch Ch N ,-", ,'- N N ,-- N ,--, ^-( N N N c,) N N N ,.. .--( ,-1 (.1 .---( ,-.1 ,....

c4 c4 (4 1.) (1.) <1.) bl) bl) bA o g g c`b = ,- Q C.) c..) a) a) -1.7.i 00 0 g cg g o ,),./) ,)c.) ,.» c» -0 -0 V) V) = C.," 0 V) ,,,, ,,, V) V) ,D CI) Q Q (/) VD (/) Cf) V) r. . . . ) •'—' ,/, I I el— ) et)— MI— el— ) e —I ) 0 0 ,4z:) r . 0 ,z) > CL) 0 0 Cl') Ci4 0 C)1441 0 7:1 ›.-, t).1.) ca, :1,., ID.° .,..., ça, —4, cl., 0 „, bfl bl) 00 tA bA cl, cl, 0 ,._, cl„ 0 6 O (71 ,..cya) ,.. a)c) Cb t'Zi a) czt a) o g g g ,.9 ,.., u (D0 ct a) 0 -Z = , 4 773 , 4 7:3 - - '"C) eL:1 S' 'd = , (...)Qc...)c..),..c, ..oc...)c.)QQ(...),.9_, 5 zoozozo o ,_-,,=:.,==oo c 00 Ct Cd CZ$ C:$ 0 s-, u. 41 41 41 4.1 41 5 r 1 0 0 9 9 q = . 1!,' V) V) CA 6., :. V) (/) V) c, V) V) C 4 E 1.., . ,.., ct ,a) . 174 • Z-21 •ol • ;7', • ;"--,' -6:: -6-D • .6-=1 ('Q 0 u • 6-D "0 ' i--1 7-1 ' 0 ' 0 • s-, ct (i) u_., (1) 5,,0 -0 -0 -0 ,ID ,C gD ,SD ,SD (1.) › › aD ,-SD ,-CD ,ID -0 ,SD ,c2 0 1.) V ell CU 1..) 0.) CL.);-, V V V CL) CD CL) V , .,.., V Cl I:-. GO . 2 a 121 Do a4 OUI

• '5 '5' 'S 45 '5 7,; -9,c.)1.)_c ,c (L) • au 17.' -Z"), ('3' a) a.)17., ) t_) cij „ cu 7:1 C-) >-, Q, 1, 73 CI 1..) gu,i¢ ct E — cq ct 0 0 E o E ci E 2 2 2 o kA,A,9-, 7,g).— D C) C) 4 D 4 4 r- a4 C) 44

• a4 104 o4 (04 o4 c4 a4 o4 o4 o4 o4 p4

r- Ch cN Ch Cq 00 \ID \.° rn vn r- Ch kr) CD kr) kr) CD 00 CD CD CD CD -4 -4 -4 N N N cr; cri cri cri 4 kri kri kri .6 .6 06 N r— r— r— r— r- r— r- r— r- r— r- r- r- r- r— r— r- r— r— r- 0000

•

Cc)

• N cA cq rq rn „, cA

VD Ch Ch,. VD Ch 7:1" 7t" 71- 71- "0 (› .7r 71- 7D VD 71- 7t 71- 71- 71- 00 00 00 00 Q,, 00 00 Q 00 00 00 0000 00 i I 1 I c/ I I ct I I 1 I I Ce) Ce) Ce) on - on on on on on on on .(D r, r- tc) r- r, d- N r, r- no r- no r- r- ch r- ch In Ln r- ron- \ID Ch Ch 00 Ch Ch 00 CN Ch Ch \ID Ch \ID Ch Ch 00 Ch 00 00 00 Ch Ch Ch ,- ,-1 Ch ,, ,-1 Ch Ch ,-1 ,-1 Ch ,-4 Ch ,-t ,, Ch -, Ch Ch Ch ,-, ,, • N C‘1 ,-, N N ,-,,-, N N ,-IN ,-4 N N ,-, N ,-! "--, ,--, N N

0 0. C4 VI V) (/) , , c.4 CA (/) C/D CA (/') (I) (,) 0) 0) 0)0) 0)0) vi 0) 0) a) a) 0)0) k-.,,o bci t'A tO b-0 bk) s, ..+;-' bi) bf) bI) b1) bO b.0 g g ° g cgg cg 1-) g g g (9 .4 ..4 .4 ..t .4 m a .4 ..4 .4 .4 .4 .4 ou 0 (..) 00 73 7., •,,,,, 7, _,.., .,,,i 7.), ..r, .-ci ,-,::, cu - t-.) -,e) 7.) Tu" --r) a". ) = 4::' r4 = c'zsc) z m = z Q4 g 0 gq,„-8,9,z:* g qg,9 ..--, ...= ,.. , ,.. ..— ,.= ,= •—. ,. ,. ,_ ...,,zz" , o . 0 0 0.— 0 o c,),, 0 c..) c.) y ci) c) i ,,lj i ic A) ,.di, o 0 sn. aL, 0 -ci -,-(;) •,-7) j., 0 a) 0 4, 0 0 0 ,7) a..) c) 4 J 4 J '5 g ') '5 5 7= '5 '5 l! ja 5 5' ''' ''' -8 4- g ,,:, 4D ,2 ,2 ro :4.. 4:1 u., JD 42, JD 42, ,ID 42) . i:11 It ;7•:: (.-:;) 0 ,I • i::,' c) 0 u. (-. $.4 U-I 0 I. 4J 4J 0 4., > CD ., 1., CD i• CD 4., , A., I (-I (-I 1. ''1... 9 u.. ,.... ,;, ,...- ,- - - la ,t, - - .-0000,„,c)c).„C/D 2 o o o o l-, 3-, U. ctE1' 3-4 . c'Et' ;;a ;8 ‘ I c'l i-.n ,l. a ti.5 a t.), .„..,. at, ql.5 ”0 (1..) <1.) a) a.) a)0 » (j) ;>. > "F) LI) C )' .->- (1) I)› 2_ --- ci..) > ci.) > -c' » - '62 4 GO 4 2 a c7) 'a 64 - ' 1:-', C/D 62 c12 62 cl, 62 a 6

- --4 0)0) 0)0) Q --â.i (a I, a) C..) C-) a) ,__u Q 0 (t 0 a) 1.) -c) U 6 -0 c,.., 71-, -e) (....) 7,-, „ ,.. ›, U o ,- ". — •." CU g co C.) U ,.,11_t '4 (:.-)' r te -,-,, a." -- -5 ,es, ci ,.., = 0 (..) to (1.) ç..) a) .< ..t a) 5 5 ''' •. 1- <1.) 2 (1.) Ci 5' t...) Q czt L) = ' '" ' '(3 .' 8 u •E ct 0" ,', bc, to 0.) 5 .,..., c, = = 2 ct . .0o. .9.-, , .7.,, -, 7,,i ,_,0 a) o • b ,e 6" `,/,:' .:a a Q Q_ E PP P:1 P. H [..14 up PQ E.- d E-- PP Do Do a4 -.

o4 0014 ol4 3.4 0614 a4 o4 p,4 c4

00 00 ch kr) cl r- kr) r- Ch Cl on VD C QCV ) CN • 06 N cNi cf; tri c:; 3 06 o; 6 N N 00 00 tri 4::; a: -. 000000 Ch Ch Ch Ch Ch Ch Ch Ch Ch 0 CD CD CD 0 Cl

▪

nn 1 . 1 1-n.4 cA

71- 7t 7t 00 00 00 00 ,4D I 07) Cr) Cr) Cl CN CN CN CNCN CT\ CN CN r- CN ON r- r- 1,-.) 71- in 00 00 00 00 OC OC 00 00 00 ch ch 00 00 ch 00 n1D r- cc oc oc CN CN CN CN Cr) CN CN CN CN CN CN CN CN ON CN N N N N

c.,) cf) CD 7:1 ck ci] c4 O (ip (/) 1.) C) 0 bi) Z 0 • .-I • .--1 bA bA bA = (/) cn O ..0c'7 0 0 (9 q cci .0 c..) sa, la ,..c ,..c ,.z o 0 1) 00 U '715 2 "73 "C; cl.) o 71.1 00 3 3 O o) e, VDcip c4 I A 4 I) c/) c/D .17.i 4 i i (/) gz ..1 .1±, .477; .t5' .,.. 4-• • .--I I • v-I I A • ...-1 • •--I CA 0 I-0 CA I-0 0 CA (-) (-) CS' ) CI C.-) 0 CA Y ' (1..) a.) a P'-'1 a1•4 "Ci 77 0 77 ;--'-',n cllaL, al.(1) ;.-. 0 cl..) 00 O0 9a1 'li ='-' M a' :1 4I) a' çt -ci -8 cd ,,E,1 -8 0 0 0 0 ct _5 .. 7 c9 's -8 .1-d -7; 'a-' › -7; '5' z z c.„ = -c ...0' —•,- , 7-,.... ` --„•,1 a) a.) :6 ..c ,.c _c . .,_,--,,-, •-o c o.c. =c) 0 0 — r:1 0 b L-,, ,t-i ,sz . 0,0, . ---,_ -.-_ `"-r ,ô = g . , ,_, ;., , , ,.., . , 4•74 = n '-'' CA 1..) CA I L) I_ e ) . ,./, 0 0 cii ceo 8 .p., s, ,15•-- .,..4 ,_, , ;_, ,_. 7i • 4,_, 5 E .4 E „, v ...0 ...0 °t ,c cd c t ct , Q) 0 ,•1....?, 0..) ...0„., (L) ... 121 2 .- 4 121 — . 01 ''- 1 ..) -I) 4-6) n ZI - 2- b°' 11 ) . .: 4' ':2 :) I ' '' -4( ' I`21 n) a> 4 c 1 ) r1 ) r 2 - - v ) c47D 1 2 n -- 5

z ' = 5 0004 '' 5 ct -.5 77,; E >,c) 0 0 C-.) ,s1 0 cA c.) 72 >1 72 0 72 ,o ,sz J: ,— n . y , ,,d Z CD CD (1) 1--'—'n c9 9(7 +b >.' 41 4t—j A—, l'i ,—(121 4/7—n (...) ,Z ,2

'0 L.) U "Ci al "CS '0 ''t "Cl '-t 0 (..) C) 2 C.) C..) C.) Q) 0 (...) a) a) a) a) a) : (..) „, O -0 C.) -,-ICt ,'174-') ME .–- - g g cg g ' ',. ' g - 2, , ' 8 0 cl 'Et' ''a' ,474 Ç--) Ç-.) , , , , , , , , , z k , , : , u c -, ,) ..) - -8 -- °. ' ft 2 . , f 2, ."' . , 0- t O 8 8 = o = = = ,c_11 (NI sa, a o 2 , cc czt Q) Q) 4. V.., 4. -.; —4 00 DO P:1 cd0 W E--( E-1 E--(

p4 3.4 e4 a4 4 P4 o4 o4 o4 4 o4

cn r- r- CD CN Cl C,7 V7 lD CN V7 CD C) V7 CN CD 00 00 00 Cl Cl CV CV CV o6 c7 cr; Cl Cl Cl Cl Cl Cl (NI Cl Cl Cl Cl Cl Cl Cl cn cn cn rn cn rn rn cn

rr)

4, 44 1-, 4, 4, 1-1 4, 14, 4, 4, 4-, 1-1 4, 4-1 44 Cl 44 44 4, 4, 4, 44

71- 7t 71. 7f 71- 47=1 00 00 00 00 00 00 kr) 1 1 CIL) rn rn rn1 rn1 rn1 - M r, m CN r- oo ,, cr, -00Q0 c) m m m 71" (--.... r- r- r- a\ m m m ch oo 0 ch oo Ch 00--,---( Cq CN CN CN CN 00 c,.., ch ch ch cr) cr, r- cyn oo CN -, CN CN CN CN CN Ch n ,....4 ,--4 N CN CN CN CN CN -, -, -, -, CN CN C CN N ,...... 1 ,-.4 ,--1 ,--.1 ,-.1 ,--i ,.-q ,-.1 ,..... ,...1 ,--1 Cl N N Cl --I ,... ,--1 ,...... 1

o

3- O ,» CA (ID V) va = Q) a.) a) a) a.) -ui ti) 0.0 013 0k0 0.0 a) g g gg g c'7',:_ .= Q ,...) c.) c._., ,...) o a) -6 co co a) a.) Q " ci -ci -ci -Ti -7) 0 0 0 0 0 o o o-c o 73 o o o cl o o 0 o 0 g g g c9, . R 4 = ...c 0 0 0 . - c.) 4 4 = ,..c = ,..i 4 4 cn y 0 y y -g ,' c& cn ' , 0 0 0 0 •0 c, Q) '' ''' 2 , 7:) 000 g cJ-' E c,1 ,12),, -7,(c E'7§ . . . E '7 '7 '5 '5 5 *5 "c) s,„;- 7J

., . <,..) ci.) 4 , a),, Lu % :!sg_, g ,, .1 „2 , ,c" ,i' g, g .4 à fs' r ,L 9-. ,„ ' r 2- g g g c, g , • Q) a) 0 5 E (1) a.) jp 0 0 o jp ai) al:, ct ct tx0 to bi) 4-) t te et .0, * Li cd .01 cd " ",.0 > 4) a) az a.) "CDa.) 2 2 2 ct c7; a) a) lz; 6,75 '4 4 62 62 a4 00 DO

0 0 g = g c9> o ., (-) 0 „, (--) __ ,__ ,... .._ .,-- ,n ,c, >, cd Q, ct ,... ,... ,.. 44 - - c c; a) a ) a . ) a) 0 cu cu (I) CD 4') (I) CL) CL) CI) (1) ;-' '-' Ct C.) 2 4 't Q) a) a) cl-) al :-4 :-3 ;:j -1-,' 1-" " (....) '-' c:::), (:-_,) (LI (:), (1_-.) c.) c. (. (..)7,'1:3 =)-1,CD C-.) U n -' a) a.)p z 7,,i' '-.) ct -,-1 4-6 ct ,,EcDc-)

o4 o4 a4 o4 a4 o4 o4 o4 a4 3.4 o4 .4 .4 o4

n cg vn Ir) kr) CN oo oo 00 o0 o0 C) 00 71- 71- d' cr; N cn cr; NN C 00 71 .6 oc; cr: In tr) tr) tr) (Ir) Ir) \.0 r- N r- r- 4, 4, 4, 4, 4, 4, 4, 4-41 4, 441 4, 4, 4, 441 44 4, 444 44 44 -

C) ce)

N rn „. „. „.

7h 7h 7h 7h 7h 00 00 00 00 00 CIL) Cr> on on C11) o. o. o. Ch OC Ch CN CN • —4 4—, CN Ch Ch

c/ (,) (1.) V) V) V) CL) t:10 cll a...) cl) te biD b•O 0.0 c:1 Z Z s-, (7S ,4 (7( cz: ct cll ,4 c.> - 4= 4= cz, 0_L) c._) c.) --, cl..) (1.) -(5 "r.) "(:) a a = a V) C/) V) g2. g cVd- ,-= C) -= +-/3 4 - -. 4 bf) C .) I C . ) "L'," • c^7, (/) •• c./) ci) 1 C -) C .) • Dl = 000 j, 0 0 c") SZI, ça, SZI. d-, fa-, ."ti - 48 -,8) 4Lr's '5 j2' -5 -,.- -8 ô) -- - ir 8 3 3 3 :-R, ...,'..- -c 3 3 <1.," -c• — •-c — 3 > 00 0000 , , , o 5 , ,7,-,,„,_,Do ,.., i=0 (ID ,./, ci, 4_, (ID — •-,4.4 —,44, ,_, a.) ,..•-•- ,__ ,- , 44, E '-' ci ›ai ›cu 4)'-' ctu ai ai ai _4, Ç 2 12) (21 C4 — er .62 62. i;-,"

a '5 4a' 4J *5 '5 '5 o :40 49, g g 2 L) u u 77 • 7 77 7=3 (1) 77 77 77 , -7) CL) (1) CL) cL.) C.) CL) cl)

g g cg Cg g 1(.7 1) C = vis, z z , W cq

a4 P4 4 .4 iO4.

n 0.0 . 7h Lt) cq gn oo oo C) kr). CD. . C•i N 4 o6 y) c) c) c) c) c) c) c) c) --, cl c.) NNNN NNNNNNNN

ci) a) (71

U 1) (1.) 1.) V V a..) a.) V V up up u) •-n 0 1. ct bA bA bA bA bA bA bA bA 0A to bo bo a.) a) 73 7i 7i 7i 7:; 7i E L7s • -to • (ID • a 0 M (I) Q) (DPCIMMM==(:1:1 10; P:1 S:1, 4-14-1 4-1000000000V)C4 riD O (t (7, r;4 *z.; •CT CT • CT • CT 1) • CT' CT a„) ciD (71 czt czt .4,74 .174 0-) a\ E 4Z4Z4ZZZ4=== • • 0 — -to vl) 8 r:14 • .._ • •— 1-4 " 8 E bl) 7:4 -122 czt cn 0 cr) 0 up a 00 "17 ct e •- (,) • ) -) -6 r -5 V, old E • 4><' 1 r r.00 t i rin VS O • co') C'-) 0-1 ,--a) ,--a) c9- B a) a) - ct cc) c:-.), 10, a) 5 > cc o z o Ç2)a) Q)" :-.n 1-, a.) CL) C.) E tO bA fa. •). 77 77 ..." cb P4 IZ4 ct cb 1-=-( 0 •- 0 c„ z 0z 1 1 :D 1 1 1 :D (4 (4 :D 1 1 cd0 -to 7ri n f:u ct CLcy, 0 • C.) z ct Ct '7) ° • ° 1.) (L) c1) 0 E

cn r- CNNCC NO E V-) r- Cl Cq Cl 00 Cl CD CD CN rc) Cl tri 4 1/2) 4 06 CNi 1 >. • c,1 1-) 71- ca\

c.) .L1) , " c, -5 . 2 co, cr; • •

OC '1- c ::444 cr 0 CD •-• $44 a) a.) -E4, O © tjt) E c,).(t - 6 -.) ''', lr w ct 1 (1.) " ' 1-' M V . I-) c-) (..) c) ct t---: 0 1.* 7I--71 d E 0;4 (:; .> -ko •-9.) - cv,„ a.„, ,„ ›. a.) ,__cl ,_,b-0 fa., i ›u 0. › o `-',., ct I tn cl-) ct ," 0 to CI) • c•z-,) II U I I ,... ,..'.>Cf) I 1 C/D 00 ,..., 77 ct = czi M • '-' C..) (A) • a .! 0tki 7:5 0 ;.T4 g ct 0 a) a.) ,,,c1.) CD 5 . . 0 ,A • C/D CD E-i '•Z5 g E cl cl o E -tt o N ct cdn cl (NI c,1 0-1 0.) 0-) rf In kr) r- r- oo -. --, -.

•

\ID 4a.' 0 0 In. CA Cip CA CA CA CA cn c.n cn c.» cip cA cn (ID cn ,. ,., ,= ,.c = b.0 t- .1.1) c:1A e),0 b.() bki b.() b.0 bl.1 bl) bi) toi) b.0 01) bh c.)) (n c)) c./) 0 5, = 5, = 5, 5 F. = E 5_. , = cd ccl czr c, ›, .s. .1:1 .‘-, .u-:: • - s--, i:".., • 1--, .1--, • ;:2, .r.7 . ;11 *r-," • s.-, • ;..., • i'.". z fa, cl., P.., cl., cl., CI., cl., cl., In., Ca. ca., CI, cl., CIA la, cz: 00 DO 00 00 00 DO 00 00 DO DO 00 00 00 00 00 It tt It tt c.) r_, 1_4 1_4 u., ;.- 1 4 ;- ,.., ;_, ;_, ;- ;_, ;_, I. ;_, 0 0 p, 0 ,. I L.) 4,.-2 1) !.-2 L) 2 ..E .3 1) ,2 _,) •.1) .!-2, .',?.) , .2-_), 1) I I E 0 -0

t Cf) I I 1 rn I I I I 1 7t

I ^."' I I I Cl I / I NCf.) I I I I Le)

..- o c.) = ---' 0 ct o4 ,-- 0.) 0 a),- v) ...t 0 0 0 , 00 , , o4 , , , 4

Cn 01 00 V") - Cl tr) Cn Cl Cl 00 00 Cl C— \I)n Cl 71 ,(:) op m V-) In 71" kr) 00 1,-) Tn CD 00 ,t rn r- cT CT NNClNN cc CICr; N oo 1-1

i.-n (1) ,.,

,-( czt ' N-"' tC/C3 O O t. "CD 0 (/) ct ›, O LI z .5. QC) a) a) cl 1 5 1:- " ' ,-, 4 (24 L) L)ii (t a.) .' -o O ,_, cc.T 1-7 c7sO gct § 0.; cd, o , H , , Z , , .-.. , , 4

•e

4.1 f4 o4 c4

00 ce; oo ce) ce; ce; oo 0 cy, E—c eNi ce; ce; c,-; tr; v3 r--: oc; o: a: a: o: n n•n•1 1••n1 71 1n1 ...... 1 1 11 /1 vn••I 1n•••1 Cl EEEE EEEEEEEEEEEEEE 0000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 re) z z z "5 "5 ô ô ô 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0›-„ cg g g g cg g gggggg cg g cricri cg u c_) c...) c_) c.) Eitttt 4=1 g'Z' 4:1 4= g= g=, 4=1 g= 4= 4.7. g= ,4 4=1 0000 00 0 0 0 0 00 0 00000 f.."zzzz ZZ4ZZZZZ4ZZZZ4

N CA 1 1 1 1 1 I I Cc) Ce) i CA cq

te) le) I I le) 1.(-) 71- AD AD li") te, 1.(1) 111") I

Cl) CA 00 0.4 --0 (-) 00 C1.) If.) ICI = cll (1.) 0-1 f:11 ---. 7-1 Cl) Cr) g g Ln 1: (24 (24 "1i/ li3 M 771" 4 4 VI) cd cd NN D 1 1..i 1.-1 CA 1 CA CA Cl CA C..) C..) 1

7t Ch AD Ch r- tr) (1-- CA (r) CA CA Ir) CA AD --, 71" r- cD on cN ti") CA CD oo c9 oo r- cq 0 r- cy

6 6 Cc; 6 Cc) 0 0 6 -4 6 6 n1::, c•i 6 -4 6 CA

;14 p4 o4 p4 o4

7t . . . le). 00. cq Ch CA Cn tr) CD CA CA 7t r- 0 en m c‘i E* rr.-; cri 4 4 4 4 4 Vs; kr) tr) kr; CA CA Cl CA CA CA Qn cq CA CA CA CA CA CA Ci CA CA CA CA Cl CA rra' 4E 3g rg #”4:0. ô 'ô '0 •0•6 'ô *0 •0 *0 aaaa a. a., a., a. a., a. a. a., ,„ „, „,

000000 0000000000000

c9c9a9c9ggqg qgg 090000 4`-'2 UUUUUUUUUUUUU 4 4 4'czt 4 4 4 4 4 4 4 ,e 4 ,4 4 4 Ci 0000000000000,4E4,,i,47i m zz 444ZZZZ4Z44 HHHHH

CA CA I I I Cl I I I I

'Tt 'Tt t---1 I I CA I I I

-Z

,5,9A „ ,5 „ „

ch ch CN CA CA d- Cl kr) cr) CN k(') 00 CA CN CN 7t cD op tr) cA CN CD --I --I CN kr) CD 00. d-. CD. CA. CA. . CD...... CD. . . . . r-. CD cf') CD CD AD AD CD CD cc) rn CD cD kr) CA 00 CN 1—n1

o4 .4 o4 o4 .4 .4 .4 .4 o4 o4 o4 p4 o4 o4 o4 .4

AD 00 CA CA CA CA CA kr) CD AD 00 CD AD 00 CA CN CA kr; \C> nCS oc> 6 -4 -4 -4 4 4 4 kri cl cv cq cq cl cq rn cn cn rn rn rn cn rn (ln rn rn cn rn rn

CCS ("4") • aaaa a' a a 'a' a' a 'a ±) • O F5E56 605 ° (-""-"oE56c:)o- a. Jai (21. sa., a. a. ggggqa, a., a, 0.4 a. a; O 0 0 00000) 4 CA Ci) C.")0000a4go4a4g (A 4 J cf) 4c/3 d 4c Ei c cl c' c— 7I c''cJ0 ' g4 Ç'71" "g a 4-, 4-, 4-4 4-4 4-1 .1.nn •...44 tEi C4.Eictct:Ictccl g 0 Ocztcztc1CztctCtct0 t•-n HHHHHHH (20 GP a:1 H HHHHHHZ

"45-•

0

7:3

1a. C'n rz4 0

0 a) a iia) 0 •,-.1 C/D a.) CI:) a4 a4 a4 P4 a4 2 a4 (.1) , , 1 1 t a., , n , 4-1o V) o

-. 0 0 kr) cp. 0, -. kr) -. rn kon VD r- 0 VD VD --I .i7 0 0 —4 v:D VD C•1 VD 71- CI') Cr) C-- CD (r) VD cs.. ,r r- ,r -.. oo. rn. . c,i. r-. -...... 42, (-,i 0 C•1 0 ,--, 0 4-4 ^-' 4-. ,--, 0 CD CD n4D N --I 0 --I ---( GN 0 N C cn tin cl N C.) ; 7*,

t\13 o cn , o c.: -a) un 4zc.) > = rn czs o rl 0 c.) (..> \ ..., >, W .Y q o EL' kr J1 -cs-on .-t a) (-) 1.) t,o) 4-1 0

• g o4 o4 0,4 p4 .4 o4 g o4

121 Q.) ,

. . (T. \tD 00 00 00 C14 o r- 71- C.-. VD CD 0 cn 0 cl r- 71" ta., 0 VD \D r- oo cn cn cn re; ,t kr; < con cn rn cn rn rn rn cn ,t ,t ,f ,t ,t ,t ,f 7f* 'cf 41- 71- 7t 71-

<71 ct TZt ( cc Ti CC: CZ$ CC) CZ) CI CC) [l) CT) Cl) Ct C/) C/) c./) Cl) C/7 CI) CID C/) CID C4 Cl) C./) CT) <1.) (1.) I.) C.) (1.) (1.) 1.) (1.) CI) CI) (1.) cl)

s:lis:L.C1-iszL“:1“:14C1..sa4C:14aL,fa..sa..s:a4ta“:1•CL•raal.tz c c czt 4 r:). <1.) cl.) CL) cl) (1) LI) CI) I.) CD cU cU cU

r". U.-1 E

CT)

'7) tc.)3

• F:14 )

CN 00 r- \ID CA CN CA Tr) \D TT) \O Tn CSn CO CSn TT) \ID V-) Ch \ID CA rn cn rn rn cl 7t rn rn vn vn 7t cq cl r— cr)

Cri) N C:3666666671: rn C5 —4 co

1:4 f:4 P4 4) .4 1:4 1:4 rz4 ;z4

71- 00 00 V.) 771- \C) 00 Ch ^-1 Tr) 71" 00 cq r- tn -4 tr.) co cD

oeS (:); c5; cs; cr: 6 6 6 6 6 cr-; r,-; 4 4 ,t ,t ,t ,t ,t ,t 'I- 71- In in kr) tr) tn tr) tr) tn tn tr) C71 CZI czt re) C4 CID CA CA Cfl CA (4 CID CD Q.) a) a) a.) ci) V V a) a) a) a) a) a) a) a) a) a) a) malmmmmmmCI, CI( f:1-1 C1-1 f=1-1 B z z z z 3 3 3 3 3 3 3 3 75 75 75 75 75 75 73 75 75 75 75 75 75 75 0 0 0 0 0 0 0 0 C/) V) C/) C/) c/) C/) C/) C/) C/) C/) C/) C/) V) C/) gggggggg F,(i c,',:(? zzzzzzzzc...“....)uuuuuuuuc...)(...)uu9. 9. 9.

\,..0 1 1 1 I I 71- 1 1

I I I

0 0 11-) 00 Cq 0 Cr) 00 01 71- 71" '71- 01 kr) Ch CN Ch C) cq 7t ch r- 7t \ID 7t ch cim 4 6 6 6 c2; c=> c:; tr; cy c:3 6 6 6 6

C CC C o o = C m g m (..) c.) c..) .— ,,,ms ac a) C C t..(3 4Fi ""( 3 6 , , <4 , , .o ,

a.) ,•-• C:4 r:4 1c4 rz4 c4 1:4 C/)

7t 0 re) vD 00 Ch re) kr) r- oo kr) Do vD vD r— c\1 re) tr) vD tr; tr) v7; oc; oc; 06 csn 6 6 6 6 v.) tn In v.) Ln ../-) In In In tn kr) vD vD vD (1.) ci.) a) ci.) 13.) ID a..) a) a..) y (1) a.) a.) 0 y a) 1..) 0 a) a) .1: 7:) 71 az==zz7:1 7:) 771 7i "0 71 77 77 7, 71 7:1 "ci 771 77, 77 7:) .-.t4 .- .4...-4 , .-..,...- .47.1 .-74 .474 .4;2, .-,t-4 .4.,1 .-,t, .-.-2, .-4,,, . .4.,4 .-,.,-2, .1:-2. 75 o o o o o o o o o o o o o o o o o o o oo cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn v) cn cn Q) 1.) 1.) a.) 0 ci.) 1.) a) cu y y a) y a.) a) a.) y y 1.) a.) 0., fa, fa. fa, :1.1 fa. fa, ra, fa, CI, ra, fa, fa. CI. CI. C4 C:)., Sa., fa. CI. MI Ct Ct Ct CZt Ct ct ct CCS ct CCS CI Ct CCS ct CZI Ct CZt CZS ct (..) Q.) Q.) Q.) Q.) Q.) Q.) Q.) Q.) Q.) C.) C) C) C.) C.) (..) C.) C) C.) C)

7i bA

-6-n ct '174 r=4

4 P4 D I I I D

n(:) (7n r- r— CN CN nSD D CD vo r- ro CD CD nID cq cn \ID N (0) Cr) rn rn 6 CD 6 6 6 CD 6 --, 6 6 6 6 6 6 CD 6 CD 6

tr)

1:24

"Cf c4 wi c4 a4 a4 o4 a4 4 p4 o4 o4 C/1

E-n

•et CN N vD cn cf) 1r) In oo oo CD in r:D cr) cNi cn. ce; 4 4 4 tr c4o nCD nID \CD Z) n.CD n(:)

Cc) cc)

7.0 .$1, .0 .$4 . c,14 Q) .$_4 .a.,) Q) ›,.$.4 • $_,) -5--5»">>>>»»»» 0»

(1.) u a) a) a) a) cu a) a) a) a) a) a) a) 1.) cu a) a) a) a) sz)L., cf) cd) cn Cl) Cl) Cl) ci) cn cr) cn cci cZI (L) (1.) I) a) (1) 0.) cll 1) I.) a) a) a) a) a) czi (-) r.21 121 (21 12( Q.)

4-11-.0 11 t i lDit t il I I 00 t

•71- d' t

z g 124 I I I 1 11n,' 1 I 1 I I 1

VD r- cc) CD Cl 00 cc) In In VD 7:1- 00 In in -0 VD VD Cl VD 11-.) GD r-. . . r-. rl. cl. vD. . -0. CD. r-. 00. VD. oo. 71-...... kr) CD CD Ch CD 7t CD CD cn cl cn le) cn

a4 a4 a4 n-1 a4

tn cc) cc) 00 Cl VD 00 CD tc) 00 n1D CD cn r— Ch Cl VD cc) CT 1.0 n6, NN CC O c‘i c-; cri lD nC) VD VD VD VD nCD VD VD f.") lD r- r- r- r- r- r- r-- r- r--

71- C c.)

>, >, 771 4 p4 0 0 0 0 0 0 0 0 o 000 o 0 o 0 o o E E a4 124 o4 cz4 co4 to4 124 o4 o o CL) 0..) CL) CL) Cll 0.) CL) 1.) a) a) a) a) a) a) a) a) m r=1, f=1.4 fa. 04 0404040404040404CD. ta4 Cm. SD. SD). SD. fa, cd C'd C'd C'd M m al al al m C.) C.) C.) C.) C) C.) C.) C.) (...) C.) L.) C.) C) C..) ga.

cc V') V") I I I I \ID \CD

VD VD CN ON

c.)4' OA a.) a) M = c.,) c.0 0 › a.) a.) "d a) ==--, 00 cu cdCD. m fa. C 4 a 4 a› . ) a› . )P —Z1 m9 c9 o4 1:4 8 lo4 a4 1 Z Z Z co0 1 1 II f c.:5 c.5'

ON VD r- r- cq CN 00NNC0N- o . cr) . r-. . 71-. ",:r. tr). . oc. _0 o CD 71- CD VD CV re; 06 ce; 6 re; (-Ni cei of)

a) 71. a) ..- ,-4 N = ° I--. a) a) 0 ›' a) a) ,-.4 t-- ›, d C-) - 2 a) a) "d 0) Ç..) (...3' CI.) 0 L.T4 c) cp 0 0 C.:3' - a) 1-4 C..) v) 2 0 c..' 1 ia) 1..) 3 — g ..- o ct bi)C._) = = › 13 i '?)3 . A & Q , mira,-c) .,_ v, ,,, i.) ,i . 5, -F, cq , ai , r- , c4 , : < = 1 , (..) › QD ,

t:4 f:4

-eC14-) c.ir) 0 ,-; 0 N N N v> tr; nr.:; vD VD oc; (xi cs: r-- r-- 00 00 00 00 00

kr) n•••n1 Cf.) -0 -0 -0 •.= .0 -0 .0 .0 -0 .0 .0 -0 -0 0000000 ocic)c..)uoc)c.)c.)cil)c)c.)c.)0 o4 o4 o4 o4 a4 o4 a4 a4 a4 a4 a4 a4 a4 c4 o4 o4 a4 o4 o4 a4 o4 EEEEEEEEEEEEEEEEEEEEEL) 0 0 00002),02: 02 0 0 os.22 0 0 0 0 0,-,) z z O ct cts cci c.cs cnt -0 ,0-.0 -.0 -0 -0 -0 -0 ,•0 -0 -0 -0 -0 -0 aL4 p. p. p. cL4 a,

1 cA 1 1 CA i II I cn

71- VD M t I I 71- 71- 7h I 00

,—, v.) Cf a) tu) tu) ,._ b.° 0 0 a) • -, • ,•••• 0 ,-• ;.-., ;_, a) cn :—.a) --CL)n P-, al., (....) ct f.-d C4 C/) 0 a4 tLy3 cu a.) E m ôtr, ,r) , • .E-i-'.,er'124 0 , oo , i N oo oo ).—, , , i 4:1 i a. PL.,

r- CN 00 nn CA CD AD '7t r- r- CN CA I.r) cn CN r- m oo -.. . m . CD. AD. AD. . AD. . nn. CN. CD. . r-. . . m . oo. m . cl. ch. cq. ". -. cq m C2) C.7.) CN1 Cq CC) C) C) CI CD 7h CS) AD CA

0 O 0 -t4a) O >.... a) ...a) O a.) g ,... ct (:-.3 U a) U O (3' cp 1-• 0 1, a) ‘-• 4,-•cu .-• a) P.• C.) v) 4a:1 -c) Ct 0 O Fl , UN i t L) I i

p4 .4 p4 o4 o4 o4 p4 p4 .4 4 p4 o4

CA M 00 n.--4 AD CN tr) AD CD CD AD r- r- ch cq oo CN CN CN cn CA cr; cr.; 4 4 4 tr; tr; oc; oc; cs; 6 oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo oo 00 00 CN • • q

zz zz 5 5 5 5 5 5 o o oc00000000 oui) (D 000000 ›, ›, >, ›, ›, ›, ›, ç=1, r="( 0. al. al. ai. c9c9c9ggc9cgqqgggEEE UUUUUUUUUUUUUI"uur H H H 0 0 0 70 -0 "0 70 10 '10 "0 "ti 'd '0 -0 czt (c/t) ct al a3 ai a3

6 6 5 6' 6' 5 6 6' 6' 61. 'e:g E'. ;g

1 t CA i O o i '71- iO \ID cr. i I 1 CA CA

C 7t I Ch Ch I Ch 7r \.C) cr. cD (1.)

C.) ci) (1) 1")) 4 4 a) a) Q) a) (.!.34 E'E mct (1).E = 0 h, bi3 C't C74 ,Cri c.:5 I 6 1 L) C..) H P> DO GO

Ch V-) Cl Cl cf) r- oo 7t oo CN CA 71- Cq Cn n1D 00 CA n.1D C\11 CN1 00 CN oo 00 . . CD...... v-). . . . oo. r-...... Ch. ce). C-. CD. ln CD .. on CD Ch CA CD CN ce) CA cr) CA CN CD 7t 7t CNI N on „, „, tr,

0 Q) >, 0 Q) ,... "aa.) ,.. -t_g (i) - a) Cuj (1) (1) 0 u4,1) (1) I-) C-) c u xux • ...)( L) 2 ..s4 5 û'1,--,)1)j 0 ,-.0 oU cu u uc...).,5 .... (...)Uu a, -, E L) _L) .6) = , .,. 0 Ig...› t ,,O .2a'a m c,., ,...„ 'c-.4 445 0 ,,L,-= 0, E. , up , , 0, .. H 1 Pr4 , Do 1/4..., H , ,

p4 o4 p4 .4 .4 0.4 p4 .4 .4 o4 .4 p4 .4 p4 4 .4 o4 .4 .4 o4

\.0 on on -. ken c) on V-) cn on c \-0 c0 71- Cl cl on nID CNi Cr; cc; 71: tr; tr; 00 00 01 01 01 (0 C) Ch Ch CN CN CN CN CN Ch Ch CN Ch CN Ch Ch Ch C.\ CN CN Ch

EEEEE .EEEEEEEEEEEEEEEEE O 000 0 0 0 0 0 0 000 0 0 0 00 0 0 0 0 P. aaaaa a. al. a a P. P. PL., C. aaaa a, a P. P. C-C1 71 'c78 'c C7i "Fd aaaa cl, a. aL, a. al. cl, a. cl, a. aL, ra. a. cl, cl. aL, a azazaa aaaaaaaaa aa V) (/) C/) C/) ci) C./) (/) C/) V) V) V) Cl) C/) C/) V) C/) CI) C/) C/) 1/) C/D M >Mt Mt > Mt Mt Mt CM M > Mt Mt M Mi Mi Mi CM Cd Cd Cd Cd Ct Cd Cd Cd Cd Cd Cd Ct Cd Ct Ct Cd Cd Cd Cd Cd Cd Cd Cd Ct ti ti

4-1 1...4 1 ,r I 7t (1.-) ,

• 7t I I 0 \C) 1 1 1 7t 7t , ,

a.) ci..) Q) vp ,..) a..) (1..) 0 0) O 0 = = = = 004 cr cr I 1 , c/c c/c l' a4 g:4 Ki 'A ... a , HP-1 , , 3, Do 0:1 P:1 GO 1 -.

n rt CN Cq 7t OC Cq r- tr") cl cn C) 00 m oo r- tc) VD vD n10 cl lD. . cl. Cl. . r-. 7r. . . . c., cl,. . cq. . . r- 4 Cq 0 71- r- cq --I CD m Cq r- r- CD --I 6 ;-. Q)

,aQ) a . .Ela, o O >-, 0 o a o --„, O 0 C s o Q...) 0a ,.. ,.... cu ›, o ...0 o L) 0 cl) 1-1 (-) ct czt = .,--,= = `..-, 0 (....) li ct ct c-, ‘..., (...) ^ cf -'-'O C...) -' = a.) -2 J. 0 (1) = = C1.1 gz ci) , 0 al"' cu),D '5 a E a.) c0 ',:a 0 , , c/..) 1 = U0 1 1 L)

• 1:4 (Da rz4

0 CD CN Cln cn V.D CN r- r- ro v:) 00 nID VD 00 cl rq rn rn 4 4 4 4 kri v3 v3 06 6 6 o c. c) c) o c) c) c) c) - 1.-1 4-.4-1 1-1 e•-..1 1-1 11 /•n-1 1nn1 /-1 00 Cr)

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

VD V) V) V) (I) V) (I) V) V) V) V) V) V) V) VD eD V) vD V) V) V) VD 3-1 -1-n cl.) CL) a) a) a) el.) '1-1 ..;-+ n a)a)a)a)a)a)a)a)a)a)a)ua) I. 1•-• I. 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Sal la fa SI.Sai so, la fa, fa. 1a4 fa fa Sal Sa. XXX X XXX X X XXX XX 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 [4 V4 V4 44 44

, i i r- r- Cl CA AD 1

7J CI) kr, •5, Thici o4 O4 f:4 to4 „ P4 I I ,

oo oo NOONr— m m r— cq m r— m m r- r- CD AD CA 00 00 71- AD V") Cq CT m AD V") re) 00 CN CD AD CD 00 00 AD • • • • • • • • • • • • • • • -• 6 CN 6 6 77 c) CD CD CD V") 7h 7h CD CD cq

OO NOO NN a.) c; 6 c\i c‘i cri 7t: tri tri v-; v:3 r-: 1-1

..... 1-1

-

a a a aa a a ce) ccccccczcz=c0000aaaaa o o o o o o o o o o o o .—•

(1) (/) V) V) (/) V) V) V) V) V) (1) V) „ ›-1 14 14 "S. "41 4 4 4 4i a) a) a) a) a) a) a) a) a) a) a> CL) 0 0 0 0 4 4 4, 4 4. 4 4 4 4 I-. 4, 4 w 44 w 000000 000000 0 ° 0 ° —' —' —'"' 0 0 0 0 LID . . LIDLID a a a a a a a a a ta.4 1:1-n CLI CI, f:Li ct) Y< Y< Y< Y< i< Y< Y< Y< k k 0 0 0 0 0 0 0 0 WWWWWWWWWWWW E-4 E-1 E-1 ü. ü, P. ü.

Cq I I rn 1 i .I I I 1 1

ai) 74 - a.) ct (-1 8 _H:4 , , -.

N cl") N \.C) tr) C::C CT) d' N •-•-n ONN7I- N nD oc> c--: ch c) cf-) 7t. 71- N c:; c; c; oc; 6 6 6; c:;

Ot

Ir.: ,L4 - a) 1-..( 0 a) ri CD >, a.> 0 = >, W-1 e3 ct e.-2) (....) a) ct a) --, U • —•itli CEi s. ,.':- a) c) V) ci..\ Cd N 4 C-1 0

: 1 I I 4.4 i 1 fEE1 1 1 .. I 1 T-1 1 I

C..)

C/1

c) r— (---; N ce) tr) r-- 11-) ceS c6 06 od an cs; a: 6 6 6 rri rr> 1-1 4-1 r-n 4.1 C`i NNNNN CNI CNINNNN

cr)

cd cd Cd tll (1) (1) a) a) a) tl.) (1..) ›, >") ct c7S ct C-F'3 c'd cc:1 Ct (CS PP PP PP po po po po al. a. o. a. a. a. a. a. a. a. a. a. •-•,-.-•-•- (1.) el.) a) a.) 1.) 1.) v) Ci)) 'c ci)) cv4) 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 LT

t nC) cq t i rn rn r I vn \.0 1 (>0 -- c) ..-. o a) c24 a) a) .a) c/D ,-. .L4o 7.1 '>--. .e) a) 0 -7) i r.L. r-: , o co 1.) "7=) ,t cq cl ix c11., c.i (i) , , -, ,- , , .-... , --I Do , PP

n r-- VD Ch Ch cc) Ch CD r– Ch tr) ID 00 In 00 cr; tr) CD 71- CPS C) ci.) C.1 71- If) ("A C) r– Ch r- c7. CA 4 --, CD Cr; 4 CD CD 4 6 CD 6 CD nID 6 6 6 —4 6 od -2 rn

:.‘z 0 uo :,. tri o ,... = ,7s a) 0 u cl a) E a) O 1-4 a) ID 1. ..-, o a) a) ,1-1 A (...) „9 c‘...-') 7;4 cçl• .4 0 ,24 C) ;€4 (...) .0 v6,-. ••--, ;-'1) '4 ' 8 C--": 0,,:) e, A ,..,,,,. N " . oc)- " .... . , , - , , , , c2"7" 44 .4

Ç.T.1 o4 o4 o4 o4 o4 c4 o4 a4 o4 o4

E-1 :4 V') VD C>C) fr) VD N. CT ch N ten .r) CP\ kr) CD CD vn NNNNN 06 dd CACACA E*NCArlr•INCINNNNCICICACACAcncr) cA —I 1-4 .... 1-1 1-1 C"rt

C C C C z 0 0 C .0 .0 C c c d al cci cCt CC; Cl f:tt,r2t.,Ç:I.,;:14(:)., cdCCS CCI Ct CZICt Ct czCL)t c 6 6 P4 04 0. O4 P. P4 0., al., al. p. (.• c.4 v) up

75 75 75 75 13 15 -8 75 75 75 4E 1' 4&71 3 3 3 3 3 3 3 3 3 3 '4 a 8. a) :a' 0000000000,dmmcd . . . 4 . . . Ja., a. ao a. a. a-, a. HHHH LT. 4. 4. LT.

1 In t I I 14n

t 00 00 I vn

t ct 00 a) 0.) -0 -0 c.a :- C I.) 0..) t..0 ...0 a4 cç,t 71- P4 tr) P4 1 C2I „ , E-I —c' ,c22 „ „

r, vD m rq rn 71- 00 cfn 71- n1D Cq C) nn r— r— o0 — C) Cq r— 00 V.D rn on .. r— ,../D 00 cq 0 cp ,t r

6 6 .4:5 6 ,r; 4 (5 c5 4 c.5- , 4. 71-

a) a) t, a) a) U e....) ci)co Cll C.) (1..) ci) (13 .rnI al ej I a) .4 ;... rn c.1., ci.) rn C O a) 1 --1 1 H I CCI 1 I

t"-- CT cc) tn 0 0 71" 00 N Crl 00 71- N tr) N 00 ON cr; c,-; 71- 4 4 kr; tr; nc:) ocn rn ocncfn 1-1 1-1 e-1 1-n1 1n-1 1.n4 rr)

Cl MI Ct Cl Cl Cl Cl Cl Cl cf)Cl Cl ci) Cl __ c/D cf) VC c/) C/) ci) c/) VC cf) c/D c/) cf) CA V) CI) 'E Z 4. "--' -- (1') (1) 1-) • • • Cl cd cd cd cd 1 ) ±- ) c1 ) ±D 1 ) 1 ) D± ±j 5 5 5 Ç.44-, 4-, 4-,

• '''',,-Ei— ',7 7,7d- 74 7g '''-g -cm, CD ,) Z,-?',."'ZZ*Z.'4"a'Z ,,, c„ c.,) ,,, . ,,, c„ ,,,,, c,„ ..tzc„ (.„ (,)z OE3 ac82VVVVV ct ct <,.$ ct ct m ct

4. w 4. 4. 4. 4. 4. 4. 4. 4.-. *4 '4

\4) 71 1 \C) j CA 1 1 CA cq

cts a) Q) ra •—4 a) .—n cd n IA IA 1 Vs') 1,„, rs: 06 = .--( Cc) Cr) ,/, I:4 71- # 4 . , , . , 4 1 , -. , I ,-.1

Cr) CD 7t Cr) V:D n1) CI\ 00 00 N <71- Ch V-) <1- CA 00 V-) CA CA n1) tr) 00 C) 71- CP1 00 00 t"-- r‘i cN 6 6 N oo

)—) n-• ;•14 r:4

00 C f 71- In In 0 \ 014 CN re) CA 00 C) '.0 CN tr) t"-:- 06 06 06 06 01 (71 CT 0\ ,-2( c,r) 4 4 V"; oc; oc; Cr)C-c)(nCnCr)CnCnCc)Cc) El 71" 71- 7Cf. 7t 71" 7f 71" 71" 71- e-4 •—•n1 1-1 r••••I C./)

(I) C/) C/) C/) CI) CI) C/) cz1 czi cct ct 4-1 4-1 4-, 4-1 4-1 4-1 a a a a a a cA CI) CI) Cl) Cl) 0 0 0 0 0 0 CCS Cd cd ci cc3 cd la, a., p., cd cd cd cd cd cd cd on on Do on on Do

If)

00 1 1 7t 1 1 1 1 1 1 1 C'e) C',11

771

<3.) io4 c4a4 rVFt' g

CD CD cl GN -r CD Cl nn oo r- 00 ch rr 00 (-1 ,c) - CD CD 00 cl cn --I cl ON V) rr CD • • r-• cl• vn• • cA• 4• 7f• • -r• . . . • • • • • • vn cl vn m ve) oo vn r- Cl CD cl CD CD CD cl r- CD oo 71-

g Q 73 a4 p4 .4 .4 .4 .4 .4 o4 o4 o4 .4 .4 .4 p6, o4 .4 a4 .4 o4 CiD E* n r- cn vn oo CN V3 vD Cl cl vn k.0 ;4 oo ID on oo vn 4 CD 06 ch c): "It cn cn 4 v3 ch cci oei cc') cr-; Lr; n 71" le) In tr) tr) V") 11") tr.) 11-) tr) 11-) lD VD .0 V3

g a 0 0 0 vp c/ cn (I) a) cA 11) a) 1) ri (1) I) a) I) r1) (L) O O O O O g ct " 0 .F4 0 0 0 0 0 0 0 0 0 0 0 C.)(..)U c7$ ct cd rd czteJ- z 4 a4 a4 a4 P4 (:4 4 P-4 E-4 P-1 CD (1) Q) >, ›, (/) >, V) V) V) V) V) C/) V) V) V) (/) V) CC; Ct Cd cd ct (.9 3c 5 5 5 a..) a) a) a) a) 0,2ouc...)c.)c.)c)c)uuc.)c.)c)c) ',15h-3 ,.-6camcrictmmacrsaaaaaaaaaaa C7 C.7 CD C.7 C.7 > > > > > > > > > > >

t crs f'n

rn I I I rn rn

O a) cri m (.7 3 3<1.) ,14Ei /24 a4 a4 124 CD 0 1 1 1 :D 1 :D 1 :D :D

oo r— oo ‹) Ch r— CD cn oc CD Cl cl cl CD CD oc) cl rn cl 71- CD vn oc c> 7h vn cl oo CD 7h oo r— oc CD lD cl oc) Cl r— v6 cri (----: 4 6 ce; -4 6 .r3 -4 cNi ce; ir; cs; ,4 N vn

O o O a o 00 m a ›-.. 00 o O c—) O 000 cu a ct cr; o (ID C.) -., a C..) (...) o L.) a)aa) ›,,.- ct Z ., m › c.) o C.7 (....) a) a) C...) a) E ,.m -o o a) 0 cd 1 -T.-• , , C/) 4 , ;1:1 , C) , , c.n , ,

,c4 a4 1:4 p4 ç:4

cl 0 cn oo N vn. I"- C> 0 71- 71- 7t. N 7t. 06 06 6 (,r; 4 4 06 o; o; n.o v:) r-- r— r— r— r— r— e•-• 1-.4 1-1 1.-•1 444--1 1-1 14.-1 1-1 1-1 1.4.-1 4.4-1 4-.1 1-1 •-1 1-4 •-•.1 •-1 4-1 ••••.1

V) V) V) L4 V) (/) LID (ID (/) c1) c/) LID (/) (1) LID V) (1) (1) (L) >>>>>

re) r i t i r cr)

00 1 1

n n n cr) r- r- 00 N 0 0 0 00 .0 r-n tr) N c-N1 CN rn C) oo C) r- tn r- N (o cr) c 71- C) —0 ON (Zn C- 71- N ...... --100N0cnr--(000—irnr--INcnr--0N(n-1 e-1 CNI

124 rz4 (24 in4 a4 a4 1-4

oc oo CN 00 1.r) r- CD nD re) --I Cl 0 0 71- -7 if) LC) • • • • • • • • • • • • • • • • • • • • • C) -7 Cl Cl cc) cc) •71- 71- 71- 71- (C) n1") r- r- r- r- oc oc CN N 00 0000000000000000oo o0 oo o0 oo oo oo cc oo oo 00 00 00 0000 00 oo 00 00 1-1 1n-.1 7-1 17, 1--1 "-I 7-1 77 /-4 /-4 4) 44 44 44 44 41 4) 44 4) 44 44 44 44 oo oo GO GO GO DO Co0 GO GO GO DO DO GO UP Q) uP C4 up up uP up UP uP CD (1.) 1) CO Q.) CO 0.) •••-, ••-1 •••-I ••-n ••••-n ••-I • • C a = c = 0 0 0 0 0 0 0 0 0 0 0 66 ° ° ° 0 ° 0 ° O• 0-4 O• O• A. a) In. la) 1.) (I) a.) CL) a.) a.) a) a..) 444444400,004..1 4.1 I. ;nI

Cl In 71- CD CN 7t CD r, ch r— c ,t ch ,t cq oo cq CA 00 00 CN CN n4D 7t tn n4D r— 00 n..4D

6 71: 6 4 ,4 Z). -4 c3 c\i oci, -4 6 6 6 rc-; 06 6 6 (-4

4 0 --. a)Q g (.1...5

•—• 4 CD CD — I I I I •••-

p4 p4 p4 ;4 p4 p4 p4 o4 p4 p4

N re) 00 00 rq 00 00 r— CD Cq cc) CD \.0 r- cn 6 6 -4 -4 -4 4 711 4 4 tr. tr) J.:3 Z). \X.3 00 Cr\ CP\ ON CN CN CN CN CN CN CN CN CN CN CN CN CN GIN CN CN CN —4

44 44 44 44 44 44 44 44 LL4 44 44 44 44 44 44 44 44 44 44 44 44 44 GO GO GO GO GO GO GO GO U0 GO Cd GO CO GO GO GO GO LO GO GO cdo GO

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 -a 'a'-a ''a 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a. a. Pk Pk at. a, a. a. a. a, a. P. a. a. a, P.., Ok a. a. a. (1..) (1)(1),:1) 1) 1...)(lla.)00(1)(1)0001)(1..)1..)

B EEE B EE5EEEEEEB5EEE B E • .4...E:2g• 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

I 1 1

c,) nn oo r- 71- 71- \C) V-) (0\ 00 Cq• Cq 00 0 Ch M Ch oo r- 14") CN 71: CP\ kr) cq r- 71- cn 6 6 6 6 6 6 6 6 6 c; 6 tr) Ch

r4 P4 P4 P4 P4 P4 P4 P4

CD CD In tr) 00 00 tr) C) cn Ch CD ,--I 7t In CD CD C) Cq m cn

• od o6 od oc; o6 6 6 6 -4 c\i cNi (-1 -; 4 4 4 CN Ch Ch CN CN CN CN CD CD CD CD C) CD CD CD 0 CD CD CD CD CD CD •—n CA Cq Cl CA Cg Cq Cl Cl Cq C\1 Cl CA CA

00 P4 DO cc) '5 —

- 4 rn V. a. a. a. a. a, a, a P.• a, a, a, a, a o. a, a, P. 0 0 0 a) a) a) a) a) a) a) cu cu a) cu cu cu cu a) I) a)n 4— E .:: .,:: .::, .4.t, .1: .1: :t: .:. ..t. .-:. .t: .-:_. ..- .4:: ..,., ...: .11 4—• I. • ..I . :à',9gggggqg g ,9 6 6' 3 3 3 3 3 3 3 3 3 (5 3 (.5' C' 3 3 3 3

71"

N111111 C"-• ("-- u IC'c) Cc) I

7:1 Mi

(1) (1) (1) C/) c/) • PP a) a) `C:41P4 P4EID`c't cq cq u 1 1

71 - 7t" CD CD Crn V-) Cl Crn r- oo oo oo Ci lin 0 oo 7t 0 en -71- cA cN CD 71: CD ci en oc; tri <3 -. 6 -. 6 -. 6 6 m c‘i cq CN

o 0 crN ee; o 0 ,., ,-4 ... ›-, — el i.. ) 1.) = = L) a) o o r cl .r.., W 19. L') ct 0 0.- C—) 1.) g 4.-) O• _Q) .--, L.) g '-i' = .2 U at 1 a' 14 __,1 ..• L() 70 CT \ ct 7-4l•-•1 `'"Ir 0 0 ;-' ,--( N 1 0-5, , , , N (..7 , I ri...I I

0-1 o4 o4 o4 1:4

✓ C.le) .4 \•(:) OC nCD nID 00 00 00 N NNN00 g 0 ci ken <3NNN C NNN cr; 4 4 O CD CD CD CD CD CD CD CD CD Cl Cl Cl N Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl N Cl Cl Cl LO Cl Cl

▪

-s4 ri) ri.) 1.) cu cu a) a.) a) a) a) a) a) a.) a.) cu ct P. PL. a, a a a, a a a o.., a a a., P. a a., aq a a a faq 7:) 77) 7D 7D 7D 7D 7D 7i 7D 7D 7= "0 7D 7D t) +-C2 (i) cl" Ô Ô Ô ÔÔÔÔÔÔÔS CT) g g ô eP- P- Q:5 P- (5 4 4 4 4 6 4 4 4 4 4 4 4 4 4 4 4 4

V-)

(/ (IDC./D tk0 tD-0

ri1 0) Cip Ci) 1)1)

c) f P , a4 o4 1 F-1 F-1 lN 1 D D I

tr) Ch r- Ir) Cl AD 00 AD -7 AD 00 Ch 71- 71- 71- AD AD Ch 00 71- cc) V-) 00 -7 AD 71- r- r- cN 71- 00 Ch CD (C) Ch -7 Ch CD Cl Ch v-) CD V-) 6 V-; 7i C5 C‘i c5 c-c; e5 6 6 6 .6 6 cr; 6 6 -2' 6 N tr.) cl cg-)

a o-, a O o›, O a m a a O o u o m O m g q a) C.) 0 C.) *r:-', C....) cu a) . —, c•cs= CI) V) L) .2 .-g , E N rl. c0) m c..1 cq -. cl ,.',:: Cl E.1 Cl 1 1 1 -- .1-' 1 Cl 1/4....) 1 CA 1 t

• p4 p4 G14 p4 p4 p4 o4 o4 o4

trp oo 7t- r- m- co 71- rcp trp N

4 Ir; tr; v=3, nr3 06 06 c7; es: -. (Ni (Ni (,-; 11 •-••1 N NNNNNcsINN CA CA CA Cl CA CA CA CA Cl Cl Cl Cl Cl Cl CA Cl Cl Cl Cl Cl Cl Cl Ce) Cr) -4 i4 74 -4 <1.) <1.) 1.) P-1 a. a. a, at., a). 7:1 rc; 0 0 0 0 0 0 gggggg 444444

oo oo tin rn 7r C- C\iNN 0 0 C0<2, •--1 N

V') 0\ In VD rn oo c-ri 4 4 tr; tri cq Cl CA CA CA CA CA CA CA CA CA CA

▪•

rcs 1.) (i) Cr) E Cr) •- L-) 0 r:4 z o .— E E ct z -7) ct NN E o z Ei (1) C N 04 d- tr) —1 L) cip n 0,r. • 0 (LI ,a) 0 _g 0 L4 ct * FEI -5•0' r, r, q -•4 cl) c) to 4-4 %-0 ci) •-• o 49 HI -7; — z — c.) ui (.:1 E • cr te-) m . to a) a) (--‘1 tc) ,C70/ tk0 (71 ".0 CA cc) Ic) cr) 0 g 73 cc) v-) 00 d- CA u rz..2 5 ,zt 0z "-av.) cl.) _Lw, (...) c.) , ta•.0 cl) (4 -0 l-c" c/ ct C DO M m z >, t---- M • (-) •0 5 eu up N N • (1) — 44. tk' (1) "c7,' 'c.c1 7:3 0 t) • Ct Q•II • Ct • •—n a.) Ci•A) 0 41 /"! • là 0 el) •••••• 0 C.) O 0 0 ;-n c24 c Q) (/ (L) DO N N kr) ,c) 'E-71 1=1-1 "0 • ^ • 0 6 ,Q (ID c5 c/) b,0 cz1 (i) (:) ;•.1 7110. ct -1n••+ • 0 cl..) • ;-4 bf) Q) - ' CL) .(1) tk0 I) a) cl el" 1.) oc.,t.L..,) E _,.00 A ct H 7i v1 n ° N d- C Ct CL) co NC cn .7) c4E , , •—• cz: c, a.) ° r tr) a) c.) " ci 71-• c-i • cl-) .,„1 r--- c./.) c+-n (/) ›, o .4) (1) C.7 up 6 ct4) 8°)E('''0 , ci oo < (4 rn oo CD lD • ›)..„c7s t:4 3 • •. r- 00 CD 7t 00 CD d' CD CD Ic-) 44, cn ci) •E g) 4 c7) 7:3 t 4 (L) (1.) `) e 47clid 2 cz/ "' o tr) C a o.-CA • 1. a E nnI(1)— E •El • "'>< 7:) .rE g 'F,4 E "cic, u ci) M " =.R 4—‘ 7.57; <1.) 0 ct 4.) a) 0 ct 0 8 E 8 -;re, c,„ . („) E ô - Q t . R; • — _ (L) cl) cct tu) E (1) Cr> 4:1 1-) if-n -C3 ri) E 0 ct o ct — o a (D ,4 _,t).10, ,„), ;- g ci z 2 -7)b•c, to2i z rip c/ o c,„ czs 7:1 ct =-4 „, • cl) ai) <1.) •07.: E 0 `J u E C/3 •-4 8• 4, - 0-- - ° t CA ce"' ccia) 1) 0 T..= il E C) -t 15 78 o ---„ , ri ^(1.) 1.) tcl-) — 1-, ^ c.)• E • ei5' E •-ez, g,-)4 cvj c.) sr .,) 4:9ct E 5 ct-'0?‹ l- Q) c, 0 tic, 4— - °"-gt • F, (i) 0 7i. F3.45 C• ;.C?I CZI >a) g -2 Ct "E

00C tr) oo m kr) ,f oo cl ,t 00 00 Ch VD CD ‹) VD n cl cln 00 VD N M Cl r- oo r- oo oo — rn d- d. nc) v.) d- d- d- d- vn tr)

Ch Cr) N Ch ,--n 0 ,--4 M M CD Cl ,--1 in CD CD 00 M Cl Ch r- CD Cl r- M , n --i Ch Ch Cl 00 (-- r- d- 00 r- ,-- VD r—i cc) ,—I r-- oo Q--) . ca. m r- r- Ch 00 - - — n Cl ln tr) M CD CD 0 Cl d d Ln cin I oo t--- vn kr) tr) .0 . ,--1Clcc)Cc) ,-1 ,--1,—(M\ONClcr)d- o. in

n vn oo vn oo vnn cl 0 N N 7t CD .71- 00 oc vn oc CD rn oo 1D oo 71- tn0 1- V") CD M Cl --, CA le) 7r tr) 1-1 1-1 rq ce) N

n M CD Ch M 00 CD CD V. ) CD rn ID r- rn rn rr.; od 4 r....: oo; 6-n Csn 6 06 --f Ctn M M 00 N

M Ch 00 \ID rn vn oo r- oo oo rn ci oo cq r- rn r- oo 71- r- 7t ln Ce") r- CD In Cl VD d" Cl CD CD Ch d' ,--1 d" r- r- cl vn r- Ch oo r- cD \ID CD CD --I CD V") CD 1.-- CD \.0 Ch M Cl Ch 00 V-) 71 - Ir") te.) r'" ,....-4 1....' , Cq Cr) (n ,--0 Ch Ch Cl in --I Cl M M 00 d-

o cln d- d- ch ch r- \ID \D d- d- CD CD CD , --1 d- Ch Ch in Ch r- Cl d- V-) \D Ch 00 ,--( \ID ,.-- --4 ,---, 00 CD \ID CD in in CD in Cl \D ,-. 00 m Ch d- M Ch 00 r- cD d- —4 . ,t rn -. tr..) te....) cp cq o oc cp . cp d- r- \ID In ,--, Ch CD ,--1 ,--0 ,---1 Cl d- rn VD r- r- 7r

rn rn rn rn vn vn rn rn Ce) \D d' 7r Cr) Cc) Cr) CA Cr) 7r Cr) CA

cr)0n \.01---NO--1 0 n000d-CN nOv:)"1-tr)

4kr;406cNitc, i--;--, 60. 0 q In 0 c7s c,, ÇZ•4 CCA i 1 GO i i U0 g i E—c i 44

E*

N oo (ln oo c:) nc) cc) ch 'n r- ch c"° 6 6 c c rri (1.; cc; d- tri v cri dl 't V N N 00 —4 —4 —4 con 1-.1 1-1

cr) cr) cr)

Cl kfl d- AD 00 On r- V Un 00 CD On AD CA r- r- 00 CD 71- 00 NQ N NN r- CN AD CD CN rn CA rn 00 AD AD AD kon AD tin AD -7 Ln ken 00 r- In r- oc AD

In Cq 7t Cq .---1 t"..." CN On d- 77 CN CD CN CN r- CN rn CN 00 d' CD 00 CD In rn 77 0000 rn rn Cl rn Cl tin 00 AD 71- CN d' CN CN Cl rn CN 71- 00 77 Cl CD CD CN cr) 7)- ,, ken 71- r--- lin 77 On cin ,f Un 71- ,-( , -1 '71' 00 71" 00

N 71- Ir) tr) 0 —1 '.0 oo If)NC•171- cncr)rn-71- (-NIC,100cr) cr) N

d' 00 77 AD 71- 71- 00 In rn rn (C) 00 Cl 00 od cr; csn 06 tc; v3 oc; 4 4 C7. (>6 tri '. 0

N oo rn CN rn CD N r- CN r- CA CD 00 CD d- rn r- r- 71- -. 71- oc m AD CA on r- or) CA AD kin AD 71- -7 r-- N AD kin CA CD 71- 71- AD Cl -7 n.C.) In -7 cin CA 7:1- 7t Ce) CN r'".

N ch oo o cy ,t tr) r- C m CN m 7t m cq ,t r- ,t cl u--1 71- Ch r- C 0000CC 71- d- CD Vn CD In CA Cl 71- cin 77 Cl -77, '.0 00 m CD r- AD On 77 77 77 71- r- r- m Cq r-

mcnce-INN(nce) n0-1 -1 71- NNcr)

00 (r) 00 00 oo r- , t o AD on 77 06 6 cr; tr; cr; cri 6 6 7t -4 (,-; • r- r-

:.- ;;.,-,o ,..t )... 'A g o 7!! .'" '") >-, NO Cr:‘E..* L) a.) czS a) :5( P4 L) N ••-• .,-, :::41 ai 7-_, (1.) 4 4Z ct ct (../2 e, ,7-'1 tf, ';"?", L) 00 c. 0 N .< , ,. ,,, Cl W , z 4 .1 ,,,. „,4 1 g ;=0 0 E0 0.4 • oo ,t 0 m c; tn oo CN CA A '.0 r- 0000 ch 0; o; ch 6 --4 -4 -4 N Cl Cl CA CA Cl CA Cl C40 Cl Cl 71- Cf-)

or) CN CN r- un ,f r ,t un ,f kr) r- rn oo co , 7t cn -. ch r- 71- CD CA r- oo oo-oo cq cNI cl cp, CD 00 7h 71- 71- CN Vn AD AD N r- AD AD 00 r- r- r- AD r- AD CN r- on r- oo r- r-

11-) 7t oo on N cq 0 ,f AD 00 CN 71- Ve) AD cl 71- r- cin r- CA 71- CN 00 tin AD AD on vn CD ,..--1 M 71- Cq CD ,--1 CN r- 7t C.") 00 ,-( 71- CD (e) CD AD 71- r- CD AD 'cl- 00 71- AD Vn AD CD ,--i rn AD CA ,--( AD CA rn 00 CN ,--1 CA cc) ,--1 CA 7t AD ,--( 71- ,---1 CA ,---i rn Vn --(

vn oc o r- r- c) kr) ch 71- oo co N 71- \ CA 71- CA CA \ \ Cfn N Cfn Cr)

un on r- cq oo r- un CN r- CN CN CD on on oo CN AD AD 71-

CD CDn 6 oc; cri -. o6 cie; co; cri cr; or; or). c5

7t be) Cr) oo on ,e) CN CD lin rn CN 00 vn CA CN 00 AD CD 00 Vn 00 CA GD CN r- Cl '71' CN vn rn CA Vn CD Cl CN CA 7h GD 71- r- r- oo ,t ,t un ,f r- ,t

CD CD Cl CD 00 AD ^ -4 00 'et C m c oo cr, oo r- c5 c5 un cl r- 71" Cr) CD CD 00 In C7\ VD CN r- r- cl oo CD AD CN 4 4 C 00 CD Ve) r- 00 CA On n40 CA Cl 00 AD

71- M Cl cl CA cn cA CA 7:1- Cl cl cl Cl 7h CA Cl

cn CA 00 CA NCNNCNN CCC c; c:; cri nc; r:) CA 00

g-4 to CA •z- o -5) ...= cks cn 0 GO cd z g c-,..) = cc,, ,,-4. !,.> ba, o :114. , , , , P , , , E-1 cn

• a4 4 .4 .4 .4 4 .4 4 .4 04 04 .4 .4 .4 .4 4 C/D

Q Cc) 71' tr) CC NN 71" r- 0 on on ,t ,t AD AD 00 CA CA CA CA CA Cc; CC; Cc) 4 4 4 v-; Ir; tri kr) \C.) AZ; N o6 0C C5 il N CI CA CA Cl CA CA CA CA Cl Cl Cl CA CA Cl CA CA CA Cl M

V") Cr)

71- CT\ r- cn un cO un un cl ken on vn (r) cr) N fZ) N vn CN CN CD \.0 CD CD "zt C6 d' n--.1 h- C n.CD 0 00 r- r- r- oo r- oo oo oo 71- 00 00 00 —4 CP\ (7\ 00 CD

cl cn on vn 0. cn c) r- c) vn oo cn N00 vn CN CN Ch CN r- c) -. o cN) c) co r- vn r- ,t CN 7t oc ', r r- CD cn --( --4 oo r- 7t VD CD o0 cl r- 0 un 7f r- 71- Cq 7r V-) VD Ci Cl Cl C.--- ‹) ,--• 7t c-.1 ,--$ --I po cA ,--; tr) VD ---; ,--4 cl --1 cl cn cn --I -. --I

CD 00 r- C ‹) 7f r- nID CD r- r- VD oo CN cl cl CN cl cn r- oo cn cq r- cn cn CD r- cn Cl cl cl cl 7t r- cl oC VD r- vD CN cl —4 cn cl cl

• r- cn r- cc) r- r- r- r- r- c) cl cl vn on ve; n6 0; 6 00CC00004NO 0: kr; c); N ,-••••-•

r- vn VD un rn kr) CD oo vD cn 00 oo CD un CD r- 71- cl 71' 00 CA Ce) CN cn cn oo ,t vn CD r- vD vn CN cl vn cn vD cn r- cn 7P 00 71- CN cn CN CD Cl r- VD N cr)

un CC cl vD CN cn VD cn cl 7t un ‹) CD CD vn CN vn cl cn CN r- vn oo c) CD CN r- CN r- r- c), • ça ci oo oo cn -, oc , t un oo r- cn r- VD CN 71- oc oo cc un r- VD 71- 00 r- cn nID Cr) Cr) kt) rn r- vn r- rn 7t- (4")

Cl cl 1- cl Cl cl cl cn cl cl —1 Cl cn 71- cl cl cl cl 7t

CD ,-( cl —I 0 CN r- c0 0 vn VD 0 ln 00 71- cl ,-:, d ., 6 c5 d c; d • cq o (-,) . -- c c, in N 0 cn -, N W C C „ o o 'co >' J, ••31 ..zc.) › C 0 ct o g cd 0 Q...) c._) o (...)ct .- 7.-_, = -0 ,. L) ,. 70 cd a) L) ,,:: -.Fd , ooO , , , 4 , , ;4 a4 = g ' ;10 0 o4 o4 o4 o4 a4 .4 .4 a4 ...) .4 0.- .4 .-1 o4 ,... ,-. ,-4 a4 a4 o4 E* g4 a.) un CD VD o0 CD nSD 00 cl r- CN Cl VD r- ,t VD r- CC

C,1 C\1 C‘i 4 4 4 V"; Ir; vl:; 06 crn -4 Cf") Cr) fr) fe) Cl e) Ce) rn on c.) cn on cn 00 cn cn rn cn ,t CD NC CD cn oo VD on CN r- cr\ co vn vn tr) CD lin In CD r- oc le) (r) n.0 CD CA CA CD 00 r- oc r- r- co oo CD CD CD CA cn oc oo cr, oo oo CN CS\ CN CN CN CN CN CN CN CN

VD CN ,r r- CN 00 rt 00 On CN 00 00 CD 00 r- CN 71c 71- CN1 CD • 00 C) 7f Cr) Cr) ln C) 7t Ch 00 Tt CA CD 00 Cn tin VD 7t Cl VD 71- 00 7t CD CA VD ,t 7t CA 4 ,t C) VD CA --I VD On CA CA 4 CA V") CN CN

n N 01 CT N C 71- N '71- C) kr) c) \.r) Cc") N CO r- cn 71- cn 71- N ,---(Cn't-,—f ,—f n—(c•--IMNN'crNLr) ,—INCVCr)N\0 ,4D

acl • CA 00 on VD CD r- C:) 7t C) CD 7t C) 00 CD CA r- CN CON oc; tri cl 6 N 6 s::; 6 c; o6 oc; c; nc; vc; od o 4

tr) r- oo r- c. cc) In vn N kf) N 71c rn vo c) tr) oC) cl tr) r- kr) \ tr-) kr) tr) Cc) ,S) CX) CO N kr) CZ) Cr) Cr) 00 CD C=> tr) c•I (-N1 cnli 71- cq

Cr) VLI Cl CA 00 r-.. r".. Cr) 771c oc r- -, -, cr, vn rn ni oo -, 71- le) CA 71- 0C Ch r- n oo CD CA ‹.) on CN ,---1 4 71- r".".. ^-' ID CD CN In lin 00 Cl Cl CA In in -) ...-1 00 CD C.-- CN CA c‘,1 7r. 7t CA CP\ CN VD ,t On 4 CN 00 00 14 cn --, n vn vn cn cq d- -. r- vn , 71- 7r 7h CA cc) cn rq rn 7f CN rD --i Cq CA

cq rn rn cl c, 1 cq cn cn 71- cq cq cq cn cn

oc rn r- rn oo r- cn VD cn cn r- vn rn oo rn 7:1" r, 1 Cq —4 —4 6 6 6 c\i -4 -4 —4 6 6 6 —4 4 cri kr) cq

= 0 cz' 0 ›-. ›-.) z = (-1 cct cci (....) c.l. (.....) cn 0 a.) ,,,., '-g XI 4Ri I H I t Cr) 1

o4 .4 .4 .4 .4 .4 (24 (24 o4 .4 o4 .4 o4 o4 124 .4 (24 .4 .4

m or, 0 -. cq r- nJo oc oo r- oc CD 7t 7h 00 00 ken n.0 00 cN • c-.1 cc; cc; oc; oc; 4 4 tr; o6 ca; ca; ca 6 , t , t ,t ,f 71- 7r "7t 7t 71" Tt Tt 7t 71- Tt Tt Tt Tt Tt 7t Tt Tt Ln Cc)

CD CD CN • nC, CD CD 00 00 CN VD 'CI' .7t 0000 00 C5 CD CD Cl CA CD cc) 7t cl --I tr) cl cl 00 00 C0 CD CD 00 CD CD CD CD 0 CD CD C5 CD r- c5 CD CD CN CN CS\ •-o .o 00

VD, 00 Cl CD CA CN CN 00 tr) cc) r- oo on c. oo tr) D .CD n ‹) VD 7h 00 VD CD 1D CD 7t r- kr) 00 Cl V") - n n n ‹) tr) CD 71 CD VD CD ID C Cl 4 00 VD ‹.) \.ID 4 V.) Cl Cl V. ) Cl nC, C5 n1D 7r 00 CN CN CN CN r- --, on -o c. --, -, cq c, 1 cN.) CN 00 ,--( ,-.

NC 00 -. 7:1- 00 r•-• CN Cl Cl 00 CD CD r- r- r- c,1 cv on on cl c5 4 4 on vn nc) nC) CD VD 00 CA Cl 00 CA Cl CA Vs) Cl 71. Cl C\1 00 Cq Cl 00 r-

• Cc) 00 CN 00 Cl 0 \.(D d- cc) \SD on %.(D cc) ln CA 00 Cl Cc) tri CS N ro; od o6 n15, 4 kr; kr; oc; 06

V) cr) —CNNte") r- r-- N ,—I in —q —4 --i •1- 0', N "1- NN n -o 'd' CD 00 00 kr) N N kr) C) n,—I 0 00 ct r--- r- cs on • 0 on vn N -. vD .r5 c.)- on .4 cqo c). ^' ,.0 71- crn n4D ,--n c-sii ,—, ,-1 ,—I T—n N ,-. .--. ,--i cr) 0000 ,--.1 .--4

Cl AD cc) 00-N00tr) CA 00 r- 71- CD CN CD 00 CN 00 VD Cq - 00tfNN 14.-) cl 00 '71- CD 00 -o •-o CN r- CN ,o tr) C5 CN r- 11-) 71- Ce) V-) tr) Cl cc) Ir) tr) Cl C5 r- Cl Cl 0") Cl Cl 7f 00 'cf Cl cc) tr) cc) cc) cc) CN CA ).-.1 7f on

n • cc) 71- 7t VD VD kr) cq cc) 00 ..mt r- vD r, on .0 -o CN CS C5 C5 C5 CD C5 4 c,"; )--; CD 0 CS C5 c5 4 6 c:; c.; 6 00

1:4 :4 c4 ,4 1:4 1:4 1:4

-o On 71- 00 Cl • tr) )--4 V") 00 CD tr) CD 7t CD cc) ,..C5 00 CN Cc)

6 6 6, c5 -o • C\i cc; cc; cc) 4 4 V. ; V. ; \r$ \fS Vn Vn tin tin tin V") le) V") V") V") V) V") V") V") V-) V-) V.) V") V.) 00 cr)

\ 0 0 ',CIP \ v) c-1 c-1 4cs n.C) VD, te) rn r-- tin cl co r-- CD CT <71- te) co c-1 cl CD 71- cl CD r- c) m c) c) o kr) o N-N N-N 00'--00 v) oo 00 --- nC)

Coo 771- cl 0 r- r- cl r- r- te) un %.JD 7t 7t r- r-

• vn Al) r-- r-- r-- r- rn r-- CD CD r-- r-- ▪ CXD CI) \ CD Cr) ten r-- cn CXD ,71" \ CD Cr) Cri Cl r-- c-1 c-1 nC) cl cl ci) cl tr) c-4 Ln Al) cl cl ren cn

te) te) 7.11- CD r-- r- CD CT te) CT r-- CXD tr) 00 cq rn • 06 c‘i cs; oc; oc; 6 oc; cS cri 6 6

v) d- cT ) cT cc) cl (T (T r-- v ) c-1 c. c‘l CD Cl te) cn oo (NI t-- c-1 v) cn \c) • c". r— v) r- r- CT CD CD cl CXD CT oo un 00 71" Cl 00 7t V.) Cl Cl \DID CT

un rn ON • o .1- r— rn oo r— v) r— ct. .t. —. rn CN vn • 71- CN cl CN \-1D \.0 CD CN 71- CD CD rn —4 --( CN co 7J- ---( cl CD cf) C) o0 (..1 ch ch .. co CN CN tin tin C) cl CD Go vn CD r- r) tin cl CN —4 cl —4 —4 rn 7r -, --n -. cn rn tin cq ..7r tr) 71" cl —4 N

E <1.) cll "CS ce, (-.1 ..cncncn—irnt-cnNrc)Ncicl mn•4 11 N r.7,)

N 0 t .., v) r-- cl cl v) r-- ,:r 't <1" ,1- G, C-- C-- c',1 ---( C'A Nt r"'": 6 6 tri 6 --; 6 <:; a: 6 6 -4 6 6 6 :..g c) c) c) c:) c)

W4 = o o ..4 .-, = c9 czt C....) (...) 14 m ., oCI) = to -rt' C....) , 1 d , i W r i g 44 0 • g g tz4 1:4 1=4 (24 E—i g v) r-- oo c> v) CXD , 71- 411) LIID r-- cl cr) v) Ao --. 411 r-- CN CD ---i Cl

r-: t-: r.-: 06 (x; cd c). cp: <::, ez: 6 6 6 6 .4 v) v) tr) v) tr) v) v) v) tin kt.) c'D 4o aiip c'D cA v: v=> .cs v:, I.7)

Cr) Cr)

• 00 D N71- 71- 7t 14-) C nC) in In Vn d 71" r- ch CD CD Ch In 71- C) \..0 00 CD CD CD CD CD CD CD CD CD r- ,t r- on ,t oc 00 ,t -. 00 „. 00 r- \..c)

-c) N 00 N r- CN 00 rn (NI n0 Ch r- 0 oo N Ch 00 I.1D nit Cl \C ,C 00 Cl CI rn Ch CD CD 00 0 Cl CD Cl d ,r k4D Cl Vn d• CD Lin Ch 71- 00 In n.0 Cl • Cl CN n1D Ch Ch Ch 00 (NI Cq 71- "I

0 (4 • cr) tr)N,CM CS\ —+kr)c)0 D C 01 in r--- N.

• n4D 00 rn rn Cr) d- CD CN Vn 00 Cl d Ch r- r- C

•

c.)

• ,t on N ch r- r- ch r- rn rn 71- 0 Cl d CD r- on rn on 0 ch Cn Cn ch m po ch \.0 m• en r- 0 on on 71- rn Ch d" 00 CD Cl nID n42) m cq 71-

0 00 \..0 Cl 71' CN Cq 00 CN In Cn oc r- 71- (....". '—' on '...0 00 d 00 71' nID Vn Ch \.0 d I..0 ,..0 ..ID rn r- on on -. CD ,--( d lin ,--.4 ,- Ch 00 ,---( CD CA Vn CD tr) n , - In N CD ,..0 d '..0 Ch d r- vn on vn r- CD 0 00 r- CD in ---1 71 Ch rn Cl \.0 rn cq on N N cq on N -, 71- ch rn 7/. n..1D I..0 .-.. 7t Cq 0 <+-n ci) C)

Ac1.2 Cc) Cr) Cr) Cin C'fn N C'fn Cr) Cci c'el

r- v.) r- ,t (-NI en • Cl 00 n.0 rn r- 71- 71- Cn nS0 0") 71- 6 6 6 6 6 6 6 6 6 6 6 6 4 4 cd 4n

cdt4) .._ :.'.z a.) ...cl) a) 0 .-- co z .-, (1.) 1-, 0 "iû ç'-:.) Q (..) ..., cl) (i) 7",z: cz9 C1:1 cn = • (....) cl) ct ..gZ = 7:3 (...) 0 ,r,0 ,,,ct cc; •(1)—) ctE 'A a -7-L) ;" cl cn 0 I I I IL)111n11 H4L.) I I . 0

C.)c 104 P4 o4 p4 p4 o4 o4 p4 014 Z w - a 6" Vn CD rn rn on on oo oc CD In rn Vn Vn rn oo Cl \.0 00 • C•i Cc; cr; Cr; rr; Cc; cri Cr; 4 4 4 nI3 n1D nSD \.0 nID n4D n1D tr) C N C ci — r cc oo CN 00 CD Cl CD In 00 r- m rn CN v-) 71- cr) n v-) C N —J •rD N oo v. ) oo 71- 71- cr) \sc) nC \ cr) C \ „CD tr) r-

CD r– c — CCrCC r- 71- CN cr) CD 00 71- cr) Cl CS\ \..0 00 Cl Cl 00 Cl cl r- oc 71- CN \.0 CN CN NC 00 Cl 00 r- CN Cl 00 CN 7-1 CN VD 00 71- 71- rq rn 71- 71- VD ("1")

N 71- (NI N r- 71- —J •-•-n NT--cnclv-)cr)vD VD in In 71- N71- Coo0OclO n0cn

• VD 7f cr1 CD (4-) --I VD oc r- 71" 00 kr) r- CN \.C) tn

Cn 6 CT 6 o6 cs.; ch -4 co.n 3 od csn o6 nc; cr QC od od

VD r- rn CN Vn Cl VD CN cr) nID (/ -00 Cl CD CD oO d- 71- 0 v. ) 0 c‘i \,c) r- n•C> Le') cr) oo ao cr) • %.0 VD 00 CSn N CT CN tr) Cr) Cr) Cr) 71" 0 n—n 00 r- r- —171-cr)N-1 C Cr) 71" ". 71" VD fn

cl tr) t-- v-) 0 d- N CD Ch —1 71- Ch 00 VD VD CN CN CN 71- 7r Cl CD CD cr) ,--, 00 Cl CI ,--( 0 `.0 '7:1- 00 0 00 CN 71- 000 d- d- cnr-- —I --, v-)N71- vs) \ID C'l n 00 "71- 00 oo CNNI----000 t"--- \C ---(N71- Choocc) ••0 —1cld- In r-) cq —1 ---1 Ch 71- cc) J---J r– In d- cr) v-) ch cl CPN 7h —1 (-Jr) —4 --, r-

Cl 71" 71" Cr) m 7r tr) m get m m 'et 7t 71- m

• cln Cl CN r- CN Cl cr) Cr) nSD 71- 00 71- \.C) Ch 6 (T 6 4 4 CC Cc; C‘i — N -4 cri 6 4 -4 -4 6 n.,c) 6

• c:4 r:4 a4 (24 (:4 124

CD kin 00 lD CD Cfn r- CN CN Cl ZJ, rn CN CD CD In in CD r- ch

QC, cc; oc; cn 6 6 6 6 6 -, cNi re; rei 4 tie; Lie; 1r; Lee; v3 nr$ 17.S • ‹D VD VD r- r- r- r- r- r- r- r- r- r- r- r- r- r- r-

71- 00C cl /D r- /D oo /D r- Ch cfn 7 - 1 cl on on on oo on Ch on ri- cl on /D 00 /D CD -r vn /D cl vn un /D r- CD 7t Cf) /D Vn vn /D /D cl CD on CD 00 r- ch vn 00 00 00 00

7t n r- vn oo rD r- r- vn Ch vn 00 r- ch ce /D CD cl vn on ri- Ch CD CD C3 on 00 r- CD rt un 00 un vn cl cl rn Ch vn oo cl r- vn 71- Cl un /D rn CD cl r- --f -r Ch Ch CD rn cl cl cl r- C-) .7i- /D --f cl -r rn -r /D -r -r cl ri- cl ,ty rn -r Q)

4-4 0 7t -r /D vn cl r- CD /D vn /D oc vn Q) oo cl ch ch oo r- ch rn oo rn -r CD CD cl cl cl cl rn cl 'as -r -r cl -r on rar cl rn cl cl cl c‘,1 nID N N

CN oo CD CD Cq nSD oo ch /D Ch /D vn Ch -r /D oo cl • CD CD C3 oc; cNi CD /D 6 od oe, 4.1 oc; cl

e)0

on 7t 00 C".. on r- vn ri- Ch CD vn Ch -r 7t CD ch 7t Cl OC oo /D cl cl on rn vn cl r- cl CD oo r- CD 71- Ch CD cn 71- C1 CN r- n(:) /D oo ch ch cN ch oo r- oo (,) vn ch c..1 on cq vn oo oo vn -r -r on cl on cq -r EL' 0 5

tio Ch oo cl r- on ri- /D po 71- Ch -r CD -r CD on rr CD un cl on oo ch c CD /D un vn -r tr) -r cl CD /D cl Ch CD cl cl /D cl cl cl oo -r oo on r- vn r- CD on vn vn /D CD cN /D vn CD oo on ri- /D CD CD /D ch /D -r 0 Ch ^-, cl vn on Ch ‹) cl -r cl cl -r d- 'Tt '--, Li-) ON /D ri- ri- 4-4 ci) C.7) ••n-•• • • 71- cn on cq cn 7r. on on d- rq

n n0 r- un on vn -. ol on -, /D oo C3 cl CD r"---- CD ,--4 f:;, • •-n m oo cn -r CD cin cl CD CD cn CD r,..- --1 -r ve'; -r 4 C,i Cr) d N Cl ,--1 ..--( cr) Ch C-- a) b1) W4 a) a) In ,-., a) o= al* raz Q) O a).) cd Q) L) 6g= r ) 1.).-- 'Ci c.a);.3 czt ,._ , 0 0 (.) a) "Fi vz u ---,,, 2 0 C3S (L) a.) 3 -- ,., : 2 L.) = > (i) uL' L)'—' C..) ci) '6 8 0 A a -c; t , 4.- up Ct —5 0 Ii -,t . -P„ ,9 0 .. 1.-I o 0 — c.) 0 g4 ;.1.1 0 ,... o4 . o4 .4 p4 ,-.. 1:4 ,- o4 .4 o4 o4 ,-. ...1 ..) o4 ,-. o4 ..)

CD rn r- CD ri- /D r- cl /D cl vn /D cl on oo /D ch vn od o.46 od ch ch ch ch 6 6 cl cl cl co; co; co; 4 4 00 N N r- r- r- r- r- 00000000000000000000000000

Cc)

n 00 AD CD Cc) 71" .C) 71- Cr) (-1") (f) VD 00 v:1- at NCO 0 00 r- c. •—n CA CA Cl N V.) 7t. tr) kr) CN 71- VD d- Cc) VD Ch d' CD 71- ^-4 po .1- po po cp, po ,—. po

• r, un oo rl on on 0. CN CA VD .1- —1 CD CN AD ^-4 00 Cc) VP , C) Cc) AD CD Ce) Cr) d- Cc) 00 cc) cc) CA Ch CD 00 d- Cl CN CA r- CA CN 00 Ln on CN CN CN r- V-) AD re) Cc) AD Cl CA CA AD Cc) Cc) 71- CN Cc) CA

vn r- vn vn Q vn oc CN C V cc) 00 CN • r- AD CN 7t AD r- r- oc un on o oo cq AD d- r- cc) ,—. —4 cq •-n1 1-n1 ,n m m m m

vn vn cl 0, r- on un un c-A r- (D r- (D cN (D CN Cr) o; o6 o6 cc; c; 0; cc; .6 .6 ()d v6 0; o6 n6

r- cq cc) 00 CN nrD tr) y) C) CA CD r- V-) r- co co • co oo cq ,f co un r- r- co c, 1 CA Vs) r- cA vD 71- CD CA r- nXD oo CD r- on r- vn nrD cN r- ,f oc .0 vn in CA rn r- cq N C — CA

Cc) VD ON kr) kr) .c) r-- on r- vn --, on CD CN Cc) rn oc 0, -, oc -, in AD CD CN ,-, r- CD CD on CN cc) d" CA Cl V") CN CN CN CN cr) AD In cc) Ln CD po n , .,.--1 C-- d- kr) 00 r-- ID C) oc r- (D cq 71- r- CD CD AD cc) CA cl ---1 po ,- ln r- V') CV AD d- 00 --.4 Cl 71- 00 71- '-' nCD

onc-Arin.l.ond--tmun munonclononvnonc‘lononcn

• cN -o cl oo oo 0. on on 0, on r- -o cl CA Ch r- on 4 C.5 C‘i C‘i ci d C NN C CD\ (Ni AD CA CA

cu ,-- -a7') a) ,-. ar) ,..(,) ),o a)o z ,a) 0 -.t a) LI L.) ''. ), ?,., a) ‘, = . ,e2-,-° 2:3 . cl.,-. c,), -,. r L3' , a , , .1-, ch F.

a4 1:4 ‘:4 1:4

AD CD CD AD r- r- CN CA 00 CN CN CN Cc) Cl in CD r- on un un • V") le) Le; V-1 OC3 NN r•-• 06 oc.; 0; (D NN cc; oei 000000000000000000 00000000 cN CN CN CN CN Ch

Cr1 7I- Cr)

E 4 CT CT 71- v) in 00 nID c D CD ---' cr) o r- cl di- CD nC) CD CD 71- CI CN CA 00 CT CD CT CD M AD M M CD C) , f •-• CA M M CD CD CD --I CA n40 cN cl pc ch r- r- r-

AD M CD r- ,t cq CA CD CD CD CD CN vs) 00 AD M d- CD r- -o vn cl cl vn , t AD d- CA , t CA M 71" 0 CA "a ^a CA 00 M r- rn AD AD • CT 00 V-) CT 1-n CT ,.rD CT V") CD rr CA CT In CN \SD 00 • V-) rn rn M AD CA CA d- AD CA AD CA CT m

00 d- 00 in 00 M -71- 0 CT cr) CT cc) cn d- r- r- r- ,4) d-CIADCAMCANcr),--,CAMMMNOOCACA 00

AD CA 71- 71- r- V") CD CD r- AD AD 00 00 r- AD r- vn oo c) v-; 06 06 06 06 nc; 6 —4 z; 6 67: 06 N

r- 00 CN r- CA CD CN CA 00 CD 00 CA CD CT 71- CD CD 00 tr) CA 00 d- CA M on on on oo o oo vn c) 0. on ,t c) vn cl v) c,) c\1 -o cv on 4 CO\ tr) t"--- CC) ^a ^a Cl \C, r- vn • vn CT M AD M r- CA CA on on vn cq cc) Cl cr. on

N oo C cn .1- vn 00 CN AD CD r- CD tr) ‹) on oo ,t cq 00 vn tc,(DcN 03, 0, CN CD m 00 7f 00 M M CD 00 CN r- VD AD CA r- on 00 00 AD r- r- 71- ^a tr) CA Cl.) CT AD V") CD on AD CA CA CA 00 r- r- In V") 7.1- CT CT --I V-) CA 00 AD 71' CA CN

n c\Irtri- rncnrrrrd- vDvnanri- ri- c-ncnri- cnrnancInc Irn

00 - 00 • d- CD 00 CT 00 771- ".1 AD CD '71 CD lD on AD CA CA CA 6 CD CV M c‘i M CT 6 V-; 4 h.: Cr) ,4 6 cr)

o 0 0 >, ,-- , - - n >, --4 ,._ cd C) ,_ . (74= cu 6..“..) 1.) (1) 0 ct (....)

r I ""\ _ ; n- a.) L) a..) ;-, a) .., (-) (..) E 1-,a) (....) c) a) ct

g4 fe4 po4 p4 g4 a4 g4 g4 g4 g4

nCD M CD CD AD CD CT CT on on C NIIDÇn 4 4 tr; 1r; n1:.; 0000 c) c) c) c) CD CD CD CD CD CT CT CT CT CN CT CN CT CT CT CT CT

71- N C N oo cN (2) 71- nrD cn tr) nC") oo NC 71- r- cr) co r- t-- "t• CS\ vn (Dn N cl 0 71- cl 71- cc) CN cc) CD N Ch Cl N vD o0 CD vD (2) N 00 c7\ 00

00 DO 71- -1 0 r- 0 -1 nrD r- cc) Ch oo r- r- cc) rn • oo oo CN r- cc) r- N r- cl CN 0 oo r- 0 -1 cn cr) oo r- t-- t-- vD vD oo r- •-n vD N -1 cfn7rvn0cr)N-IcnNvn71- rn-in-id-,--in-ivD-(-INCT\

Vn 7t vn 7t 7t 71- Cl rn 0 CD 00 71- r- Cl oo 7t on r- 71- cn vn Cl 7t Cn Cl VD vn rn vn • Cl Cl Cl nID Cl r- cl 00 kr) m m

CD vn Cl cc) 7t Cl C) CN cl CN r- Ch vD cl 00 Cl el CCN 6 6 06 06 06 ‹.) 0c3 CT 06 06 CNi Cfn

00 cl -c( r- r- rn 00 cl vn CN Ch cl cn C) cn rD, ch r- oo ch vn r- vD rn cl CN 71- 71- rn Cl 71- ,--, 00 r- CD cl rn r- 7t CN •,47) r- vn r- oo rn 7t nrD CD rn r- r--- Cl CD '71" cl vn 7r rn cl -. rn Cl 7t 71- cl Cl 7t- vn Cl 00

cl vn Cl 71- 0 r- r- cl N 00 cc) CD CD 00 D vD (D oo -. rn cn Cl CN vn Ch vD d- n0 00 C7-N r- .1- .. oo 7r oo Cq 4 C..) CD r". Cf) Cr) ", Ch vn ch CN -N cin CD CD 00 vD CD Ch n/) rn 00 CD cn CD ‘,C) ,--1 ,--q r- vD d- Ch ,..0 ,--1 -, t-- cr) 7t ^, rn cl -. rn rn vn cl

cncnkr) cncicncencncncnvncncncncc),-.71-NelNvn

r- vn vn ch CD rn CD r- oo rn vn r- vn 00 rn C•1 4 4

o ›, z a) 0 0 ct J, (..) 0 ,_?,. I.a) ct cz: a.) •,r,nM C-) O—. ..../ 0 C...) L) ccs mi . ,--;., • cl .E (...) 4-' P a) a)c.") ct cn 4--. .,, SM-, cl)E. c0 ,.. 0 i C/D i = Zr/D i IL) 1 i 1

r:4 (24 P4 P4 a4 (24

rn vD VD CN r- r- CD rn nC, 00 ‹.) 00 Cl Cl 71- 00 Cl rn Cl Cl

4 4 4 4 V-) Ir; v6 cx; 0.n 6 6 6 6 —, CD CD CD CD CD CD CD CD CD CD CD cp „, „, „, „, „. „, ", „, „, ,„

71- Cc)

CD CN 000 CN CS Vn CS') CD Vn CN Cl Cl 00 on CN CD 00 CS 7t• 0 00 on Cfn CD Cl cin 71- (n 0.) 0') kr) Cq .,t7t oo O– n n on C Cs 1r) oo —• oo oo 00 o0 oo O.\

'7) C7i cn N NNNNC 71- n7t- In cr) cr) CD cl oo ,c) tr) In 0 - cNi tr)

N CI N —• •—n C \ tr) 71- N —q C•1 - Csn (L)

o Cl 00 7!" cq CD 00 00 CD rn CN 7t cn 00 r- oo cq vn on r- r- r- CN CN • rn On Cl CN Cl Cl rn NI Cl 00 r- oo

vn CD On on Un Cl Cl rn CN 7t "I r.". 'C) 00 0e) r"-. 06 06 06 6 6 4::; 6 crn cV F7–n 7:4

cip

• ci vn r- cn oo un r- 7r cv cn cv 00 7:1" • r- r- n n n 00 71- V-) 71- Vn CD Vn r- CD ‹D ID CD CN 1D vn CN CN 1D 7f Un CD rn Cl Cl CD CD 00 on Cl 00 cn r- 00 CT CD • cq cq cq oo 7t cn CN Cl Cl • CN Cl 0

a.) too 0 Ch cl 00 ,...-.1 CA On Cl ,--( cn nID CD cn Vn Cl CN ‹) Cl CN CT Vn 7t CD Cl C---- '7f CD 00 71- CT CT r.---- CD CN Vn Vn —4 ‹) C--- 71- un Cl nrD CA CA Vn ,--1 ,---t nJD ,--i CD r"--- Vn '1- 00 'SD 00 --I --4 Vn Cl Cl rn n1D In 0") In 0") 0\1 ("1 ,--1 CN On vn Cl ‹) In 7t In CNI Cq In 0

• cl cn cn cl cv cq cn cv cv vn ,t cn cl cfn CtC)

C.) • rn r- un ,t oo CD CN CT (2C.) CD CA Cl • cN r- CD C•1 CD -4 6 6 ir; 4 4 6 6 6 r', -21 CS 6 ,-21 czi cn W ct 0 ••–n C1.) 1;34 u•-n --czt C C:-..)' ^C) cy = .4 tt C.) R C.) .2 = t L.)(z, '-c C.) E , mt 0 -.e W z I I 1124 I I I , .0 g

c5t (21 7,1 F* Z tr) tr) v:s) ir) tr.) oo oo gocnvDt--

c‘icNicncncricn4.4v-iv->v-i nc>vi).6t-:r-: 't 06 06 06 06 - 1-1 •••nq v-I .-I n•••••1 v-,•0 nn 1 1-1 vnI •••••1 1-1 ••••1 Ct`""""", ^, ^1

▪

00 'Cr 11") kr) 0 00 00 00 N C Z) N• 00 Cf) Cr) \ 00\ N \ (:) C-- 0'n 00 CA CN C-- n cc) cr) •— cc) cc) cc) 71. cc) Ch Cl 00 Cl Cl d- •—n 00 00 gn 00 —1 00 —I 00 00 —; y-4 00

N oo cfn oo on CD vn vn r- cv vn CN 00 CN 00 V-) CA r- ,--1 r- cq ,zt (-NI r- rn r- cl r- --I --n 7h vn \ID 4 r- vn CN CN kr) 00 0 d- C) r- cc) vn . ..71- . cq -71- r-- 7 vD 71- trn CD 0 00 cn c-A 1- ,—, ,i- —, ..—IN--111) \ .0 Cel ,-1 \ .0 CT ,—(NOOCAN-1 --1

r- un r- cq un oo rn O C on Po r- r- cv cq VD CD 00 r- CN \CD \cD 00 CN CD cc) lD Lie) \ID 11-) VD 4 VD \JD Cl , t v-) 4 —1 CA CA Cl Cl Cl Cl cl cc)

cc) cc) CN 00 CN oo r- on c‘l cq on CD CD tr) \.CD 00 CD On (DO t:') 0,6 pc; c-Ni oc; 6 c‘i oc;

00 CA \ID 4 v") un r- rn CN r- r- CN kr) CN V-) CD on NOONr- ln Ch 71- Cn 71- '''''' 71- vn on CN \CD Ve) VD Cl CD vD VD V") Ln 1.• CA C -- 4' V') VD CA , f CD Cl Cl 00 cr) VD CD \CD 4 cc') Cl r- CD on ,... on ---, on -. -. cl -. V-) VD Cl —1 7h ---i --4 CA CN Cl Cl ---(

C) kr) N kt) 00 oo 7r oo 71- Lin CD CD CD CD V-) , t ,t d- 00 on cq on on ch cl on ch on oo cl r- cq r- cl ch ch r- Cl 00 CN CN 00 VD r- cc) Ir) cc) d- CN 71- cA 4 4 Cl 4 on 00 CN 00 cc) r- on c‘l cq --f cc) CA cc) VD CA le) r- Cl Cl

Ce) Cc) V-) Cl Cc) 71- Cc) tr) Cc) Ce) Cc) kr)

00 VD CN V') Cl r- CD CD 71- CD Cl 71- cc) Cl CD 4 N CD CD 0 06 oc; CD CD CD 6 6 N on

o4 .4 o4 o4 o4 .4 .4 .4 o4 .4 .4 .4 .4 .4 .4 .4 -4 .4

CD Cl Cl r- VD 00 r- cA on vn r- on tr) ‹) CD d- CD 01 01 01 cT c:; cf.; ci-; 4 4 tri ,--INCN1NNNNNNNNNNNNN - n-n1 1-n1 ....

N cf)7I-

tr) r-- r-- tr) c", Cl 0\ --n CT ,c) Tt ( NI --T t.c) r-- c) r-- • nn nID Cl nn Cl 7t \c) CT r-- ,C) cc) 0\ --I Cn 71- r- fr c) r- \..0 CD 00 CT 0\ r-- VC) CT 00 00 CT • ,r1 irr.1

QC CNNr-- r-- c) r-- CD CD 7r C71 CD r-- CD CO CD In CD) n co 71- CD CDN C) r-- cf) t-- —I 0 ,t 0) CD) r-- trn • tr.) oo .c, r-- Nt r-- r-- v:) r-- c) lf) v:) tr) tr) CT CD\ Tt CT Cl 00 cc) 00 Cl cc) 00 (f)

tr) re) tr) r-- Cl CD CA 00 \CD CT CD \CFP Tt r-- c) kr) , cnoft000r-oc, cn c, ) cn CT Cl cl) cl) cl) CT r--

c;) cXD CT CT CD CT v.) c) r-- tr) cf) CT CT 410 cl) CD

ton CD 4 kin C, cri 6 ocS c, i oc> --; tri r, i 06 CT CD c:rn

• ,t 00 r•-• C NC 0CC 0\ CD kr) r--

00 ,--( 7:1- CD .-1 cn 0\ Cq C) -. nC) 71- nn 00 7t r- 00 nn CD CD tf) nn 71- r- cq cn CD CD oo CD CN -. 'I- r- tin .r) ,--1 --1 .1D Cq CD Ch Cl r- oo o 71- C) '-f 'f C) CC) 00 In Cn CnN Ch ,--1 nn Cl r- cA cr) Cl CT Z) r- rq ,t cn -, 71- Cle) f, Cn '''' Cn Cn In \.0 Lin ---( -. -. CA gn n400 ---1 -. 7f.

N( fq) -,1 '7r Cl cf) N cf) Lr) cr) tr) c,) Nt fr c, )

nn 7t nn r- oo o oo C) CS) cf) <1- CD r-- cf) coo CO 6 cS 6 6 6 I.L; cS 6 • cc; ,-4 c,i c:; cNI

cl .:- 0 .-- = (1) 0 ›-, 0 cuo s-,cA LI) >•= n o a) a) C71 "' Ct U CZ: --( ej U ,-= U U ' 1) ›, cu 0 I.) U ."1' "') ,..4 • --, 1-, ct r'l •, n•,::) :z, 1.) 0 ' s-• r--: 75 •.--, , p ° C) 8 r-- N a.) DO I.) c-r) "CI N N al-, --4 (1-) 71 0 7" Û, ---4 C/) '.: 1=1 1 1 (._ C/) 1

o4 o4 (:4 (D4 O4 o4

\C) CX) Cg) r-- CT CT Cl cl) CT kr) 6 c:) tr) r-- CJ\ Cg) IC; If) q"..) (".5 r-; oc; 6 6 6 Cl Cl Cl Cl Cl Cl Cl cf) cl) cl) cf) CI) cl) ,-y

00 cc)

N AD 71- AD rn 00 on 00 oo o •—n ni:) o 71- oo cs o cs cin d- 71- on r- cl 7t 71- o cl on d- on oo 71- cq vn 71- cc) oo cl r, 00 r- r- r- oo oo oo r- 00 Ch

00 un CD r- r- r- O on -. r- r- r- 71- o r- Ch Ch r- 00 tin 71- AD 00 cA rn rn --I 00 Cl Ch AD 7t. 71- r 71- 00 CD CD In AD In CD In AD Cl 7t r- AD rn CA —I 00 71- 71- vn 71- cl AD AD CD AD cD 00 rn .4 Cr) CA ten un ,--1 ---( cA --, AD cfn AD AD --( ,-...-( ,...-1 00 7t rn

Cl AD 71. 71- --I CD CD 00 00 N cA o 4-) on on rn Ch on CA CD tin op 7t ch Ch cA CD ch op ch Ch CD op CA CA r- cq oo cl vn ,f on cq CA kr) "" "" ^. 7t ".

CD 71- 71- rn 71- rn Ch CD r- CD CA Ch Ch CD AD r- Ch AD Ch rn CD CD -4 -4 6 c7; cr; \ci 6 -4 o6 c,; -4 6 6 cri

00 Ch CA 71- 71- r- vn on 0, ,f on NC on 71- o AD AD CD rn Ch AD un AD In -. CA CD CA In rn on o cl ,f cD vn 00 "-I CA CA Ch AD 00 7t r- r- cl ci -, Ch le) 00 \C) r- r- oo on ,f cq cq ,f on ch c.' cn r- 00 71- cl

AD CA 0000 00 71' Ch ". Ch 00 CA CD --I CD AD Ch 00 71- --.1 cp cs1 cl r- in ,n ch c c N In rn CD --I ,- Cl Ch Ch r- on Z) 71- In 71- po ce\ oo on --, cn r- Ch tin --( 00 AD ---1 CA CD CD CA AD ,r on r- r- AD CD CA Ch CD —4 r-- __, r- ,_, cl , ,--, ,-, 7 n r 1- (NI .0 cc) --, oo cq --, 71. 71- on cq -. n1D

c‘i on 7t Cq 4") on 4 4 on on on V) ". on -. cl c, 4 -. on

Cl tr) Z) on on c‘, .i r- n.ro cs cl (:) 7r oo 00 kr) CNI CA '71- 6 CNi 173 6 Vi) tr) 4 CC Cr; 6 4 CD CD CD CD 6 6 4"; ", 7t (NI

Cc) I I I

.4 .4 0.4 .4 pi4 .4 0c4 .4 p4 p4 .4 1.4 .4 .4 .4 .4 p4 .4 .4 .4 p4

4") CD CD 71" 00 Cl Cq Ce) 00 '4 CN Cl 4") 00 n1D r- oo on 'Tt In Ch r.j cc; cc; c-c) cc; 71- 71- kr; \ \s3 r-: oc; oc; cO ofS on on on on on on on on on on on on on on on on on on on on on on •••-1

M VD n0 0 00 c 00 0 Li-) CD N N ocNootwrc ao r- M 00 d- M CN N 00 N N N - r- 00 r- CX) -00 tr) — r-

7t CD n-••4 N NC CD cT C'0 VD 0 0 Cr) tr) C 00 M CN CD lD r- Cr) N CD) Crl Cr) OC 1-1 00 VD Cl CN 00 Li') VD r- V;) N 71- 0.n C.3. 71- 00 Cl rn rn CSn Cl N r-.

Cr) 71- d" Cl 0 (C) N 00 00 Cc) ‹) tin *I N r- CD C t--

cq CD \..0 CT r- C) rn r- r- 7t cq m r- 71- • oc; 0000 C tri 06 v3

M VD rn r- C) CN 00 CD r- in CD rn vn CD VD 00 71- m un vn oo 7r oo oo oo ,t n.4D cD CD cD m VD -. M rn co lD cr) lD tin tin 71- 7f 00 n.C) CD CN CN CN r-

•--1 .71- 00 d- C--- -. VD —1 N N 71- CN N -. r-- -. —, CY \ r- r- t.0 --, csl 0 VD —, kn N M Ln tnN-,MM00,-.M 00 Inkn 00CNVDCN 00 r-- d- M d-mtn00-.0Nin-. NknMCD•d-r- N N M M d- vD 00 d- -. N N d- r-- d- d- Ln d- -.

tr) N N rn cn rn vn cl vn un m cq vn cn

VD cin 00 N D Ch ^. Ci C) 00 00 C" CD VD r- C". CD kr; CD CD CD CV rr; -4 CD M c‘i -. -4 7r 0, - N •-4 W = 04 z • 0 o = O ›, ›, o ›, g ,0m = L) n 0 . U m cr. ' ›, U C) a) 1 = o m .74 ,.. c..)= • (-) ,.'. ,4 75 -5, c) <1 0c''d g 0 = •- 71- In •t-t 1 .. 1 Ç.T4 , 1 C) () 1 1 ". t 1 Ci) g

.I., 0 1:4 p4 ,_ ,_ „. . o4 ,-) o•-• n-• *.- 1-. (24 24 )- 1-1 En0 ,-1 In CT Cn g „ cncn.N000v00 tr).0t---71-. cr) n c:, c): c; (o 't 6 -. -. c., 4 4 kri kri t- 06 oc; cr c, i c, cf.; mencincn E-1 7t 71- d' 71- d" 7t' 71- 71" 71- 7t 7t Ct Li.) In kn ,-( ,--1 ,.-1 ,-1 un ,. _. _. „ _. _. -, _. _. _. ,. ,. -, „„ _.

n D 00 C- )e) Ln 00 CN CN nC) cc) Cl CD 00 ln kr) r- CN nt CD CN CN nt M CD r- 71- 71- tr) 00 CD Cl C.) CN C- Cl Cl Cl nt Ln M Cl nt nt M \so c, n4D 00 00 --I

00 CGC 00 CN CO r- 00 r- Cl 0 0 cl N vD r-- cc) CD M nt CN cl m cl CN Cr) po cr) 0N Do CN CD CN Cl n1D CN nt 00 00 CN CD 00 CD cc) v0 Occ)o0 c-- oo incnoo Cl nt nt CI N cr) oo c--1 71"

CD CD n-( r- CN 00 nt C\r CD M Cl nID CD in CD 00 cr) cl cl VD 11") tr) (1") CN CN CN te) cc) cc) —4 CN1 Cr) Cr) ÇA —I

71. r- tr) CN CN CD CD d- VD 71- CN 'Tt r- CN r- nt ,..1D CN 06 o6 Lr; n6 '.ONNN C\CN0NNNNN0;

7t- cV —1 Ln oo r- cc) VD VD ON CO CN Cs\ tr) VD r- 71- Cl CD C- nt In 00 00 Cq 00 tr) (NI 00 00 In CA VD CD M 00 m nt CN Cl nt CN 00 m CD CD r- cc) (c) nt oo r- eD cq nn cl m oo cl --, 71- ch c.,1 ,f

kn m 00 Ch —00 00 r- CD CN CD In cc) CD CD Cl c-c) oo oo in —1 cc-on nt CD CN kr) m r- nt Cl r-- r-- r--- tr, cr) cc-, In 0 In [-- • m 00 nt kn nt CD nt CD in oo oo cc) \\ID Ci 0 0 in M nt Cl CN VD r- tr) Ch —I N •—n vD —1 cc) r--- Ch

• c•1 c•I kn cf-) 71- d- c-1 -zt- vp, ‘.0 cc-) cf-')

0 n in c> cl tc) ,-.. 71- -. ,.-- M Gi, 71- N Cc) CN r- --, cv --i CD nO CN cj d c5 tie -. c\i Ir; (-1- --n ....1 cr) tr) .-1 oo g') (--: c,i .6 e5 c,i cn --' . c•I r-- CD -1-

= 0 0; >-, z (ci

G5' L) 1-* q c5/ a (.) a) O .0 C.)

I I t I I I IH4-11 I L(i-'1)1 I g P64 4 p4 o4 o4 p4 .4 a4 4 .4 E* • Do Z) Cl CA le) n.C) oo oo In 71- c) cl c) 00 • cci cc; Lri 06 c;; cs 't 0NN0000 cr: kr) kr) v-) In Ir) Ir) tn tr) nC) „, c/o

-

kr)

CN 00 L C N 0 00 \„0 r--00 r-Nr---0,00,0cr10n 00 tr) d- d- cr)NCh C") VD N r-• c-r) V") d- r-- Do N 00 C 0000

N VD VD dr 00 kr) c:t VD kr) N oo nr) en c N N oo C d- dt r- \.0 c,-1 ken o r-- ken c-1 cr) N N VD ‹) 00 dr nC) N v") 1r) 00 CD VD r-- oo N VD In kr) (-NI r--- 4 c-A

r- ken On o0 0•n r-- r-- CD ^d 00 dr kr) oo oo oo r-- ,c) ken dt ken 4 -dC\INN ,--(cr)--n cr)

00 00C r- (NI (:), a, ken co; t-: kr; c:; CNC00NC

dr 00 00N v-) CD r-- c') V") cr) 00 r- cl VD N CD N CD VD VD 7t CD dr cr) 71- V') cn cq nr) r- r- cq oo vn vn Ch r- kin VD 7t Ch (4") en d- r- CD 00 N VD Cq CD CD CD CD r- rn -4 cq (-1 -4 r- cs1 -4 rn 4 cn en 4 cl -4 ken

cq CT -4 cn VD r- 0 r- C N un cl r- ,f r- -4 CD v') 00 dr cr) lf) 00 vD) '71" ken N Ch CD ‹, CD CN CN 71- In d- krn .er) Cq n1D r- cn CN) r- CN 0NNOtnCnCn c") r- Li") V') cn nC) Cl CN m-1 1-1

en un nc) cl rn cn cq vn di- en en cn vn (NI vn NN vn

In 0 71- In oo o un VD 00 r- oo N 00 N r- N CD CT • oc; -4 6 en —; 6 n.6 -4 cNi ce; ken' 4 -4 c:=> cr) N cq ci") N

O 0 O 0 0 a) o (...) U ›, " a) 0 ›,-- ct= = ct 0 (...) (1) 0 Cd C..) a) ..= at) › -0 - al ,,..0 .- o ed rcl i DO ,4 , 4=1 , U , , 00 u

.4 1:4 4:4 tD4 tD4 4:4 tD4 r:4 4:4 .4 a4 f:4 1:4

<) —4 7/. dr 00 00 cl. vn. . r-. . C). . CD. dr. CN. . . ON CD cq rn rn 7t 7t tr) tr) VD N oc; cyn cr; 0; 6 6 r -- rs r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- oo co 1 nn1 1n1 1.n1 1( 1-•••1 r•••••1 n-•1 CD YD Ch cl 00 os CD Ch 1- r- cl oo rt Ch un un o CD 1- cn cn --I Ch cl r- 7t oo len 1- Cl CD 7f 1- 1- —0 cn Ch cN 1- 1- un Cl r- oo r- oo Ch Ch Cl yD Ch r-

0 -(7) CD cn 00 rn YD cn Ch 0 CD vn lin cn 0 1- un /- cl vn CD r- cn r- o oo tr) \C CD r- tr) 1r) Cf.) 00 Cq 00 Ch cn cl rn oo 0 YD 1- --I un cn cl un cn Ch cl r- r- r- CD Ch r- CD cl 1.(-) .ce, 71- cncqc, —fc‘ld-kr)cNINd-r--.0clN ,—(r-

75, 0 m 00 rnor o — 0 00 N c\1 r oo - cf) r- 71- g7t- 71- .6 tr) c, -71- tin cc) cl (r) ce,

JZ

. •5 4•-• 00 Ch 1- oo cn oo cf.-) c) In ce, oo • an .6 cr; cNi en r-: o6 .6 oeS cn c:sn cen N

bA

Ch Ch un cn cl oo r- cl un CD un vn 71- rs". 00 CN 00 r"-. kr) 00 ^-4 00 vD oo oo m oo oo cq oo Ch cr) , t tr) vD CA CD r- oo 00 cl --I rn CD r- r- c) Ch CD Ch un cn YD 00 cn cl cl nLD cl cl 71- YD cl rn cl cl un Ch

0 ai) N r- cl vn un r- CD rn --t un tr) n n n cl Ch r- 1D oo 1- 1D 1- r- 1D oo un cg-) r- ce, cl kr) c) cl cc r- r- 7:1- un tr) oo tr) m In vn cl v.'D Ch CD Ch oo cl os ‹) 0 -. 0 c,-) --, ce -. oo Ch rn ---4 CD cn YD CD 00 CD vn oo rn Ch un rn CD ---1 ch nID vn rn iD 1- YD --, —1 un oo --, YD --I oo un --( cl rn 0 —1 rn cl --1 r- \. cl 4-1

•n• c/)

N 'I' cn cl cn cr) 7f- cn 1- vn cn 7t cn cz9

c.) oo n1D YD r- rn YD CD 1- rn cl Ch CD rn oo un un cl un • CD cc; CD CD CD (Nr -4 4 - N Cr) bo czt •

".C) 0 75 z ec) Ii 0

7:1 ••-••• o4 p4 o4 p4 0.4 014 p4 o4 p4 p4 al 7:1 00 U n ri — oo. . .1D. . r-. . le). 0. . m. . cl. CD. CD. 7t. YD. . un. un. r-. rn. 00. cl c:s Cl rn rn 7t 71- 71- tr) YD r- r- r- r- oo oo Ch Ch CD CD 00 00 00 oo oo o0 o0 00 00 oo oo oo 00 00 00 00 00 os oo 00 Ch Ch 1n1 /-1 •-n1 ••n-•1 •-•1 v-1 1.1 1-1 •-•-I •-•.q ,••/ 1-1 1-..1 •-•n4 E Nc,.rn N -I n (L.) rn Ch cn CD -0 d- d- ./D Crn C) r--- N oo 71- 0 ,z) r- N dl- cl -4 CX) ri) rr.) kt) 71- r- n.C) r- co r- r- r- oc r- r- n1D r- r-

ch oo v;, oo v-) r-- oo oo oo oo oo cc) 7t- c (Ni t-- N ff 71- d- ir) co.n cn ir) ir) d- Ln d- CD In r-- r-- CD r-- oo d- d- 0 7t. tr) (Ni (T CD N

00 N •-( 71- •--4 Cr) Ch tf) \ oc oc oc \.0 —1 CD o 71- oo oo s=) r-- N •—n Lr) oo .1- cc) -(ren-,NNcncsiv:Dunrn,-+rnrn-l nCD-1

N 71- oo 71- 7t. 71- in ch oo cr) c=) cr) oo • c‘i cl; 0.6 (a: oc; -4 -, o6 06 oc; 06 k.6 oc; -4 cn

Cr) Ch vn un Ch C -r ccO Cc vD 00 1-",1-1V0NCI)Cr0.00),-,"7:1- 11-) CN1 c) oc; cg) t--- kr) 71- cn te) r-- r- N oo rn d- d- rn C:h o0 -1 un N N r- N rn \.0 -n -I nC) r- N N

N N ---4 r- cq rn -...( rn --( cq vn r- ..(D CD rn r- rn Ch CD N rn 7i- N4D CD .(D \.‘D CD ---4 CD Ch N un r- rn CD un c0 rn --I r- Ch vD r- ch un oc Ch CD dl- r- ---, --1 vn ch vn un -I rn rn 7t. ch nID 0 0 0 (D r- cl --( CD CD n r- --I \,‘D N d- .(D rn 71- oc -.. ,..1D rn rn rn Ch --I --I un d- N ..(D .(D

d- rncnd- uncnrncndtd- cnrnNrndrununrnrnNrnrn

vn ch N rn Ch oc CD

4 P4 P4 4 a4 o4 o4 0c4 4 4 P4 crp n.0 o0 cq. o0. oc. . r-. CD. . un. . Ch. . rn. CD. . vn. . r-. CD. CD. vn. . . n CD cq m m 4 1- 7/- 7J- 4 In .(D \ID r- oo o0 Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch il •-n 1n1 1.-.4 1-1 1n-n1 1n4 1.n1 1-1 1-1 1.-.1 1-1 cc)

• AD 0 r- CD on —1 r- oo C) vD cn ON AD Cc) on CD 71- CN CN CN rn CN Vn Cl CD 7t VD VD VD cc) CD Cl 00 ln lin cc) CD 7h r- VD CD cc) on vn vn trn CD • ". 00 r- op op ", 00 -. r- ch .. r- ". r- n,c) 00 -.

CN 00 CN CN 7t '71- CD Cl rn CD 7f 00 CA AD 00 C.- r- cq CN 00 .. cD 00 CN CA r- Vn CN CD rn 71" CD CN rn cq CD 7t rn CD CD 4 71- n CD C) Vn CN Cl CD CD AD AD cn oo cl cN) on cq c) -. cl on cl 7r ^-1 00 00 C, 1

ON cD r- cp vn v C C c) r- vn r- cl Ch AD .. -. on CN Cl Cl CN CD r- vn 71' Ln CD In Cl C"... CD 00 '..CD Cr) Cl Cr) Cr) Cr) Cr) 0..) ,-, CA CA kin CA rn vn cA cl

ch cq 71- z) cp .. .71- 00 cc) CA Li') vn V") CD CD Cl Cl rn 00 vn 71- CD n • oc; v3 C CD oc; od cn -4 c\i nc). v) CD

on vn cN r- 7t C..". CD d- lin AD AD AD AD AD Ch 00 00 c.,1 r- AD Ch CA lin Cl CN AD r- AD CD Cl CD CD \ID Cc) 00 AD cc) r- cq CD 71- Cl CD 71- CP\ cn CD 00 CD CN Ch AD CD .. CN CD .. on (D • oo on oo cl on ,7h tin —1 cl cl r- cq vp

r- 7r ". '7,- on co r- ci\ (-A CN on on oo cs, cq on vn -. on oo oo r- on --1 cl r- r- vo r- on -. AD ,-. CD 7f- 0 c.,) r- oo CD 00 d" r-- CIN CD ,-. Ch 00 On \..ID oo r- on on oo CD CD CN lin AD AD ,-. AD CA CD Vn ,--. rn '71- ,71' AD --.1 --1 r-- 71- Cl ln r- vn CN ,--i on .4 r- 71. Cr) ^-, 71- Cr') Cr) Cl AD ,--.4

• c•I on on on cN1 c\1 on d- d- cl on d- on on d- cn vn •71..

.. CA CN cc) CD CN 00 CD AD CN ON Vn rn 00 kt.) .1. tr) CD op

7( ,1") CD C•i CD CD CD 4 CD CD CD CD 4 CD CD CD cr a,

i rD4 (:4 a4 1:4 c24( 1:4 i-ÇD4 1:4

• 00 00 c m ch ,. cp —4 71- vn CD CD CD cl on on tin CD vn od o6 oc; crl c56 6 NNNNcn 4 4 .4 tr; n,6 oN cN CN CN CD CD CD CD CD CD CD CD CD

CN 00 v0 r- CD 00 cl VD cc) VD 0 D N r- CD CD 00 cn 1/0 CD nID 7t Cl 00 v0 CD v0 v0 v0 r- C O N cl r- oo N k/0 cl N 00 00 r- r- VD O r- oo \ID r-

U cJ CD m 4- Ch kr) Ir) m 00 \ID le) CO (14-5n 0 00 r- r- O N - r V0 cr) r- . • b.0 cc) cc) CD oo \.0 CD ch V ) \CD v0 c0 \ID v0 "1- r- Ch CN r- 7t VD r- oo r.". = E • cl 7f v0 N cl ‹) t•-, CL) r- CN m Ce) 71-

v0 CNN7tNr-ooNoo m 0 N CD \ \ N \ID N N \O CD Ch oo CD NN0N v0NCh NNNcn

CN \LD rn rn v0 \.0 CD \ID m d- oo r- \O CN 00 CD N 6 o8 cl 6 od oc; N: CD N

N --I CN r- m \c, v0 CN v0 cl r- cl v0 r- v0 \.0 oo Ch r- • 00 Ch 00 cl CN m -7f m m m v0 7t cl ‹) v0 71- 00 v0 oo r- v0 oo cl v0 \ID --I ci 71- V') N 00 n(2) r- • N --- NN -. \ID-CN m N un c‘,1 rn

0 CD Ch N N — CNNNr cn vn rn 71. 11-) • Ch v0 '7F 71- 7t CD CD CD 0 oo r- 140 \CD Ch CN Cl Cl Cl 71- r- 00 r- oo Ch Ch r- cl m \ID CD rn▪ VD 00 VD VD C v0 rn -00 O r- rn

cl v0 v0 v0 m m m cl v0 m 7f 71- kr) '71- Cr) In

CN 7t71- ri- v0cl N0E--v016 rfiNr-\CCN

ChN W *4 = o ›, ›.1,-.-a) a4a) (1) (1) g ,-- u-i' 1-. 10 C.) 0 1-1 U (D Q m =•--, E ›, Q.) Ç...) 0., mt =-‹ 71- vn C .. m -c-"zs -. i N C7 C..) p i ü. , , WF1 , i cl e Wi 0 n o4 a4 ›... ,-. p4 ,-.- a4 o4 )-- o4 a4 ,--. ..-) -. ,-. ,-. s-. a4 ,--. E- 00 O0 0000Nv0v0NNr-0o gONI.,00

n - t 060e5 06 ci c5...c ,ic--- 71:71:4i, ; c) 0 CD CD CD CD CD ,-- 1-1 1nn1 cl cl cl cl cl N cl cl cl N N cl cl cl cl CA N N

ce) E CA cc) VD CN cn VD cc) d- CA VD r- cq cv cl VD C- 00 ) CA CA r- vn N 00 n--1 d- tr) C N VD 00 r- oo r- r- VD r- vn oo VD CN z

kr) N Ch r- CD Cc) VD 00 00 00 cc) CD Ch r- csn cn cd 4.J 71- CD CN C) *1- cc) 71- d' CD 7t 00 CV CD CD In tO CA VD 00 r- CD CN 71- 71- (NI 71- 00 Cr) C:\ r- VD CA N cr) Td- (Y)• N

N r- cn oo r- r- oo cl cl r- CD tr) cl VD CD CA 00 cc) CA CD CN VD CA VD CA CA CN CA CD

VD 00 00 71- CD Cc) 00 00 cc) CD r- cc) CN CA Cc) 06 of6 c‘i crn oo; o6 o6 oai oo; C5 cD

00 ,-.1 Cr) CD "71- r- cn cD cn c,1 CN cl r- cD oN vn vn ,-, CD cn co, tr) r- cq r- CD r- oo r- r- CA CA cc) CT\ CD Lr) 00 (----- CN VD r- r- 00 VD ,-..- In If) CD d- CA CA 00 CN ,-.1 V) in .---1 t---• ,—IN,—(000C1,-4,--i,—+,---(0"NON ,—Irr),-4N)—( N

,c, cin r- .--, kr, 1/ 7t CN Cl VD CN cn VD 71- ,-- d" --I N r- oo r- ,-, oo r- oo C r- ,..o cr-, ,f CN r- vn cl vn ,-, ,t r- ,-, r- ,, it) 0 r- C r- Nt tir) CD , --1 cf) cf) 11") 0n 1.1") CD kr) Nt n,CD CD N d. r--- (-NI n.r.) d- •71- Cr)cr)NN d-,—,NI---Cr) In

kr) cc) 71- tr.) cn 71- N kr)

cA tc) 00 CD (r) 00 CD VD CN cfn vD 4 6 c\i c: 4 6 6 c; c:D V")

= Z = 0 0 0 v) >, ›-, 00 ›, ti) = = 0= —,tZ •5, ,_, M (1) —.1.., (..) .., C.) 0- C) sa. a) a.) c/D a.) --, ct ', ci.) 7,1 o ,C) 0 4=1 a ,, ..;-4. t'.7' (..; (1) N ' 'c."' c N ,. cD ' = N .15. Ri ;-, cl ,',74 ,-(1 cl ci ,c1 E--, L) , , , N , , E-1 1 CA 1/4...) 1/4...i 1 CA 1 1 s..: n...) 1

• c4 fc4 0c4 p4 ••• 1c4 p4 .4 4 .4 .4

r—r- cq v...) 00 71- r- CD VD 71- CN C> 7t cD cn tr) VD tr; If> v:*) 06 06 o: o: 6 6 NN „, -* CA CA CA CV CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA t.r) tr) 00 t--- oo N N oc oo

• rq CN Cr C • cq r-

71- 00 "" "" Cq N.0 r- C Cn V') 7t ""

CN \JD cq "" 7t \CD n • C). CT nC o6 4

cc 7f (IN n.0 oo VD \JD C) • tr)

kr) m N tc) N 0c cfN d- 00

• cl cq or) .tr) r-

U-) 00 C.-- 14-) 6 6 6

124 ic4 c4

kr) ON In oo cc; cc; ccs; tr; vs; N N N N N N N N N N N N Q 00 4 a) 0 . i 00 . 1m.) "0= u Cr) :. cA ,. (11-.)' = 3 2 z 3 m —__,-,°-) — 1..)— z ,... ,.. 3 ,4 ,..z 01) z z\ c.J ct -.:1 0 -c:7,, -.,,,1 0 0 __, ccs, +-:_ ,,0 .e5. . cA tu) czt •-, \.0 1..) F., ...z c.) a) to ,4 1L) ;- c, 5 v) ,_, ›, z ,--- 3 0 ep-, 0 0 a) 1-n 0 -0 -ci cl) .. 2 .- ,..9 c'zi ›, 'cl ai ,--..-. t 4. ., , 0 Q _ „, ›.--,., — , ),— ,.c . »,,,, a.) q.) ¢ 4,-d di g ., , = (1) 0 ••_,. c.,) • -., .,..-c to „, • -, „„, ,...... - , , - ci..) „.., P4 CL, (/) I--, u-) 0.) X vp '' di (4-4 E ,_,-, ion czt u CCS al 4 P -,t, a ) . .(:) P 0 - • - P "d c.) u "0 u M c...C: el.) P , E 7, ,, (1.> • - z ,.= 0 0 tu ct >C 1- - L;_c2, cl-) 1-) cip ''(/) -c) 4Ft' • -, . '17) 11 ,4-, ; m E 2 V It 2 .7' '' 77'1 6 •E — cg Ls'. .,,.., S' 2 .4,1 1_, z "0 17) 2 . 0 E ,.,.. -5

, , 5 (t) 4...' 0 ''".' Cd ° . (7.' ..i ,,_i• nn ,__,0 u 4-4- 4,,, -o- c.), --. 0 '7) Cl, = cl) ,-- V ) lcl v) ° 8 -714, + - ' ., F. 0 ) c, ,,A 'e 0 Ag „ =,;., 1.) — mt F., -,_, -,-;,- to E .,>_:' 0 ..z ro•,_ c,..,,_0 ,. ,a) 'Ercl S=1. bdt,', a) -.) 2 0 0 ..,,,- in, ct E — .t„) 2 1., t ,._„, -6,2, cn .z 0 :5 E vt '''' --- *F (A • - ' — ,, = g.:3-, 0 in, ct, - E ''' ' 4c2 f.,, 4-.' ,...ct ..0 — a) -0 1-I a) -7, n- ct 0 ""C) $-, lo 7 ,t ' 4 ..= , - - , a ) t 0 -5 • , •--- 2 c',' 0 ›, --' -7'7: E ,t r- i —, "t 17; E c zi a) a o ,c, c . ,5-, ,.':' 4 4, ° 0 ,L, . g Q., ,4 = ,._, Q c., , .... . n 11 (/) 0 E ;-„, o z . .__. 41 ,,, ,#, Ct Cr ejj ;.., '' U 1(/3 CD (1.) 2 Ct C. › ,f2 -- -- E 1 z = = -,5 4a' t.3 <;.). iii,;;‘"2 t,b 0 & (ID «-a — 5 z Q .Ï-• = C71 › 5 g E z ..o ..-',., -o 0 § 0 ,, 0 ;.. z > czt ,.. 0 ›, -5 btu.,,,„. E) tu.4= ›, z 0 ,,, tu tu ,,. 0 z >., ,,,,0 z f,,,, C-) › I.) . a) •-, CrI ,0 c.4_, ,,,s = -12 4-)', cl) a) . 7-. cl c..) c) 0 C.) 0 --, • a) c.-,b- -.= s,...., E„ ,..8, u ,_ „ •`- 10 cl a) 4 '.7' c-i-,L? CD 0 0..) V) Z ,e = CI) a) 42 1..)- -,9,_. c, , cd 4 ,-, '' Ce • 73 r-.) s' 4-1 g - a) ,__• ,-- to,z) exo.[n,1 k.). L,./ = c;, ,r) . 7,, .... C.5 , . -0 '4' z 5_, 0 .(7, „, .... • 0 ›,,.....: tu .tr.,' (.1n4 1.) • ,, . Cf) M (/) cu S' . ,..., z 7::1 #., Ct$ 0 4-1 ,C:5, .., ,...4 CZ3 4-1 '-''' c., t c., E .. . ..z V, z -c, ,,f-' ›, ,-, 0 ,- c:N tu 0 0 ';', 2, a) C.) 0 CL) '''' --"1-) ›.., • ,-n ,--4 u L., •-.). (D ..0 Z ;.-7 U V) ..,-E5 t) g:' 4--, , ,i5 -,„ ,-,-, t i-4 Ct CL ) -. -_-: ° C t E . Cd ,-O cQ Ol. ,,,Up z 5-,Ci] V , ' a) •cn ' -' "cu O .,,., a) • , - -1

4 '-- n . I =I) -,-'E ' ,,,-< 0 = Uce .-- Z 9,, 4za3, ,,„..-°-) .,- . cz;• F, -C 0

d' 4 (i) .— u) , 1.) .._• > ,..,'ci.> ct'a' -.,1" . `, 4 = e cf)2 'LI2 0 -8— . :,i4,5)

,4, o. t - 0 m fa, a) (1) ,4,-,0 ..a - 2 c;N '- o , 00 E a) ,C) 0 _. .- -..' '-` a) CS ,..,q-' • ^0 C--) :=1 -9, I-. 0 t+. 'Fzi •.• ---'A, .,_,.q (D cip' .,_ 0 qt,- 0 u. ,4 t.0 F, cz „, , .—Q a) c_.) t4-, -70 a) 2 7:3 „•-•,,z),ba->cli.:ccl 8›,,, ,o(:) ° E = cu 7„c 4_, _ ._ ,_, ,5 ›, z 2 .a C 0 p_4 q.,5 ;-' -"i^n-' . Ci) c/D -4.-n ^ Cf) a) o., CT n 4-1s., ,.. g 2 0 ••• . (1,)") czi -"0 mvp LI _7),L) ":L.). • >.--' cb 00 ,7:j 1-4 2,,) vi ,-- C) 1.) fZ1 I-, u . 0 E 4--n czr,.Ei Z ili S. 0 CL) ,77) 7' : 1 i ,-',' 2bi) ,.. 5 8 ›.., c„ ,.(2 , ,1.1 0.) 0_, E .., . 5 1 p. c, an a) c t cn 0 0_0 , ,.° z. 0..) •-(I) '' -Fd".— "e• m 0 m C4 '-' c: 0 4 c . czt. c•i-. ,. 3 z ,-,,, a) '-' ..-'' (e > 5 0 -d S::). • -, •,-., Q•'F.'," a) (I-) a) cf) 4 •F-9 -8 SZL-n ra 0 4--1 ' 't ti ,=1. CD ;n4 U.1 CZ;.4 c4•7, Ca(1) › :i C\ 3 = tij -2 2 g -- 00 c„ 1,.,,-, ,.,i 0 2 .A,.) .. tu ',71' --3 fa. ' 00 i ,--, cl, t,-, ir, 4_, .., ,4v75 0 ‘a' co $.0 u 0 S=1.1 a c >, z1 -0 a.) ,.S.71, 7-,d u _ c> ci) > 0 ct ;._, -5 ,,,„ a) -F2, -; I--1 c...) ,_c,1 ,:e •c,' CS •lz •,-. z 4-( . r ) 0 4_, S 2V) cA (1) :iz"U ,-cs' y..,•cie .ci ,-, 7) ,_ "0g ,-/i0 47d_,C4 .41) "(19cp z E ,-. "12, ',7; › (1) (L) bA › .-' " ct 2 cr ;-n g. 1) " 0 &' 'y,:0" 9D ' ' ". 5 0 a.) b A z t a .) e - , r.") 0 0 c., 0 t . F_, •,,-.3, . -c, 2 a) g: ,..,ct H -5u '-c=' g r:.E = '4. cd u° ' =m — CN V0

,---, N -- ,----. ,,---. ,'--- ,---. In kr, ,' 0 0 .---. .----. -- —, — N: t--: v-)-- --, ir: r-.: r-: — — r...:, ,--: ,--: — — . ‘_., ‘...... , ‘.....„, ‘..-, r--..._, ,—..,NN ,—,

0 0 0 0 0 0 00 00 a) a) • • • ••••I • •

'C.E1 -r7:1 CI CC: 00 00000000000 (..) C..) C.) C..) . R'' Fi 0 0 00 0 0 00 00 1:121 {.! Et: a) a) a) a) a) a) a.) Pc) PO aa up up up u uuu u a) cv a.) 5 5 or.) (..) COL (a., 0., o-o-c-8-8fa., (4 (4 (4 (.4 t) 0 0 42, CD CD o4 o4 o4 o4 o4 CD C) C) CD C) CD CD C) CD CD 4 4 4 4 A A

cdo Do 00 0000000000000000000000,0 ,, Do Do Do Do Do 00 DO 00 c4 0000000000000000 LO DO 00 00 00 a!i (21 7 121 - 1- a f r= 1-

r- r- lD VD 'AD \AD C0 (40 C0 rn cq CA CD CD CD CD CD CD C) CD CD CD (L) (L) 4 4 cri cri cri Cri cri • N N CA N 6 CD N N N N 4 4 4 4 Cr; Cr.; f:4 IIIII,,,,,,,,

ct 4D r- C-- r-- CD C) 71- 71- rn rn czs ,12, ,--.i ,---4 CA Cl CD CD CD CD ct ,..0 00 00 V0 V-) V0 CD C0 140 V0 4040 CA Cl V0 V0 V0 V0 V0 V0 V0 V0 ("0 (10 C 0 c0 N r- r- VD D I"- r- r- r- C o r- r- r- r- r- r- r- r- CD CD lD \Jo Cl Cl 0 D D n,ID Cl (NI NN Cl (NI

CA CA V0 CN Cl V0 CI 00 CN V0 CN CA CA V0 0') V0 CN V-) \ID Crn Cl CA CA CN CN r-- 00 C N CN r- 00 C-- 00 CN CN r- 00 r- 00 r- r- 00 CN Crn CN CN CN C C CN 0•n CN CN CN CN CN Crn CN CN CPN CN CN CN Crn CN CN CN CN 1-••1.4 1.1 1-1 1-1 1.4 1-4 1-1 '-•••• -6 6 6 .6. 6N- . CN CN -a,. - 0 0) ..... n0 Cr\ CA CD CA CA

CN CN CN CN CN CN CN CN 00 00 00 00 00 00 00 00 00 00000000 00 00 00 00 _,--t--1_ -,,--i -_,. ,--( -.„ ,--1 ,--( n.--n ,--( rn rn rn ..t- .471- 75 ,-743 (-A Q Q. c-1 cl cl -,cl N„ cl N -,NI N-, cl.. -,N -, -, N,, NI-,

V0 \JD r- 00 CN CD Cl (40 7t V") t"--- rn en en cn ct 7t .et 7t .et ct

,-, CV ,-..... VS. c...... -- k..,-, 0 ,--, — — ,---. tr) .----, (-4 -7 0 -- c) ".7.' c•-n -. Cl ,4D G) , n n --:•• c-: ( 1 -" ,-1 -" ,-, ,)", )--, u-, 7" v-7,- (i- rn ,-. -- 4 - (-4 r-- --.r: r-,--, _ .,-- _-- "--.. -1:. - ,- r-Q r- r-.2 r,r: -i trî --' r: kr — ,- -, ,- t---7 , N._, .,_t--- ._. --,_, ._, ._, ..., ._, ._, ._, ._, ...... _. --

Q) Q) 7J 'Ti 77) 7J i., s-4 C•14 1:1-1 = =Z = CZ1 Ct O •00•-• • •-•000 • •-• ci.)1 1) ti) 'Ic. .5 = = r.:4 ctc10 0 •-- ..•- CEi 'F'71 'Fii 'ECI V-1 g ,..0 • c7P . Ff ..• ,•Z .Z cu 0 (-)U(..) QUO cu ,__ oo o o o o UD C1)

00 00 Q) 00 0000000000000000GO DO 00 00 00 00 00 L) 0000000000000000Cd° Cd° Cd° Cd° 00 Cd° °O a a a) 121Dal

• CD CD CD CD CD CD cq CD CD CD Cq Cq Ch Ch CD CD —0 7t d- CD N 6 6 6 6 -4 -4 cNi c•i tri tri tri v6 nc; oc; oc;

• gD gD crS crc czi ch 4: ct \.cD cq rO CD rn c) ,t) rn If If (D tin tin vn tin tin Ch cl cl Ch AD AD CD CD 14") CD (1. 1. cl C ‹) in v..) In in ln cf-) tn cc) tn If cc) cc) CD 7t 87)' Cq r- r- r- r- tin lin cl cl cq cq d- 71- 7t

N In Ch cl AD Ch cl CD --4 --( Ch cl cl —1 ---1 —4 --; CD CD --, CD CD Ch r- oo Ch r- 00 Ch Ch Ch Ch 00 Ch Ch Ch Ch Ch Ch Ch Ch CN CN CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ---n —1 --I --I --f --I --I -, --4 --( ---1 --.1 --I --( --1 --4 --4 ---1 ---1 --, --, --1 "---2( :7-21 23 eT45- - 4 23 (36 N 23 2i..,- 23 ,211 -- CD CD (27,-, 2i7, i:::: 27'7, .-,..- cipl- ceiI -,-. -,-, -,Cl „,cl -,,-. -, cl -, cl ,.., --( -,ccl cl cl ---f --1 cl cl cl cl --...—1 --I--.. •-••.1 •-•-I •-•-• .--1 •-1 cl AD Cl. CD. Ch. cl . CD --,--I CA ,--. Cl Cl-' C71 ---( i--«--,-1

Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Go oo cc DO oo oo oo oo oo 00 o0 00 oo 00 00 oo 00 0000oo oo oo oo 00 oo oo oo oo oo oo 00 00 oo ---., --, ,-, ---,,. ,-, -, ,-. ,-. ,. ,. ,-, . --, ,-, --, „, -, -, ,„ -., --, --. „, -, -, -., ,, ...., -_, -, AD ln 00 oo oo oo oo oo oo oo Ch Ch oo oo Ch Ch cl cl Cl cl cl cl cl cl cl-, cl CD CD Cl cl-, ,..„CD cl-, ,--,Cl cl -, cl-. .--,Cl Cl,-. cl,-, Cl--, Cl--, Cl,. --,Cl r- r- cl,-n --,cq r- cy-.

cD c) oo Ch CD Ch CD CA cr) CD 7t In AD 00 r- In In AD tn AD AD AD r- r- r- r- cq r- r- r- r- r- cq cq ci cq cq cq cq cq cq cy cq rn cy cq cq cl ci

- ▪ • o

Cc)

..--, /*. •---. "C ./-1 s S , ,-. ••-, ..--... c., ,., ,--.. '-' S. , .... ••-n —— S. .••••• n C,". " '-' `:i- •-. C`i .--, N \CD N tri 00 S . • . ,....., . ,--, VD 06 s ,C) ,...." r:, r-: ,----. •^. ,'-' • '-' ,r 4 '-' _ N. .71- '° [--

—.. 'I n ' R -- N 'I,-; '-' D 4 (0-' `__D 4 4 r: c4) c-,"; -1-' C-:-.--' — N r: —, r-: —, — — tr) r: •-• ,-, Cr) VU -: n s...." `,...... / `.. , ,...... , ‘...... / ,-.-, ,...... , ,...--, •••.-..' ‘..-..., ,...-..., ',..-./ ‘...../ `...-.., ,..-./ ' ,..r, `...../ `,...... , ',...... , `...../ '..n..,

"Ci "Cl -0 -0 -0 -0 cf) cd cd FE,' *FE., .a,' . f":1 etc::'3 al. sz, cd cd rz4 -tc -zc a4 1:4cd cd r:4 a4 2 a sa., ,..4 ,--u „,=, = ,,,, u - _ cd ''', ° cd -ci -ci „, cd °ct cztf!' 4 = = a4 r24 'EL,' 'fa,' Q) cl-) .- . 7(-' .-,..Crd .---Cd p4C) a4C) M C 1 T- 1 e3 0 0 rdi -ci c7" (i) s. ;, 7:) ,.1z ,.1 up fn., -,(7: (L) I. s. _ (L) (1..) C.> C-) , , Cl., (L) c..) Cd C1)n cl) 0 0 -z - ct ct u, <1.) O cd cd , c3- c.i fxp,800t)".7g22:-;-,)0044000 c.) v) F-,; „ cf) cr) 2 x x ,, , = z z ,.., 7-1 C-) C-) bf) 00 I)5. -c) "0 -Cl -c),,4 _,C1, (=LIC)C), u0 0 <-), b 0S -5 0 0 (:) cd cd 6 c 6 2 75 75 44 cil, :8 75 75 .0 .0 75 o o 75 75 75 c) = = --, oo cn =1 P:1 oo -- cdo = = CD < = PC1 = = = PO cl n n n n '6 n n '6 n n n n '6 '6

• cn cn „. „. 71- '[/' 00 r- rn un un o0 CD 'd- r- vD vD rn acS o6 c5 c5 -2 -2 4 4 4 If) \C). 06 0.6 0.6 6 N c..1 cl

cd cd g:, rn r- VD vD cl r- r- o0 7f rn rn M cl vn vD vD NNNNCCNOCOC C) CD CD CD Cr) 71- kr) Cl In Cl ln kr) 7r kr) c5. C) 71- r- t• 71- cr) cc) cl --, cl -, Cl Cl Cl cr) cc) cl cl Cl —'- NN NN NN

CD CD cl cl CD cl Cl cl cl cl Ch• Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ch ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 23 „„ ,„ „ „ oc Cq e) cn cn cn cn cn rn rn ren C) rn CD CD CD CD CD CD „„ ,, ,, 5--5 ____5 555555555555555555-555555 5555. -5_-CD5 CD N N cq

Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch CD Ch Ch CD CD Ch Ch Ch CD oo oo oo oo oo oo o0 oo o0 oo oo oo ch oo o0 Ch Ch o0 o0 oo Ch oo oo oo o0 oo oo oo o0 oo 00 oo oo 00 00 o0 oo oo o0 o0 oo o0 ,...,, „, -. „. -, -. -, „. -, -, -. ,, ,-, „, -. -. -. ,-, ,, -, „.... „.„ , , „„ „„ „ „.„ -„. -- „.„ ...„ , ,_ -.... ,_ -... „„ S._ ,_ ,„ ,, cD cl rn rn rn rn CD rn rn -, rn -- n --I C) --I --( CD CD --.4 --( --q 8 -, „ ,, ,_ 555 ...„ ,_ „„ „.... :-. -- -- cl cl r- cl r- r- --., r- --. -. r- r, r, -. -. -, -. -. ,. -, -. -.

Ch CD cl rn 7r vn oo r- Ch Cq CD -7:t rn un VD N 00 On CD r- oo o0 oo oo oo oc oo oo o0 Ch Ch CN Ch Ch Ch Ch Ch Ch Ch CD CA Cl cl Cl Cl Cl cl Cl Cl Cl Cl cl cl Cl Cl cl Cl Cl Cl cl rn

▪ •

<2)

C--, 'as "..-^, ,-----, ,---. Cl ;1".' (7.17' ,•-•-n N- N- Cl ,---, \ID , ,__, r-s. na ,-• ,..-n CA ,-.., ,—, ,—. .71- ,....:: .' " /...... , ,..., <;' .--" ,C) \ .0 '-' <3. /....., /..., 7 ,-. V../ V-1 1--.1 ... (.4 ,—, ,.... .--, 4 ,o, CA --7 ., ,._, , ----. 4 n • ,••••, v—.• ,,-.... N • V5 '—' C4 00 "—' Ci C--. ," VD S . ^ VD •—• y—, C-- — ,..CD tr) .. In ,--, ,N• N.- s • _. ,t CA "-A- C-• .- — C-: C: r-- . NC)‘. 4- -44 CO n30, ^. •—n —. ^. VD N - ...... - . ---, --, n tn r) in -4 t---: v5. vc) Lr1 ,--, ,-. In s s s v-, ,o s s ..-44 ,._..4, ...._., ...... 44 .._.., ,-.., s.-,4 ...., ,-.44 ',4-.44 ,...... , ,-.., ,.....4, ...... , ,...... , ,-../ ,..._. , -.44 ,....., ,.....44

"173 7i al., tal4 (c tz4 a4 u) cl) 71 7i b-0 bA 7C 7C 77 = Z C11 fa, S:), •F14, • a, i• a% -7)0000 ca ca T.; • a-, a-, a, a-, rz4 rz4 Q., sa, 5 ct c'l ct •5-. z = -1:3 -zi -5 -5 a4 a4 4 m m m cl. m • - • ,-, (.) up v) ci ot U ,____m c7i.- -..1:4UUUU g — , fc , i — li ) 02 ,i ). ; , 5 5 5

. 17,(') . 2r,L) • 2 ) CL) Cl" CL) a4 17-1 . CI' Ct Ct C.7; CA C.Z; C74 0 ct ,11..) ,1.1._), 011 CI * •=1 ...' f:). tal (...) L) E LL, 7' 7' V) V) .'--' ---' I)) tri) kfL) .q (1) (1) 0- b c It t at ;73 E ou 4 4 4 kit., „I) ct› ct or, 7.) . b.o . b.o . b.o . txo.,a, , 0 Cl Cl ..9 4 4 gi Clq- cl cq cl ci N UL)U = P:1 E:-o'- E:-: P 4 Do U-,

Do 00 00 GO 00 GO 00 (..) U DO C00 V) Cd0 Cd0 DO U C4

71- vn on on on on vn vn vn Ch Ch d- '71- r- r- r- r- o -. or; or; or; or; 4 4 4 4 4 4 (r) tr; tr) tr) \f3 CS\ CA CA CA Cl Cl Cl CA CA Cl CA Cl CA CA Cl CA CA Cl Cl Cl Cq Cl CC)

7t Cr) czt D 4: Ch 7t 7t CD V-) V') r- vn Cr) cq ,1D CA In 71- ch r- r- cq (-1 c) nC) r- r- 7t 71- Cq CD CD CD r- o 71- in CD CD d- d- tr) kr) n1D r- cq cl r- r- -, cl cl csi cq vn vn ci cl (-1

Cq Cq CD Cq CD CD 7t CI CD CD on on —I ci on on ch Ch Ch Ch Ch Ch Ch Ch Ch Ch nSD Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch „, „, „, „, „, „, „, „„

", n Cl CA n.ID CA CA -O- 1 23 CD CA CA 7h Cl 25 23 'Co') on cl• cq• cq or) on CD CD„, 0 CD CD CA CA CC CA CD CA CA CA Cl CD CD CA CA - - —N---'71" N NNNNNNN

Ch Ch CD CD CD Ch Ch Ch Ch Ch CD CD Ch Ch CD CD Ch CD Ch Ch 00 00 Ch Ch Ch 00 00 00 00 00 Ch Ch 00 00 Ch Ch 00 Ch 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 DO DO 00 00 00 •n n n n n n n N NNNNCr) 71- 4 Lr) on on nCD on on on • CD CD CD r, r, „, NN - -- N--'r- NNr- r,

(D Cr) --, Ch Cl tr) O N CD CA Cr) 7t C tn 00 If.) Ch kr) CD CD CD 0 CD CD CD 0 ,t cq Cr) Cr) Cr) Cr) Cr) Cr) on Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr)

•

Oc)

,----.

„,...", •1 .---. ,----. ,----. Y:) ,..0 CO ,C, ,-----, t.....„ ---. ‘0, ---.. ir'—', oô . r---.- h s h V) ,---. ,---. ,17 --. " ,D ,.0 N —.. C- r : h — ,--,71- . kr, s . •,--, cr• hi Y-, — Y-, h . . h h h-- C--- h h h v,...... ,-i. h: , .._.-vi' —(4 ...._- ,--0 -.._,' ,..._,- 4 ...... ---' ,....., 4 ,..-`-- .....r- ..._.--NCI ..- ...... „ .....„- -....., -,...., - .....- - -, - -

a) a) -cl cf) O ..-- • •--n (cl g L) H a2b.1) 1.) a.› a., Ti Z = "C3- .—C4 .—c/7 ---, 2 o g g a) a.) .",c Ct ,D EEE u :---,CL) .--i'l) "ri "0 ".--,_ CL) (1.) CL) 0 -d "0 ',..., ',..1 0 0 b.1) b10 s-, 0 0 0 MI c71 '-- ' • - u 0- › › --, "r3_,' • C cl >, ›, '0 "0n cl >, ›, ;-' ;-' cn ccS czi al ct cZ •c--) (-) q ,---, • •-, -79 6 6 9 9 cq 4 4 ctuuu ,...., ._..i 4 o4 o4 -2 .2, (3 Q) !=1 'a' . ,LI. ,-, (-) (..) u u cl) ,'-n a) a) 2 2 4- 0 a) ', .2-", 5 g -8 8 › ›- ,= -5 -5 g E, Ea::),..n Cl) ,,,, •„, 0 0 z 2 2 e)C Q)-0-0 0 0 0 m ct Q., 0 1-) (1) LI-4 '' V) V) V) › Pg a4 a4 '(''' (14?) 0 C) < 4:

2 2 c '3 E'S 2 2 '"D)

c) cl cN1 vn vn vn CN CN CD CD CD CD CD CD CD CD CD CD CD 00 6 6 6 CNi cc; ce; cc; V"; nCS o6 0.6 c7; csn on on on on on on on on on on on on on on on on on on on on on on ,t

CD 4D \ID nC) 00 N 00 \ID an g g ch 71- on g g: g -g aD g g vn nID CD CA Ch CN CN cp 71- 71- Vn Vn Vn Vn CD CA CA CA Vn Vn on Vn CD on vn vn cn on VD VD on On Cl CA Vn Cl CA CA CA --. Cl Cq r- vn r- r- „, 1-4 „, CNI N

CA CA Cl CA Cl CA CD CA CA CD Cl On on CD CN CN CN CN CN CN CN CN Ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Cr\ CN CN CN CN CN CN CN CN CN CN CSN CN CN CN CN „, s,s.„ -„ CA CA CA CA On▪ on on CD On On CD On On On On on on On on on on CDCDCD0 ,—+CCDCACCDCACDOCDOCDCDO,„ , NNON CA (1 Cl CA CA Cl CA Cl CA

CN CD CD CN CN CN CN CD CD CD CD CD CD CD CD CD CD CD CD CD CD 00 C> CN 00 00 00 00 CN CN CN CN CN CN CN CN CN CN CN CN CD'n 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ,„ :Sa 4 4C.:»z le) V") 11") Lin n73 VD ,-n ,-▪n n••-q n, -`n ".n •-n •-n •-n n ▪ "••••. "• • ,•••••n 1-1 /-4

un • m ,t r- c7, od un nc) r- oc CN CD on rr vn • cy cl c,1 cq cq cl cq on on on oo on on on on on 7r 7t' '71" Of') Oo) on on on on on on on on on ci on on on on on on on on on

-

CD ,-n ,..n r.... /-, ,..... 1... ,...... /..n 4--n 1/". Lr) C N . r-- N -- -. ,--n '....." —. V:) --. --, R --, 7i. N N N R --. kr; tr, ,-.., 7} ,..-, ,-, ,-, R —. —.. —.• R R —:. ^ R VD cn -...... le, v-) t-... ,... -. r- r- __. ,...... , ‘...... , ,...... ,.._.... ,...... , ,...... , •-_.., ,...... r-:...... ,-.....,r: .._., r-- -_-,r- ...... ,r- r-,...—, t-- ...... ,

Sa, ta sa fa, (a., (2,-, = ci.) CL) a) O.) 1.) CL) cd cd cd 0 0 2 2 222 4 P4 s:4 >, --, 0 o oUt..“..)c...)C.) bz as) bsoU c9 cl, sa, sa, sa, p."iil u z -il il almcdczs ct 1.) a.) a) a) I) ci.) •— • — • — c_, C..) U g a.) a) a.) a.) cp ni -C) 7:1 o - 444 ) c) ) - cl) (1-) ' '4-- c'71', 3 3 3) (.!..:) (::.) c.:D' c.:..") ci (..:-.} a) 0..) a) 0 0 0 0 0 p., cl.. pL.4 p.. p.„ p., O .-tt tt tt i4 '-'( "i ct ct ct ct ct czt cd cd .--4 -A co (1) (1) 1..) d) 0.) a 'a' 44 I CI) (1) a) ;: -') Z ; L-1 a a cd a cd 1..) 0 q..) ;.. :•- 0 0 0 4 4 4 4 4 o o o o o o '7) -Izs -,z) >a) 0› ,..) o › › › a.) a) a) a) a) -- .-- cn (/D cn 0 o 0 vp cc) = = 1) cl" Tu"' 42, -0 -0 -0 -0 -0 -0 . -.5 ..'5_,' .-.5' .t,::1 .3=,::1 cd cd ct C( 4 4 4 P:1 ,‹ ,‹ C) C) < < < ,-. ,-. ,-. 1- 4 4 4 4 4 4

,t-S ;cf .,cf '6 2 2 2 2

000000cC00000000tf00000000 • 2 2 • • • • • • • • • • • • • • cc; or-; cc; 00 po cl cq cA cq cq II d - d' d- d d in Ln kr) kr) in V) tn V") In

a) a) N Cl cd 40 ,CO cd .0 .0 cd cd rn cd ,CD cd cd cd C: 0 r 0 10 Cl (NI N c,4 cl Cl Cl c,1 c; c: et 71- V.) (1.) cn rn en cl cl Cl Cl 7:1- rn rn r- Cl Cl cl 7 1C 71- Cl N r-- r-- N cs1 cfn cc) cf') 1- Tt 71- r-- r-- N cq cl ,„

— N c) rn cq rn rn cD cD cD cD ch ch ch ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch o -taai ,„ ,„ n 0,.) ce) ce) d' Ce) d - d- d d" d' d" d d d- Cl Cl V-) ln V) Cl ce) in aj CD Cl çR 0 (-,1 c cl CD cl cq CD CD CD cl cq cD 1:2) N N e) e) e) N (-NI cl e) N N -, -. e) e) cNi

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 0\ Ch 0\ Ch Ch 4-1 '•-• 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 cs• ,--1 ,--q ,--( ,--1 ,--1 ,- ...-1 ,---1 ,--1 ,--.1 .--1 ,--q ,.- .--1 ,---1 ,-..-1 o pz ,, ,„ -- ,, pz pz p...... ,, ,, ....„ , .„, ,, N00 00 00 OC 00 00 DO ,..1 e—I ,-I /-nI /-•-I r...-I /-1 1-.-I 9-1 1..... 1-1 1.n1 /-...4 1.-1 1-1 n-... \ \ \ \ \ \ \ \ .....„ .....„ \ ...., \ \ \ n, ,-.1 ,..-4 ,...1 ,...-, ,-, ,- ,...-., ,-, 1..n. 1-.41 1-.4 1.4-.0 ,-. 1n4 w-I /-n1

cri o E czs „ r- po CN CD cl rn 7t r- oo CN CD tin V) V) vn vn vn vn vn n4D c'Z'4 rn 7t rn rn rn rnvn vnrn rn rn rn rn rn rn rn rn rn

• ▪ ▪▪

N ,-.., ,n ,,....., ,_, .,---. '' Cl N N 0 01 n n-n N kro ,-- ri ' 1.--- r.: ‘.0 n r• ‘15 - „----. ih, h: h: h: h: ,-. '--: n-, h: \c, - . h: 4 N t--- h: r"` '-' '-' h: h: 'd- v-) Vi h: .1. . h h: h: ,-, R _'-' ., 7 .4... _Lr: , '-' "-' N. '-". . 'd. ^ ,-. h- ,.6 r-- --1- r--- r-- . 4 -N N- ,-....,- ,....-.. •,-, ,...... , •..... ,----, •-....., ,...... / ...... --..., n...... , ,....., ,.._, ••.--, ,..--, ,...... , ,......

Q) a)Q) › › ,-.4 a a a 'a a) a) o o o o Q) 1..) 3.-1 I-. cd cd c 0 Q) Q) 0 s. 1-. • > .-)' --j cd -,-a, + n ...4 8 800 ut ) c .2 ., . . 62 '62 4 O a a (1) ›, ›, ›, — — ç-) C) o c) o oo o 00 lot) 2 q a a . -0 -0 -0 -0 "0 "0 m , cll ci) ccd ccd cd 0000 a ac ct n 4- 4-+ 000 000= = `—duuc...),66 00 0 cd cd cd cd 4 4 E .,:rE z z .0 .1) ill ,21 0 0 0 u cu a a a 1) ci) (1) a) (-)C-)C--)UU U (')UU(*) • > t4) •,_. . • (1.) (1) cll <1.) (1.) C/D (/) LID O 0 a: a a n .';' 0000 4-, 4-, 4- nn CZ 4: 4: -fD -CD ,JD a: 4-, 4-, 4-, 4 c1-3 _;.-1 d < 4 -t < d

oo cdo oo 00 C) 00 C) oo 00 oo cdo 00 Do oo C) C) oo 00 DO Cd0 00 :D :D

fi Ci 1.- ^ 124 ri *-‘‘ r-4 n-4 a4"

0000000000 00 CD ,--, CD CD rs N ri" ri- 71" 7t 71- AD AD AD AD • C,1 tr; .16 06 06 06 O o; 6 6 vn vn vn vn vn vn vn vn vn vn AD AD AD AD AD AD AD AD AD AD AD AD

C 4: 4: 4: r- r- 00 C -0 g: .0 .0 4: 0000 C gz ,a 71- rt cc) cr) N n4D Ch Ch CA CA r- cl N r- N c) 'n 'n 'n 'n AD nID cq 7t 7t cq -1- 7t on o0 cq -- CA CA le) V-) 7t Cr) Cr) Cr) Cr) • (..1 Cl Cl (-1 '—' CA CA CA Cq

cl cq c.) -, o rn cA CD rn CD CD CD Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch "-- rc) V-) V") V-) V") V-) V-) V") ln V") V") le) cc) 71- In• 00 7f Cr) Cr) CD'n • c) c) 0 CD C) CD C) CD CD CD CA cq o c.1 cl cl cl c) • N cl c.1 N c.1 C-71- cl cl c.) 1-1

CD CD CD CD CD CD cD CD cD CD CD CD CC CC cD crn Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ,-1 .--( 1-4 ,--I .-I 1-4 ,-s 1-4 /-i .--i ,-I .--4 .-4 ,-( .--4 ,--I „,-.1 --, ;t5 ;6 cô CZ 25 C-) 23 25 23 23 25 23 23 23 23 23 -. c.) c.1 c.1 (NI c.1 cq NN NN CA ,, ,,-, ,_-, --,-, c.)-, cq .--1 .---( ,..-q — ,--4 ,-.1 — ,-1 ,--i --I — ,.... ,--1 ,--1 ,--

N rn ,t vn AD r- 00 Ch CD --I cA rn AD n1) AD AD AD AD AD r- r- r- r- r- r- r- r- rn rn rn rn rn rn cn rn rn rn rn rn rn rn rn rn rn

▪ ▪ • ▪ •▪ ▪

Cr)

,------d------,--, •,---- c-i ,c, • - ,--- ., c-A cA .--, N t--- l"-- ,-. —,

,r) 4 .---. • - '71" r-- 71' r-: " '---• r-_..-- t--: cl-,_ -1-- ct, 4- - - - - nto. v-i. r--- r-- no- - - - ..-- .._.., ...... ,.._...• ,...... -• ...... , ,.._.. s•-• --..,

•-• ..-4 a) a) a.) a.) 77 j-• 1-., L) L) SZI• Ct cr) CA C/) c, ,--4 ..- (1.) (1.) . . , - , - , - , .. el ) 0 a . .) 1 . ) 124 .-- ,-- = a.) U g) (1.) a) ,•zi "cl 2 cl.) ;._, 1..) ci..) • . — z a..) a.) (D ci.) a.) -c -c) (1) ei.) cp 0 0 0 0 ,J.,, 4... ;--, 1- m 5i U L'—j . ;_. ›.., k....) CJ U U C..) • 4 ,,t :8 0 . 1( . 9,,,t t --(1)(1.) C..) U C/) C/D CID .-C-'d (.4 v) - u CU '/ CO a.) a) 0, ra, Q 13 o4 o4 E-. (:-.5 ci czt ct v) c/D u u u u - 3 ,_, uo = 73czt 77) m •-cl "I= ct "z)cz g u t --4 • --, m cd mi c d cA c/D C4 Cf) C/) (:) 0 E E = 0 M M 4-t: -<41-) › ›' • ;-'--1 7. "",7'17d7d --(5 77; 0 0 <9 g 75 'A 'A c,i LI LI il LI g P. P. P. a. PP =C...)C-)E-(E--, fil ==(...)U

GO 00 00 C.) L) 00 00 Q) (..) 00 oo 00 00 c,0 u0 C) 00 00 cd° U0 DO c4

r— r- un un vn un vn un un un un on on cl cl vn vn c) un oo on on cc; roi tri ue; tri tri te: Ir; Ir; Ir; kr; n,c5 o6 oc') c3; c3; -4 lD \-0 lD lD C-

C: tc-) Q C C g: g: c d (.) r-<74 ct CA CA CD CD 04-) on oo oo lD ,t on Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) ,t Cr) Cr) un un ,t ,t •-. 71" 7/- 7t 71- et 71- 71- 71- 71- 7t 71- on on (NI c‘l

CD CD CD CD —4 CD CD CD —1 CD Cr) Ch Ch Ch Ch Ch Ch Ch ON Ch Ch ON Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch

CA• tr) r- Cr) r- r- 71' ‘.,0 71- '71- 1-7 tc)• 00 00 00 00 00 00 tr) CA CD CD CA CD 0 CD CA CD Q CD Cl Cl CA CD CD 0 CD C) CD CA Cl • CA Cl CA CA Cl CA Cl CA CA Cl CA CA CA 4-, Cl

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD Ch Ch Ch ON Ch Ch Ch Ch Ch Ch Ch Ch ON Ch Ch Ch Ch Ch 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 •-n n n, • •-• (.1 cl N CA Cl CA CA CA Cl Cr) Cr) Cr) Cr) cy cl cl c,1 c‘i cq cq cq cq cq CA Cl Q Cl CA CA •-• •-•

00 0— on 71- un r- oo Ch CD CA Cc) 7t \SD t---, 00 00 00 00 0000 00000000 Ch Ch Ch Ch Ch ON C1 cc)c) c on on on on on Cr) on on on on on on on on on on

• -

,-, ,-, VI) c/D ON -. ,__,N ,--, N ,-, N' N ,-, (4,--. — ,-, .. ,--, ---, ,- --, N,-, 4 P:1 r-- ,t „, — — — 1--- in r .... lr, _. .7.r r -... ,..,— r'.. /..-.. N ,-. Ln' ,-. ,01D, --, 00 :',..1 .- CA V5 ... 11-, , kr1 ,,,6 tn .---. .•-n - -. ,,,-, .- r': cl' Ir, ir: Lr: '''''' 4 tr-7, - vD — kr, — (---- ...... 4 — I-: t-: 7.... t-: `if! t--- r-- 1.ri 7r . 'Cr r.". .-.1 h ,-. •-n ,-...... , ,,,, ...... _.,r1 ....,r""• r".._. ._...... _. ._. ._...... _.r-.... ._.,. -.._. -.._. -. ._. ,,,, ._. ._.-- ._.—

g

cd cd z o4 o4 O000 cA >s, cc cc' UL) 442, Q) a) (L) pc-1 • • •-n • •-n • •-1 42,

472, uL) 111) Ifj vi) Ti 7.)

L) Do c.) 00 c,0 u u 00 00 00 00 00 00 00 00 00 00 U0 C4 00 C°°

kI I kr) r- r- r- cq r- r- cl cy kr) kr) VD VD VD VD r- r- r- kg-) c,r; cr; 4 lr; kr; tri tri tri tri <3 <3 r-• r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r-

42)

V-) Ch NN V") 'cl- 00 NVLC NN czs 71- 'C 00 (C Ch cN ,r ,r r, d- d- rn d- d- ,t r- r- r- ,t ,t 7t ,t tr) kr, r- r- ,f ,t 7r kr) kr) r- r- 7t ,t cq 71- ,t 7r 7r 7r kr) kr) (NI N r- r- r- •- t 4.) 0 c4 .5, --1 --.4 0 0 rn 0 --4 ----1 --I --.4 ,--I CD ,--1 Cl rn —4 --I 0 0 0 —1 ,--i Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Q)0 Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch .---1 ,-; ,--1 ,-( ,..., ,-.1 ,-1 ,-1 ,--, ,--( ,...-1 ...-1 ,..4 ,..-1 ,-1 ,-.1 ,-..-4 .--( ,--1 ,--( ,--, ,-, S. Q I n ---CD V")--- .---,..° --- VD --- VD ----' CD 25 a5 i::::. 25 ia--) .--'------,..... 25 ,--1 --I ,.....---,,--( Ci Cl NI Cl Cl ---1 CD Cl ,---, ,- CI Cq Cl CYN Cl i7,71 ,---t ,--1 N ,...-4 C.-1 -C-7n1 'i) i'. -C.) ,--1 -C-- ,--1 N Cl N ,--1 ^-1 ,--1 Ci N Ct 00 bi) 0 4-, c.) 0

" COCO CCOOCCO C CCOCCCO 0 01 01 01 01 01 01 01 01 01 01 0\ 0\ 01 aN 01 01 01 01 01 4-n • 0000000000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 C/D CZ • 1-n1 1n.1 1-1 y-g ..... ;L1 01 -)-- kr) r- r- r- r- r- 00 cf) cr) cn d- d- d- 4 4 71- k7 00 CNINCNIN CACA NNNNN N CINNNNNN •—q ...A 1-1 1-n1 /-1 1-1 1-1

" 4 W A.• t"-- 00 Ch 0 rn tr) r- oo cs-, CD m ,t kr) Ch Ch Ch 0 0 0 C cD 0 0 CD CD CD Ce) Cr) Cr) d- d- d- d' d' d* d- d- d- d' d- 7t d- d' d" 7t • , • -

00 )

,---, ,..—", n.0 /...... \O ,--. .. ,0 PC1 n Cl' ,-. .--, ,-- ,-... --, "-",---.^ ,...... , ,-.. t.....7 ,----. C.-- .7t0 in -- V.', Csi rn:, v-, (-1 s c,i' Ir: c‘i - 40 — r-: CZ-N r-: — 4 4 4 ,-'7-,' s:' s ,-, 4 ,...---....,,...--,,..../ s•-, •.--, s.--, ,--, — ,---, s.... ,...." ,..-., n.--, ,....., ,..-., •.-.-, ' ... ,...../ ,...-,

"0 "ID 70 "c) -0 ,-- .-4 f:D., Q (1.) Cl, i E1,' . 5 .' n''S 5 '. QQ Mi MS 'CI M M M a4a47- IC4 P4uu Q ell c71 (:4 .^() "C) "Ci I'Q 7:1 7i 0 at ci.)'-' 00 .7.: . 7,' CA OA Of) › › c0 C40'--cc-accomMMOCDCD, 7) 7 D 0)0 0 • M M— -, . . 0>-, >-, .4 1244., 7 c c • 9 ?j,-8 ''' 4 . §-' t t'F., --,,.; -,-,; -,-; q 20 (..)(...c....,Dfa,oucuuu .„7,b.ot,ououoott eu c..) 0 <1..) 4-1 4-1 0 0 „...,, (7) ,,, (7)m,c4c4 0 0 0—c4 uco uu= 00 V.c....'-j u u 4 c. 1. I., ), 1-,, 1, ), 0) ci.) > > > > E E O a.)'-' a.) a) H O - - 4 -c) 3 3 3 8 › › (1..) a)a) a) 0 0 -t -- 000 -„) 00fm.,10..,C1...... ,-.,0 (1) cn 8 8 cd ô ô 7} 73 7j a:, ,-54, -c 42, c, `,1 i.1 V, ô ô u o o o 0 0 6'',. c/D UD ao ao = C) C)

0 uo 00 00 c,0 cdo cdo c,c) cd o 00 00 00 GO °° 00 VD O Cd° C4 VD VD V) Cd° Cd°

O

• r- r- oo co oo CN Ch Cl Cl cc) cc) v-) cf) r- r- o Ch Ch CD CD O > ✓ 0000000000 (7; -4 -4 -4 -4 -4 c‘i cNi 4 4 r- r- r- r- r- oo oo oo oo co oc oo oo oo oo co co 00 00 0 1:2

42,

• M M M M gD M M gD M 4D C) Cl) u gn M ch Lic) N r- CNcq CNcl ch cc)Ln cnIn

o NmciciclCicl)—(mNchINNONNO—;m CNCNChCNCNCNCn ChChCh CNCNCNC...... n CNCChCNCT .....n Chh Q o ONCOCCO—cclCOCCCICON—IN CI oo o Q 0

aLt CD CD CD CD C) CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD C) CD 0 71 CN Ch CN CN Ch Ch Ch CN CN CN CN CN CN CN CN CN CN CN CN CN Ch CN m= oo po oo oo po oo oo oO cO po oo oo oo oc) PO PO 00 oo oo co oo oo

-S-- SSS5SS5 00CN r- oo oo oo C).S-6 O-e) a.\ 742., 4 4 4 4- 4 kn 00 cl c-)1 cv c-1 ,„cq cl cl cl cq cq c-.1 cp c) c) CD CD CD cq cl c.)1 cA cl c.1 cl cl

O wo • r- oo CN CN CD cl cc) •71- Lr) r- OC CN CD cl cc) 71- Le") CA cl cq cl cq cq cq cl eq cn cn cc) cn cn cn cc) 71- 4 4 4 4 71- .4 4 4 71 - 71- 4 71- 71- 7t 4 71- 7t 71- 4. •

cc)

r-. .----. ,.--. GD 0 r-: VS' ,t5 ^ .---.. .—.. • VS. r:: ,--, ,---, ., ., c.,' n -1- 71:: ,.....,...-, 1..., N N N ..t. r... ,•••• r h: ir-, : '-' :j•..' ,^...., ., , In z. ,.----, ,j- if-) c•--, r-- In '''.. -- oô --- t---.- - d-- -, ,-----, d- r--- -"7 7 ' t-: ".".. -", '-'.. r-: t-Z r--. n t• r'-.. r--- _ ,.... ,..... (..., _— — y-, ce — <1- —t —.. ,—.., •.--, ,...., ..—.., ,..--, ,....., - ,...., s._., s....-- ,—, ,--, N—.. , ..—, ,—, .....- ,....., s.-- s.-- .....,

,-- ,-- Cll CL) •,-, • •.-n ;-, 1--, clCL( alczt , C..) U .- <1.) ...... - 7C -Ci P4 124 8 F) 2 C1) CL.) = . czta . czta (1.) cl) • "" ' ."1..11 (....) 0 Cll 0 C9

•'-' • """ C71 CI u ,—, CL) >u 0 u u u (..)u 6" ,00 00 0e,0 1 0,. 0,. Q 0 li CL) CD CU ci) ,9.)9-2n n ,.? E g Clà 0 ' Tj 7..i ,C) ,0 Ct • ^::: ,Z-1- '-:- .=-- ,---'-, ,c ,.,0 CD.. CL, CD, .—,CZ, .1:: ,.D 0 ov c.) L) cla m d < = P:1 6-6 6-6 66. 6,-i

u Do Do Do Do 00 GO GO C.) GO GO GO °0 Q.) 00 00 00 00 00 00 D D

)-1 1-1 1-1 •-1 )-1 )-1 1 .‘ 1-1 C4 r.1 (:4' f=4 f:e )-1 1-1 1-1 1-1

• V.) tr) Ve) 71. 00 CN CN CN CN V") V) 00 00 00 CA CN CN CA CA

mr V-) tr) tr) 06 oci ce, oc; 06 cs-n 6 6 00000000000000000000000000000000000000 00CCN

• czs 00 tr) M CN CD Cr) CD AD AD Ir) (r) N 4: cd M M 4: Cq Cq C) CA CA CD CA Cr) le) Cr) le) r— r— r- cl cq cc) r- r- r- ki) on on r- on vn r- vn vn 7t vn on on r- oo oo vn vn vn „, cl In cA cl cA Ite) (C) CD r- r- cq cl c.1 cl cq

,--, ,--, CD Cl CA Cl CD ,---1 CA CD CA CA CD 07) ,--i CA ,--( CD CA CA ,-- ,--i CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Cr\ CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN ..--1 ,-1 ,-i ,-1 .--1 ,-( ,...-1 ,--1 ,.-. ,-1 ,-. ,-. ,-( ,-I .--1 ,-.1 ,-..4 ,-.1 ,-1 ,.-i ,-. ,-...... „ ,, ...„ .....„ ...... „ ...., ....., ...... , ....., -, ...... „ ...... „ -...„ ,, ,, ...... , —...... , ,, ( ,71 CA Cr) 00 00 00 cr) Cr) 00 Cr) 00 qD CO N m r— on on r- i_::, m m „--I --I --.cl CD-- CD CD cl „--I CD cl CD n--1 CA CA ,-1 ,-1 ,-1 _.cl ---1 cl -. cl -C„-1 . cl

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CP\ CN CN CN CN CN CN CN CN CN CN CN CN CN CS\ CN CN CN 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 •—n •—n • •—• ,„ le) le) V") le) 47) V') le) 1.(-) In AD AD CD CD CD CD 5;2 CD CD CD C CD CD CD CD CD CD CD CD CD CD CD CA Cl Cl Cl cl cl cl Cl Cl Cl Cl C- -71 e)- Cl Cl Cl (--n1

N oo CN CD ^-c CA on tin r- oo CN CD Cl cc) 71- Ir) AD Cr) Crl on .71- 71- 7t 7t 71- tr) tr") In V-) ln (r) Le.) 71- 7t. "71- 7r 71- 71- 71- 7t 7t rt 71- 71' 7F 7r 71" • • ▪ ▪

Ce)

,...... Cn q:, ,- -, ,-, • ., '—s ,...-, 00 ,---n r- — r-. r- VD ,-, ,-, ,6 --. \CD VD Cr, , s, 7 ‘.0 4 r-: r---. v--, ,-. N (-.1 N ‘,0 N N . • N N N -; \c7. [ -- - - __-, - __ -- _. 1.,.: -- c.,1 tr-;. - r - - t - - , 0 - 4 c - - : li-, 4 t----:' - ,..6. 1--.: ,,,, v:: - ir. (---- -ir, ,--,-- r-- ir, ---, kr, ....._.- ..,-,- ...., -....,,_,._. - 4 ir .....,--. ._, ,-...... C ._, ,....., --, ._,10-, ...... ,-. r-,.., --,_, c..1 ...-,,_, r- ....,-- --._, --

z z o .2 "c) "Ci 771,) 72, 7,ZJ i6 i6 ill —,1 ,s4 li 70) " t:', CI, • •-• ), S21., Sa• (1.) (1.) 1..) (1.) C) (L) . 'r.)-' • Fi . 8 6" ct ct M ;-.. ",,--, ", Cl) (1) '-' '-'" 0 0 aC2 2 f2 Ic=d-' t t 4 o4 4 Q) Q) L) L,, c:j 0 0 O 0 =-, , P4 ,_, ;_, ,9 0 0 .5.12 s,) 1) - - - - -I-, 1-i 3. (L) (1-) ,.

H E ( „DE -9 •E(I. __,. F;) , -'0' ( t 4 4 4 4 I') (j) ) (1") 42' ,-C E GO GO CD. ,CD =-. CD cN c, cN ch CD CD = = C..)

c4 00 oo 00(21 oo 00 oo oo Do

• vn vn vn oo oo on on on on on on on Ch CN CN CD ^-1 00 00 CD CA CA CA Cf; rf:) Ce; 4 4 4 4 4 4 4 V"; AD AD AD AD oo; CN• Ch Ch Ch Ch Ch CN Ch CN CN Ch CN CSN Ch CN CN CN CN Ch CN Ch CN

cti 4n \40 r- on oo o C czs c.) ch ch on ,f in d- CA 07 CD tr) \,1D \.4D Ln AD v.) tr) V7 V7 v7 r7 r7 00 AD 00 00 r- r- n r- r- 7f 71- 71- 71- fD AD V7 V") CD 71- CD CD tc) Ic) 7 1- 71- Cl Cl CA CA CA CA CA N N

CA CA CD CD CD CD CD CD CD Ci CD Ci Ci Ci Cf) CA CD CA cq cq cl o Ch Cr\ CN CN Ch CN CN CN Ch CN CN CN CN CN CN CN Ch Ch CN CN CN Ch CN Ch CN CN CN Ch CN Ch CN CN CN CN CN Ch CN Ch Ch CN CN CN CN CN ,„ ,„ ,„ ,„ ,„ ,„ ,„ cN CD Ch 00 00 00 00 r- oo oo oo oo oo • c-A cq cc) m cl Ce) Cq on cA on CD ,„ ,„ O.00 00000 N N N-NNNN NNNNNN

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD (D CD CD CD Ch Ch Ch Ch Ch Ch CN CN Ch Ch CN Ch Ch Ch Ch Ch Ch CN OC 00 00 00 00 00 00 00 00 00 00 00 00 00 0000000000000000 00 00 00 00 • •-n •-• NNNNNNNNNNN 00 00 00 00 00 00 00 CD CD CD CD CD GD GD CD CD CD (D,„ CD CD CD CD CD GD CD CD CD ("1 Cl cq cq cq cq (NI cl cq cq e';'1 cl cNI cl cq cq cl

00 CN CD CA Cf) 71- V7 AD r- oo ch CD cq c tr) n0 71" In V7 AD AD AD AD AD AD AD AD AD AD r- r- r- r- r- r- r- oo 71- 7t 7t 71- 71- 71- 7t 7t. 7J- 7•

▪

Cc)

,----. n-••n r-- ,---. .,---, \ D '71- ,-.. r-- ‘.0 C,i s ''-' •- in s -- ,---, • — .0 s: — ,---, ,---. ,---, ., N. c'•-', Ac — ,,•:, cn — in — -- s — r---- 06 s: • s: ., ,- r- .---, . N - ,--. : C.", „---, v-, ,..0 m cn --:, ‹; 4 ,.---, ,•-• -7 .-. N N. Lr, ,---, -- 4N .. —: 4 .— —. t---. 4 —. ---, N. _ .—. 4 t--.. t"-- -, ‘.0 s- — s s , i. -. ,..0 v-, 'ON v) -- r--- 1---- -. ,-., ,--, ,....., ,--, •-n n-..., ....." ,...-, ,...... , •-., •-...., ,...... ,.._..., ..-, ..-, ,...... , ,-.._., ,..--, ,.....,

•"t. al-, e t:i "Ci M 2 z = z ct ct a4 124 .7) o o o -ci -ci ,-- -ci -ci -d -ci -7J :4 :4 0 0 71 7) s , cd (9., 8 -,-,el -,,-, me! 04 0.1 1..) 1..) 0 0, 0-, 014 ta, 0 ,__ ct ct . ,_, . „ ,-, (-) (-) 7-7 Ct Ct i. 1-n 41 CZt Ct CCI Ct Ct 1.) cl) bo to f:L, fa, +c-71-• 6c., 2. 0 0 48 4 4 C) C.) C) 4 p4 4 4 4 _2 2 < <ç Do 2 z,..,, I.) cll • li -8 71 —8 —8 —8 —8 —8 L) Q) tu (1.) a.) ',.a 0 0 as ct › › 2--.) -2 fa, c.) L) L) /:::,,2 v) c.) O 6 6 6 6 6( 6 6 6 6 6 g00 ,..0c7i ..0 bi)ct t.c, 0 -0 cd _ -0 -0 E. C..) C..) Q) (...) (..) Q.) (..) C.) C..) (..) E--, E-.

L) 00 00c,0 C..) (..) C.) C4 U0 00 00 00 00 00 00 C/P UP U0 UO 00 00 Ci) Ç.) 121 IC)

n n CD Cl Cl Cl Cl Cl CA cc) r) ,7t cc) cc, 71- 7t cc) ID 4D 06 ai, oc; oc; oc; 06 oC oc; oc; 0.6 06 crn C 6 6 6 6 cNi N cr-; CN CN CN CS\ CN CN CN CN CN CN CN CN CN 0 CD 0 CD CD CD 0 CD CD

00 00 -. cq N C - 0000 CN CN ,,c2, r- N (O LP) cd (D Cl (O r- r- r- r- r- 7t 71- 00 00 cl Cl CD r- • 7t 71- 71- 71- 7j- .7j- V-) 71- r- r- CD m Cr) .7t r- r- cq 71- cl cs.1 cl c-.1 tr) cq NN

CN CD CD CD Cl) CC -7 -7 CD Cl CD 77 CA Cl -7 -7 Cl CD Cl Cl Cl r- cr., CN CN CN CN CN CN CN CN CN CN CN CN CN CSN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Cr\ CN CN CN CN CN CN CN CN CN CN

SS SSS SS.5 0 --• 7d- 71- •—n oo ii") CN CN 71"• 71- CN cN ch cN Cl CD, CD CD cl CD CD CD CD CD --I aiS es) e) e) e) e) e) e) el el el e) e) e) e) e) e) e) e) e) e)

CD C) CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN 00 00000000 00 oo oo oo 00 oo oo o0 oo oo oo 00 oo oo oo oo oo

-4- 00 0000o0 cl cl cc) cc) cn cc) cc) cc) Ir) tr. 1.1 1-1 1-1 1-1 1-1 •-n4 r-1 CD CD CD 0 0 0 0 0 n -S -S \ \ \ \ , \ \ \ cl cl cl cl NNNNN NNNN NN NNNNN

-00C\r- CD cl C,1 cf-) o0 CN CD Cl cc) If) r- 00 CN oo r- r- r- oo oc oo oc oo 00 00 CN CN CN CN CN CN CN CN CN 7h 71- '71" 71- 7h 7h 71- 71- 7h' 71- 7t 71- 71- 71- 71- 71- 7h 7t

•

cc)

/-, ,-...., ,-..... h .^. N AD ,--, h

C'`' r -: ,45. ,..., .--n ,-. /.., .-. h h ....--. ,-..., "...,r r-- -: E--: -,- N -d- in' -: 7t N h s -.. s - s 1--- :--- „—.4 --..4 ,--, - —- — — u-, (.„— t.- 4 ,, ,....,, ,_—...... ,N — ,,_....—_, 1--:,._.—_..,,...„--,,,,...„—_.—_—...._,— r--: t--..: r--: —N g" r--: t--: — N.' N N

,..c00,...,..t • ,_, >-, ›-, . _-_,- . "" CL, CL CL = z O "Ci "ti "c3 0 9-'qc9"-m6-'MC41 241=4›.,>,M M 0 t4 Mcdc9-' 1:4 C..) Ç..) 124 C4 c/) v) cn = c/) c./) CI) M M "Cf "0 CL, CL, fCL 0 1-60 ).-,-1 J-61 7-1 • 7-7; ' F-,' 0 t) C..) 0 a) M M M 1.) (1)(1 5, 5,.Z 5_,..5, UU TI:FLE5E O . . . . „ . (41== M M M M M M M 45 't' a 4 a 4 'a' a a a a CI . ) 0 (L) C..) L) L) (1Juul.) 1.)›» -ro co ==(1)0,1)a.")-0,sz._000ctc71-- 0,-0,-0 E" E"e"ee" 000 m Vc w ct CD ,- -. -. a4 a4 c4 GO DO GO 00 GO < < < Z a0 PQ P:1 = =

00 GO 00 U0 Co0 00 CIO DO C-.) U OC cfOC D v) ci) c4 Cd° Cd° Cd°Ç.) D° Cd°121 Cd')

o oo oo oo r- r- r- o co co .0 .0 r- r- oo oo cn t(-)

cri cri cr ori 4 4 4 c:). cO oc). 06 CD CD CD CD CD CD CD CD CD CD CD 0 CD CD CD CD CD CD CD CD CD CD 4- n1 1-.1 1-.1 1..-1 1-1 1-1 1-.-1 1-4

• c7: czs 4n or-) oo oo U M M C C ch ch \.10 nC) r- r- r- r- r- r- r- 71- 71- kr) vD VD 00 r- r- r- ,t ,t tr-) ln VO ln tr) tr) kr) kr) tr) 7t 7t cq cq cl r- r- r- r, r- r- r- cl• 71- 71- 71" 71- „.

NOMMO—IN-10MNN —i ... 0 CNCNCNCNCNCNCNCNCNO\CNCNCNCNCNCNChaNCNCNCNCh CNCh Ch CNCNCNChChCNCNO1CN CN CN CNCN CNCNChCN ON CN

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \\ - \\ \\ CN Cl 00 00 Cl 7h Ch d- Cl 00 CN CN Lr) 7t 7r le) Cl cq cl cD - C N----,„ CD • CN1 N N -1 C \ -E• è' N.- 1 -• N

CC CC CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CN Ch Cf\ Ch Ch Ch CN CN CN CN Ch CN Ch Ch CN Ch CN CN 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 pz -, • kr) (f) 14-) N N ,L ,L S-- N N NN NNN e) e) e) e) e) e) e) e) e) e) e)

Cl Cr) 7t V-) 00 VD r- Ch CD CA re) le) O r- oo C C CC CD CD CD CD 0 CD ln In ln ln tr) 11”) If) tel V-) tr) tr") tn kr) tr) (r)

•

Cr) fe'')

o g 4 ,---.. — . ,..., ,---, ,---, ,---, _ s• s., ,--. ,--.c,-, 71- ----, ----, ,- rs- r-- r--- •-n") r-- _ t.„, 1--- 71. -- ,c).' ,__, _ \r5 ,,r3' 4 . .. - -- .---. . ,..0 rc, vi,a) r: cc..) c•6 cc, i--- r•-) N •71- c•-, c•I Lr: •- '-- c-:: c-: _ _ . cc _ . r-- r--- ,r „ -.1. ,•,-;" •r, c•4 - cl - c-- ci -, - - O4 ...... , •—• ...... , ,-, ...... • --..• --_, n-• .._.- --- .....- ....- ---.• ,_., ------,

(9

c)

"C 7:) "cf •71 •rsa • •7:1 "Li 7.;.',71. -0 -0 -0 f=1 czt (i) c., o 7:1 ct cz: 7 :1 1 (7 c , 1.-n ;-,0000ctc7SiELF:LcllMmic'zIc " - >-, • (G4a4a4f.:4 1 tr' $,P-n-' Q.) r:C1 1:q f:q f24 124 _ e e L)15 e ‘," e ▪ 0 0 1.) cu a)C..),s a) 0) 1-) I.) HH " c c 12 12 IR 12 "C 1'2' 0 0 at (L) a) a) 1.) a) c/Dzz I il cas r.J (4 •-n •-n 000000 O O (i) MI -0 -0 t> ta. t> !--> 0 0 -11 cd0 00 L) P> P> P> o4 o4 CD

CA L) cdo DO V) 00 z CA 00 GO Co0 Cl) GO °° C..) GO 00 DO CA °° 00 Cd° 0 n C=41-41- r:4Pef=4(:4r=4 n- çi1-4: Pe -•)^ •- ' 1- 1-4. )- P4C4a4

trà AD AD on on on on 00 00— vo AD Cl CA Cl CA CA Cl Ch Ch cl) Cci 00 0000oc; cs: cs: a: a: a: a: 6 6 6 ci ci ci ci ci ci ci ci or; or; CD CD CD CD CD CD CD CD CD --I --I a) --•

4.0

Ch 4: 4n -ID 0 Ch CD oo r- ci cN oo m un N --. 0 op r- oo Ch CD tr) 00 r- r- 71- o 71- 00 oo 00 oo oo oo r- r- C CD CD 00 0000oo tr) 11-) tr) 7 cq kr) 71- 7r 7t 7t 14--) tr) Cq cq cq 7t t cq ci cq -. N c1 s+-4

cl --I ,--( CD on --( cc) cl cq cq cl cl cl —4 CD CD cl cl cl CA 3 Ch Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch Ch Ch Ch 01 Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch .> ,--)...-_ ,--) ---,--) ,-4, ,---1s ,, )--( , ,-4„ s,--4 s, ,--i,-4 ,„ ,, .--1 ,, ,--.1 ,, ,-1 „„ ,---) s, ,--s, ,-1 ,„ ,--4 s, ,-4 s, ,.--( ,, ,--) -., .--.1 ,„ .-.-1 ,„ -7, , ' 017 kr) 11") en 00 ln 000-) Cc) 14-) le) ,--( ,--( ,--( ,--1 , 1 CA CA CA CA CA CA CA ,--4 CD CD ^-I )--1 CA Cl CA Cl -0 0 Cl S.-- ,--( CD cq --( S-.. ,, ,, ,, ra, - - ' CA CA C..,1 -i";1 N CA CA CA cd 00 -C.) .C.: CA CA i'-'1 -C.). -C-hl i* -1 ( A. .C1 C' 'i CA obe) o

4-1 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD 0 Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ci] ,„ s, s, ,„ ,„ ,„ ,„ ,„ s„ Ch Ch Ch 00CN r- r- r- r- oo oo Do oo oo oo os, cs, cs, cs, 00 s, ,„ ,„ ,„ s„ s, s, s,

• " •4 1.),9 Ch CD CA on Lin AD r- oo Ch CD CA on 71- AD - NN CA CA CA CA CA CA CA CA Cc) Cc) cc) cc) Cc) cc) cc) -14g 4-) tr) te) tr) 14-) Ve) 11-) 11-) tr) tr) 1,r) tr)

• 7

71-

Cf)

(--- r"--7 n ,__, oe--,„ •- r"-: N . •--, VD . •-•-n N ,-. Cl • N N N C .-) ,-. 71' N: N v-:' i---..' te-7' vs, t -- N V') 00

cp = Z = = cd O 0 0 0 0 0 0 0 0 0 ,.= 4-J — • -, .. • • .. ..., . —r u c:3 CI 'Cl 'C 1 ' E 4C-C1 Cl 4 F71. c- ,3 U U U C.) C.) QQQQ U c) c) 00 0 0 000000 ct ct (›il 1)(1.) 1.) U OUI) I) ,- ,-= FT1 s.... -a S-I 7,--n I. I, i-i I-1 2 2 (...) z C..) (4u u(4 u upu uup 0 up0 c....)(4 up (..)up up u up u 0 o ,c aa, 4D ,.0 41 ,ID 4n 4D ,ID ,ID >v ›. îli C) (D CD CD CD CD CD CD CD CD 4.1 w.1 PC1

DO 00 cd0 00 DO DO Do) (4 Cd0 00 DO 00 00 00 V) V) V) VD V) V) V) V) 11 (21 C4. n-S - .4 1:4 P4

n cN1 cl 1;) vD v:D v:D vD d- d- CT 0 0 r- r- r- r- C. Ch. Ir; Ir; ir; tri tri tri n.6 nr3 . 14 1n.1 1-1 1-1 1-1 1-.1 q-.n0 r•-1 1-4 /-1 1-.1 1-1 1-1 1-1 1-1 (N (N vn4 1n-1 ..... /-1 /-.1 ,..n1 1•-n1

7t 7t Ch Ch CA CA CA CA CA CA

cl rn rn cl lD rn .71- CD cl cl ---( --I CD cl CD CD CD rn CD rn —4 . Ch Ch Ch CN r- 00 00 Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ,--(,,...-4 .--( ,-( ,--i .--1 .--1 .--q ,--n ,--1 ,-.1 ,--4 ,--.1 ,- ,...-t ,--q -..1 -..1 ,--.1 ,-..-1 ,--4 ..„ ,"-...„.....( ,, .'-•. —• .•.„ ,„ -.., ,.-.- ••.-'... , ,, •, ,, -, ,„ ..., ••'.. ,„ .....„ ,, CA le) ,---1 rn cl cl vn tin rn cl rn rn rn --4 rn ---4 un vn cl CD CD cl cl cl cl CD cl cl CD cl CD CD CD CD CD CD --I --.1 N rn rn cl CN CD 00 CA CA CA CV-' e) CA CA -C.) CV CVCe ) CA Cc) ....C (-7 1 ,...i

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CC Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 0000000000000000000000 00 00 00 00 00 00 00 00 00 00 00 00 00

2,1 25 25 23 25 25 25 23 25 25 25 23 cq cl (-NI cq cq CA CA CA CA NN CA N NNN -f- 1 "(n1, Cq CA CA .C7) CA CA NN

Ch CD ,--I CA • 71- nID N oo Ch CD CD cl rn er) 71- 7t 7t 7t 7t 7r 7t" tn In ln ln IIn un vn vn • ln ln ln ln V") te.) In ln in In In

• U •

C/) /...-n ,C) .,..., 0., r": .71- ,,...... ,--,. ,-.1 ,--. n-. ,....„ Ln c- r- ,---, ._hr": .. ---. . co — , 4 N ,--, t-: .71-.. • ,c5. \ 0 tr).. r--: '',1-' 7r. ,--. 4 kr,- it--,... --. C,..I '''' - v--, ,---, ,---. Cf, R r- 0, ."'. ' ‘.0 ‹) 0 -°-- h: CN. 't„ 'I-, h: h: N:. c-: N: N: N — h 01 ,C, •.- .—. r.„ r--. r.... ,,,-„, ,..._„ ,...... ,....." ...-, .----, ,...... / s----, ,--, ,...---, ,..-,, ,....„, •„_../ ,..--, •„--, ,..../

ai

•-c) 7=3 ,Ti Ij n-c) U a.) (-) a..) a.) (..) '-4 o o o o o o ctctc7ictctc:-' aLt a', r:lq:4 a4 c4 = = f=4 1:4 CI (IDCi) V) C/) (I) V) O 0 0 ›-»- .--1 .--1 ->1 .->l -->-, ›-1 >-1 • •-. -17) -cf 7:1 '71 "Ti "Cf c/) V) Ci) Cil cl, ct cZt upc") cc) (/) v) cf) Z = O 0 0 0 ai ai cZ al caf caf

a; o., . s, a) 0) a) Q) Q) a.) 4D t t t > > 1.) (1) > > > > > > 000 0 0 0 0 0 0 O 0 o -CD gD gD 0 0 .0 .0 .04D 4: -0 -.C) LL( 4-( 4q < < < 1--. ..-1 < < < < < <

cy,

0 00 GO 00 Do GO DO (..) 00 00 GO 00 GO 0 00 00 Ci) C4 00 00 DO C.) C.) D C:1 Ç_) 1- -1 - 1-4 - - f=4 1:4 P4' P4 )- .‘1- " (24 IC4

oo oo oo oo oo oo oo 00 CD rn cn cn 71- 771. 71- Ch 7t 4 N cNNr o0 oc> cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl

43 43 cd cd cd 43 r- cct cd 43 Ch cd 43 4D cd cd 4D 4D cd (T ch Ch Ch CD CD oo vn oo CD CD CD Ch Ch Ch Cl Cl ch ch • v-) vn VD v3 7t VD vD vD VD 7t cr) rn rn Cl c•I cq co 00 7t (3) r- r- r- r- r- r- Cr) ce) r-) Cr) Cr) 142 v.) r- r- N N cA 0 c4.> • Cr) Cr) CD Cr) Cr) Cr) CD Cr) Cr) --f CD Cr) Ch Ch CA Ch Ch ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch :: C=, Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch

Q VD vD --I VD vD 7!" vD --' 71- vD CA :a CD ,--' CD CD 0 CD CD CD -a -a CD CD -a -a CD C) cl., ON Cr) Cr) Cr) N rn cd 00 e) e) e) e) e) e)

4-n 0 -= (71 SD. t4.-n CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD CD O -Ci CP\ 0-n cin Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch 4.J ct 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 cyc --, ,--1 ,--.4 --1 ,--.1 ,--.4 ,--4,--, —4 --1 --1 ,--1 --i •-•-n .--4 --n -„ , -..... , , -, ,_ -, -„, --, .„, 0.:1 ON n CD CD25 --- '-- ---C) -----' --1 ,--.4 ,-- ,--1 --1 Cl Cl Cl CA CA 00 CA CA CA Cl Cl CA ClCl Cl Cl 00 CA Cl Cl Cl Cl Cl ...., -.... CA Cl Cl C-.- Cl CA Cl Cl CA Cl f")' Cl Cl Cl Cl Cl

4-1 1)

ta.c '71- n4D r- 00 Ch vn cl Cr) vn vD r- oo cN • ve, Ln le) V Sr -. \O VD ‘.ID vD vD vD VD VD 15 vn vn un un vn un Cr) un un vn in un un un un vn

▪

Cr)

,--. ,---. ,...--. ,---. ' NN ,---, •-. '71' (--- r-- ..----, tr: r '-'... :' --. V:I; N a ,-. L---:- tri . .--, N ,--, ,__, (-^", r-- tr, N r-- N V5 V5 r-- ' ,--. ^"-`. ,.., . " "•-n ,,--", ."---, - ^ ^ N - ir, "---, • frn 4 4- .--' -; NNNNN - ,--,CG ...... , ,...... " ,..._,cf) ‘....., ...... f.1, '''.., 7_,' 'r.a., Z, rC:, c..1, ..,- ,.._, ,.._, ._, .....-r.-- ...... ,h- hi ....

fai 1:24 ci ci -C1 • = = 1:4 124 7i O 00 CD-, CI, .-, 7.1 g g 's2,:sa,— In.., 2 C.) C.) Q.) ,.2 .-ci -i5 ,I 8 cll (1.) ., 0 0 - - - - PC) = 7i ,/i u (..) c.) (...) (..) c.) 0 0 0 0 0 3 3 , i. 1-4 i-• S--, i--1 0 0 CL) (1)= '' 0 00 "Ci "Ci "C) "Ci "C) .--n .--I 4) 4) (1) N cuuo Qua) 1.)=-(0, ,, PP PC) PP PP PP OP PP 1=1 la) C)

DO U0 (..) U0 c,0 00 00 GO 00 GO GO DO C.) (..) C.) Q) 00 Do 00 Do 00 121 :D a) :D :D a:1 :D a)

7t 71- V> lD V> le) CN CN 00 00 00 00 00 CN CN 06 CD CD CD 6 CD CD CD Cl Cl cr) cri cc) cri cr; cri 4 4 4 V3 C,1 Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr) Cr)

CDCD Cl Cl ,.(D m „ JD ao C cn cz; 4n 00 00 CD CN CN r- r- CN CN CD CD CN CD CD Cl ci Cl CD cq Cl v0 1/0 (-10 71- 7t •7t. 71- CD CD CD v> V> CD CD CD v> cr) cn VD cq cq cq cq ty0 v> iv> Cl k( ) kr) 4-) Ce) m r- cq (Ni cq

Cl Cl ,--, Cl CD -. Ci Cl CD CD —0 CD Ci CD cr) CD cc) CD CD CN CN CN CN CS\ CN CN CN CS\ CS\ CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Ch CN Ch CN CN CN CN CN CN CN CN CN CN CN CN CN CS\ CN cri r- cn r- r- cn r- r- r- oo cn oo 7r 00 oo oo • c‘i Cl CD Ci Ci CD CD CD Ci CD CD CD CD CD CD Cl Cl Ci Ci Cl Ci Cl CA Cl Ci Ci Cl Ci Cl Cl Cci Cl Cr) Cl Cl Cl CA Cl

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CS\ CN CN CN CN CN CN CN CN CS\ CN CN CN CN CN CSN CN CN 00 00 00 00 00 00 00 00 00 00 00 00 00 00 000000000000 00 00 1n1 1n.1 rnI 1-1 N ..(7;r7i (..*; 4 4 4 4 4 4 cq cy (NI cq (NI (NI cq N N cq N N N cy c-A cq cq cq cq e) cv cq cq ci cq c,1 cl cq e) cq e) cv

CD Cl c II lD r- oo Ch CD N cr) v) r- oo r- r-r- r- r- r- r- oo oo oo oo oo oo oo oo oc un vn un un vn vn If un vn vn vn le) le) 14") le) un un un

,----, ,---. ,----, ,----. N. „----, ‘.rD cv ,.o N ,1- ,- - - - -1- ,---, r-:. v-, v-, v::) vs, ,c, N ,---, -- • • - - ,----. ,--. v:, -.- r-- N ,,--, N d. N r-- . N ,n .. N: __N n.o - crn - r-- -, - ^ -. ,0 •-n " V'', --. ,0 .... - r.--- - - . t-- .. .•

7= "C) Ctfa, a4 a4 ct (7; -Io -o -o -o 4:i •4: 00 0 0 cl cci czi ai 4-, -1-J ..1 ›l"-. .'-'

f:4 , 0 0 a4 A i I ? ?- cg g 'F'0d '0c•4- uuci->toc.)c-),', -,:ctmctuc_)00— 0 0 '7" 7,), c`1;,) 'cip , u. ,..c) _0 Q., 1-) 0 0 — ,r. (r? in tn. › › to to" - o o cd cd F4 ,9 (/) 0,, 4 4 4 4 2 2 '' -`i ,..c '''' . z — < ¢ 11 11 la., 0.., 0 0

cdo uo c.) cdo up UP Do 00 U0 DO uo GO CI) 00 00 00 00 Cdr) Cl° D'D D') Q

o-4‘ )-(Ceaef:eriaefl-5' )-4'. 1-L-4' 1-Pel=4P4t):4" a4

VD kr) r-- kr) kr) c::; (nn •—n cr; cri tri kri tri tri tri tri cs: (' cf) cc) ce) cri r) Cr) Cr) 7r 71- d- d- d- d- 7h 7t. kr) kr) kt-) kr) kr) •-n.1 4-1 ..... 1-1 1-1 1-1 1-1 •-n1 1n•0

M N gn M gn M OC 4:0 M g3 gn M M m r- gn czt M gD 71- 71- 1I-) V-) Cl Cl 7/. (4") re) 0-) Ce) 0.) 0-) CD ce) Ch Ch CA CA n n 10 r- Lp D GD GD in CD C) cn cr) cn cl) oo 00 In cf) 71- r- r- r- r- r- r- rq tr) te) m m Cr) Cr) ts") kr) -. CA tr) n0 N Cl cq cq 04 0 ,1 cq N N cA

er) 1-1 1-4 Cr) N N N CS\ CS\ CS \ CS \ CTn CS \ CT Ch CS \ \ GT CS Ch GS, CS \ \ Cr\ CSn CS \ (SN CT \ \ \ CS \ Ch \ CS \ \ CS \ CS \ CS \ \ CS \ Cen ...... •-n • r- oo oo kr) tr) c, oo 00 71- 00 00 CJ\ 00 CD Cl N C C- N -C N Cr). Cr) cc) N N

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CN CN CN CN CN CN CN CN CN Ch CN CN CN CN CN CN CN CN 00 00 00 00 00 00 OC 00 00 00 00 00 00 00 00 00 00 00 00 ,„ ,„ ,„ ,„ 4 4 4 4 4 4 4 4 4 Ln Cl NNNNNNNNN CA CA CA CA CA CA CA CA CA CA CA CA CA Cl Cl CA CA Cl Cl N CA Cl CA CA Cl CA CA CA

CSN Cl Ce) .ct 111-) r- 00 CN CA cIP 7t r, n,C) 00 CN CN CN CN CN CN CN CT CN CN CD CD CD CD CD CD CD CD V) V) V) V) v) V) V) V) V) V) In \CD VD nC; VD VD VD

• • • -

e•-n N

tr) r:

'1-‘) ,4 O 0 o > > q Q) (..) U a) z d 77)7)7.17077177)-7) ,--1 --6 n O 00 m 7771(:) = 7: 7: .y) (ID 5-,' . S-3-,' . 21. ' sli .. . a - ,' . El ,' ' i 5 -' . ›,>-,a-,== 0 › › Od cc ,nd cc M M • = = 0 0n M U M M '.C3 Z (..) U ‘2411:4(=4a14 M M M 47j •+F.- \--...e)_,,,,) . ,zt oUU cn cr) cn up co ci) c/D UUU M a) a.) a) -.-.' '''L' c7.: (7: czt c.,:s czt cc.1 c,:s a =sa ta ;P-33.----1==-=-=c,:mc4c1)c./),,tm .T..,ü....1...2.,4-4;.1-,LI., M=4 0 0 al ai ai ai ai al L) U U -7)7)7,2,-.., - 7 > > > > > > > Q.) 754)52 ,--,° U 0 0 U U M M M M M M M E-, E--, E-1 = = CD .; a4 4 4 › › (4 ,- ,-. ,- n-• ,- ,--

00 00 Do 00 Do Do 00 GO DO GO GO DO GO GO DO GO GO C-) C) DO DO 7

7t 71- If) CD CD CD N N an an an rn rn rn rn rn

4 -4 o6 oop; o6 ct; cr; cr; cs; ct; ct; n1D r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- vnI 1-1 1n1

"MMWCNCT\COQUUMMXD,0 00000071COMMMMCD00 0 00CD OCC00000 1 .0C--MMMM --, 1,-)EntfltnI4' NNNNIntntnin I .....

Cl CD r0 N CD (10 cr) rn rn CD CD CN CN CN CN CN CN CN CSn CN CT\ CN CN CN CN CN CN CN CN CN CN CN CN 4-f CN CN Cs\ CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Orzt Q) Q) n..0 CD 4.-) eiS CT'S' "- cr, CN CD C) CD CD CD 'Efa'c cq cD c.1 cl cl cq 1:21 1-" N NC") N r0 N N Cq Cl cc) Cl (c) Cl Cl Cc) Cl Cl Cl Cl

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN 00 00 0000000000 00 00 00 00 00 00 00 00 00 00 n n33 n43 \-43 ,C) C--C-- C-- r-- Cl Cl N N cq N N N cq cl cq cq N cq cq -S- SSS S. SSS SS N N NN cl N N Ncq N N N cq cq

00 CN C) Cq rn r- 00 CN CN C CI cq n0 nID n.0 \..(D nID \0 n..0 \-0 '.0 \.0 • • 4

Cc)

o = 4-1c

o O4 "cs

Ca. Cl, ctc,71 ca1 c7; cd a4 a4 ... -,,,,,,,.0 . 1241-,E'ic .,, o o = m o oo u u a4a4 ›:=1==c2AA88888Û 4.-• .-• _,o-c4 omm '' 0 0 0 1,) 4_,L) ; a) a) a) a) = nn tl..) rit) UCJILLI-).1)4) ;n0 ,.., $-1 b1:1 t.0 .t-' _go 0 ,,, vp (fi c4v") " ..1 ! tcf.81 to-lo, ,I,),:) (I,2L:) ifl ,i-1 ( ) cn 4-)

0 00 00 Do Do GO GO 00 00 GO GO DO 00 GO 00 VD 00 D° 00 °0 00 VD 00

C'S oo r- r- r- cq CD CD CD V") v") CD CD CD V-) V-) V") Cl CA 0.-) on un vn = O; c; 6 6 on on o-; 4 0000 N N 4 V-) 1,-; 00 00 00 00 00 00 00 CN CN CN Ch CN CN Ch CD CD CD CD CD CD -4 -1 •-n -4 -4 -4 cl cq c..1 cl cl cq

,t m m •,r on m c) c) r- vn un vo m r- 00 m M .7r 7r c7\ cN NNCC10000 r- r- r- r- vD kr) tr) m tr) tr) c) r- r- CD CD CN CN cc) cc) C__c‘i r- r- r- vn r- r- vn vn on ,-, Cl -c:t cA o

Cl 0") ,--1 0"1 Cl CD ,--( CD ,--( on --, C) -o -. -o -o -o -o --, -o -. --, ch a> CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN Ce\ --n ,...-4 ,- -4 --. -4 ,-4 ,-. ,-4 -, --, -4 ,-, -4 -4 -4 ,-.4 -1 ,..4 -4 ,...-1 ,--1 pz "27-, ------..-, ------..n cA 1-1 N t-- . . 1.-1 1-n.1 cl ,-1 cq -- e) N e) e) e) e) e) e) e) o -..._CA --4 Cl ...... --( Cl ,---1 Cl ,-( Cl -,-._--I Cl c-(-.„ Cl Cl Cl Cl Cl Cl Cl CA cl cl CA cc) CA ce) Cl N Cl e) e) ccl Cl Cl Cl Cl Cl Cl Cl e) Cl Cl m ,-. o 0

C1L 4-1 60 0 000 6 0 60 000CD00000 o "7) CTCNCN CNCNCNCNCN aTCn CNCNCNCTCNCNCNCNCN 00 00 00 00 00 00 00 00 0000 oo 00 00 00 00 00 00 00 00 4.-+ m -4 -4...... -4 -4 5555555555 55 55 -- 5 36 00 00 00 00 00 00 00 00 Pz Pz Pz Pz Pz Pz Pz c"Z ;6 00 ClNCI CINNNN NN 010ANNC101Q0CD NNC1 NNNNN NN NNNNNNNMm

a9 ;49? ra.4 Cl Cc) 7r tr) r- 00 CN cq cc) t.r) VD VD 0000 CNCN CD Cl Cq Cl Cl Cl Cl Cl Cl (C cc) cc)cc) cc) cc) cc) cc) cc) cc) cc) cc)cc) cc) 71- ‹) 0 VD n1D \.4D VD n1D VD nID nC) VD VD VD VD

▪•

..._ „---. ,--. --. a a C n ,---, N , r.---, ,-.... ,...•- tr, ,,.."-n 0 ,•n••• a..6 -,-1- N ,-, et -- _ N N ci 4./^4 r-. cn-:: r-J --:. (4 ,--, ,ri,i- L ri 4 - - __ cc.cc, - •---.' ri 7 c4 7 ‘.0 7 7 r: 7 r: 4. ..n- ,.,--,— -,- _ r- - - - 7. r 7 - - - v." ' tc, - v-, 4, r-s- -, .- - ,-,-, 1- - 4-, r-: - r--- r------, (---: 7 c--: r: (-----• :r --; R R -, - ,- - -, cr, - N -, - -. - N v5 r: r: ._.., ,-...... ‘ s. --, ...., ,-, ...... - ...." \ -..., ...., ,-..- ,,...., ,_-, ...... , .....- ._.,o •-..-

70 70 ma mCI, P4 P4

0 0 ..ci ,..ci = = 0 a) a) a) cd\ d, ›. - .c.:2, .2 • ••-, • ,-.., = = = ,L.) , a a ,,c0 0 0 — •,-, c ,:=> r ) t'c'Eic.,,,,„1= 0 'et' .ct€1 1:4,, a4 000COQ 0 g q " N(), (..)c.)(8 0 0 C m cd _17, $-n 1-n 1-1 4-4 4-4 (..) L.) U CD C. (.. (... ° ° 4-1 g 4-,) .2, cu cu ,in) g g a) 1..) = = c) 0 a) a) 5 5 ct ,ct >>>00 , UUU '5 oo 00172 0—,,A,,-,, 001 ,7=1 70 "Z: „170 40 ,g0 g0, cD cD g:, -,74 "_, ,_, g0 40 r--- CEti -c-Ei .8 -10(4 4 4 < -t: -‹ < g) ala d 4., cq cq C) C) c , 1 Cl Et Et, Et C) C)

DO GO cd0 GO 00 cd0 00 00 00 00 °0 00 CID up Cl.) cl) 11 C.) C/D CID Ur) CI)

C4 P4 f:4 0--T 1-j‘ 124

r- r- r- rq cq ci rn rn 71-. lD. cl. cl. cn. rn. . un. cl. cl. cl. 7t. vD. • n,c; 06 o6 06 o6 C c; un vn nC) r- r- CN CN CN Cl Cl C) C) CD CD CD CD CD C) CD Cl Cl Cl Cl Cl Cl Cl Cl Cl Cq Cq Cq Cl Cl Cl Cl Cq Cl Cl Cl Cl Cl Cl Cl

• n1D g0 ( cc M 4) r- oo m 40 m 40 m 40 M r- CD CD CD CD r- r- r- r- CN CN CN CN CD C) CD C r- ln CD CD r- r- CD (D r- r- 00 00 Cl Cl (-V C cl C C r- r- ,n ,-, CD CD cq r- r- r- r-,„ kr) tr) kin cq v")

• CD CD Cl Cq CD CD CD CD --I CD CN CN CN CN CN r- CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CC ,„ cc) cl cfn cn 71" Cf) Cf) 71' 7r 71- .1" "71- d VD VD V) VD v:D \C) 7t cl cl CD CD cl cl cl cl cl cl cl Cl cl cl cl SS SSSS cl P.-1 cv ,t Pz P) N cq cl Cq- e) cA cq N cq cq e)

CC CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CC CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 „ . ... CD 0 0 CD CD CD CD CD CD CD CD CD CD CD CD CD CD • cn rn rn cn cn rn rn rn rn rn rn rn rn rn cn rn

n op ch rl rn 7t VD r- 00 CN CD Cl cr) 71- /D r, d- d d d 7h d- d d ln tn kn kn VD VD VD V) VD VD VD VD VD VD VD n.10 VD VD VD

• ▪

00 (-1-)

vD v-, ,--. ,-44. t- 4- - t-4: -4- t--- "4-, ,--, t.--, .47t. 4 "4-4. ,4-4. ir: - ,---4, :::4---' t---: - -- - oo .7t- - 4 ,..o - d- .444-. . . . r-- - N v-i' c•-•', - ,444-. ,40 , DO le: 1-- ,-. r) •I• ^ 7 -: ,--. ,S5 t"----' o.: —..t---- 4 -4 v-) r---44 - - 441- - - - Ir: -NN--r--- 4- -: 4 ...... , ...-4 ..-4...... " ,..-4 s..-- ,...-4 ,_444 .--.4 ...... ,..-4 ,.....4 :..-4r s...... ,-.4 ...... -

C7S CZI a4 C4 ...4 (1.) CL)

(! U- 6 = = • Z Z 7: 0 •...... ,,,. --, a) a) ,,a) a) 773 .2—" 1 (!'--, i: il '''''' (:)-' --. 0 0 .-,c,71 4c7d G m cl q cl ,Fde 4c"t' - 'c'cc' ' &I rt c'7,::' 'ezi a) a) a) (1.) O 0 c.) c) 0 0 0 O 000000 mQ:5; ' 1Cm0 i E 50 (-) 8 c.) c....1 QCL)-.) CL)" CL) c I.)-) 8 Q) cl) Cl.) a)n a) a) a.) z ipl 121 1.)n 1--'.) 7 CL)—' U7--' C.) • ," "" 4 4 S-. ;--, ;-. ;-. 14 ;'' = O O = O CD O C) C.) C.) C.) C.) C..) C.) C.) ;5 t-' t-' t C) 1.) CL) c.4 cip c./) e) ci) c./) 0 0 C"-) , , ›• › (-) '' (I) 00 00 Cl CCI (CS CCS ''' ..C: g : g: c4D a) 42) Z '77'.) T.) -0 -0 ,...,1 c..1 ,_, ,_, "0 c) c) CD c) c)

000000 0000000000 00 00 00 (..) L) 00000000 00 00 DO 00 DO 00 GO C.) C..) 00 on r21

1-J 0-5 )— (24'

• n4D tr) tr) in ln tr) --I ch c) c) c) c) oo n 4 4 4 4 Ir; V-) v-) A$ AC). A:> A•5 o6 oc') cr cs; cs; CA CA CA Cl Cl CA Cl CA Cl CA CA Cl CA CA Cl Cl Cl CA CA CA cl CA Cl CA CA CA CV CA Cl CA Cl CA Cl CA CA Cl CA Cl Cl CV Cl CA Cl Cl

et 4: 4: „ip cl cd ch IC) v-) M 4: c10 CA AD 4 4 cri -in r 00 cq cq cl 7t vn vn cq cl c) rn rn c) CN CA CA CA CA CA CA r- AD AD AD AD vn vn AD AD V-) In V0 AD ,.. V-) tr) vs) V-) le) cA cA CA AD AD CA CA CA Cl ve) V-) Cl CA AD Cl cA CA cq CA

CD CD cn cl rn on on CD c,1 cq CD cq cq cl (1 cq • CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CSN CN CN CN CN CN CN CN CN CN CN CN CN 0•n CN -„ „„ 4 ,, „„ ,„ • CA Cl CA CA Cl --I CA CA CA CA CA Cl CA CV Cl (n-1 CA Cl CA CA cn (NI rn cn ren cv cv cq cv c,) cl cl cv rn

CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ,„ ,„ „. • ( .1 ( 771 -(2,1 (-"") CA CA „„ 0 0 0 „, ,„ _„ „„ rn cen rn en en ren en rn rn rn rn rn rn

C cv rn 71- vn AD r- oo c) cq rn 71- vn AD r- AD AD AD AD AD AD AD AD AD AD AD r- r- r- r- r— r— r- r— n4D AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD

V) e"-n 00 I": ,•—n r•—n .—I t--- ,---, ,—, in n s ,—. ----. •- , s -- . _ in t--- fn V') 00 00 00 00 ,i::;• rs: .1,7. r...: t....., _ ., ,--, „„ oo

— 06 ir-:' •-. tr Le, -. r.: r--: l--: ,--. -. --, C/) ,...-, ...... " s..-, ,..-, ,....., ,..--, ,....., n....., ,...... , ,....." n..-, ,-.., ,...." ,--, ,...... , ,.._., ...-, ....." ...... , ,..._."

oc 7) >1 T7J 7i "Ci 7C g cg cq' CL, CLI CL CL, U U M cci cI:1 ct (:4 = = a) = a4 124 = 0 0 "C) "CS CD --,o li. , ,=,• -, ar) . „, .2 0 0 0 0 . 2 0 0 0 • —, • 2, 8

ci p -, ,9 -, d F.1' ,,,c)-, ,i F,,d 5), c; ' , ';', ci .1 'Fi' 'F'zJ ''`E. t4 . '-' 8 f , 4 8 8 4 0 0 z = 0 0 0 0 8 oo,,,,i3O000P-,=.40 ,L1.2 _512 u a) a) —, Q) ct) 7,z(D•aNIM ci) a)UUL“-) a) a.) ,a) a)CnC/D a) 1-n •,_Z ' = s--, 1--, s--, ;-4 - 1..) ,,, (,) s. s. 1-• u-, , , ci.) :5 e.( = u Z "-) Q , , 0 c.) C.) tif') " • (ip .`/) P) c.) c.) c.) > > c.) (4 —. . (4 up (•. 1 0 (/) :(2 ,A ,./.3-0 v.)-0 0 0 ,), -a on on -a -a on -a .0 u'A 72 72, 1-z3 °à -.0 -0 -.0 -0 CD cl (-NI CD CD N < C og1:1 = CD CD CD C) C) CD < < C)

cA Do Do DO DO CID 00 00 V) U0 U0 CID C4 CID Cd° a a D a a QDQDQ DC4 CA CI) alV) Upa

oo oo oo tr) v.-) cf') 7h 7t cc) on CN CN CN CN CA CA Cin 6 6 •—n ,4 cc; 4 4 le) le: lei le; r2; oc> oci cl on on on on on on on on on on on on on on on on on on 7t N NNNN NNN

c7; N czt „C• „JC .71- op c> ct cA In Ln n4D nc) Ln CD c) Ir) Ir) nrD n):D cD c ln In In kr) c•7) CD CD CD CD CD qD tr.) in on on N cn on r- r- r- rn rn rn on CD CD ce-) cl c1 cq cl cq cl cq on cn

CA cc) CA on on on on on on on on on on on on on on on on on ch ch ch Ce., CN CN CN CT\ CN CN CN CN CN CS\ CSN CN CP\ CN CN CN CN CN CN CN CN CN CN C7\ CN CN CN CN CN CN CN CN CN CN CN CN

tr) Ir) Ir) 1.r) Ir) Ir) ln Ir) le) le) te) te) le) te) te) Ir) \C) C) ,-. . . , n, on on on on•-•nn on on on on on on on on on on on on on on on on

CD CD CD CD CD CD C) CD CD CD CD CD CD CD CD CD CD CD CD CD cN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CS CN cS 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 -. —0 CA cc) ce> cc) cc) cc) cc) cc) cc) Cc) cc) cc) Cc) cc) cc) cc) cn on on on on on on on on cn on on on on on on on on on on

oo C7\ CD N (4') 7h VD r- o0 Ch CD cl cn 71- te) ‹) N N 00 00 00 00 00 00 00 00 CO 00 CN Cf'\ CN CN CN CN n1D n1D VD V) VD VD VD VD \-0 nD VD VD VD VD VD nD

• $ • • •

Cr) 00 Cr)

6 txc) ,a 1.Q . (.-,o ...' -a o Q)

Q) ai r) Q • ;-n 4-) O P-, cA N (1) ( c s. .--, Q)0.--,0 Q) cd 5., • -0 0,0 a s-n • -.▪ -.., w _a 0 -ao up Q) u., cu Q) Q) a) c,-, a) .--, = o ,. __,› 0 75 -c) E 7:3 74.- a) cd cd a) ., ,.. a) -c; 4,-=, .--,c.) o .,t' d = a) - -, c.) › 013 O al., = 4;4 a a.) - ccs I. (...) a —c/D _a a.) cd u, cA cn cd 6 c..) 74 4 '.5 4_,C -4 ;-, 0 4-4 cd 0 ,-,-cd 0 4C74 2 ,..- .,..> .4 a)a);... ,4 $., -7; •--, 0,0 c.) 2 0 1) 45 -6'-, › cd o cd 1-,a) ... 0 Y cA c/D .4 y -0 '17 ..= a) cd e" ---, 0 . .. "d > › ..-1 a) = .'1) 1.) SZ1-4 ..'-' = o 0.0 0 0,0 . o- 0 -. (...) Q) :E....., •5 = .2 R ,.., .– Y c..> c.) › c4 $.., 4-' -4-, • -cj > = c.4.-4 y = cA a) -6 1.) a.) a) ,- "c) (c,'' 6-, ,e E CA (1) 0,0 0 '-- a) Y -8 -.s 0 a) c -4 -4 up ."., „, o up u) a.) up a e c:f .-0 0 ° 'cl tiD _ a,—, — 000Q) ,.. czt a) '1:5 (i) '1Z; Cd '-' t--2, 0 CI, ''-' ' r., "= a) 4, . ,-, n . c.),_, .0 41 t=1- a a N = = = 4 cL) cd tu) .- cd cd 5 s-, 0 :1 ct .,-, C) '"e/ uc- )1..) 1-4 = jz b!) tV) E (/) 2 •4-+crd -6 a, .r., t) (.4 Q) 0 k . d .,``' °-)" :-,- 0 a .> = 0 , -,3 , _,-1 u ..,-;., ,,, to rz2 a.) -cs .,- m = -- a PC) q.,) '''- cA -Z cd "Ci 1.) › Y I. c.) up 0 >- "t71 2, "ci cAc`i c.) Y f:), .2 d 10 (L) Ci.) (1-) 0 n - • Cd S ,,, Ct ,. b-C1 0 $.) l'e' ' ' ,-r::, ;-, -i-+ 1-, a) cd •"d . (1-) .C,2! Z g 0 -.• • --, .> .,-. "c) cd '- 4_.4 • = ,--, a .,,,;_-,i 7:') 0-0 a' = • = a) -0 Cl) '' 0 0 s-, • -. 0) ,.= .c7' 0 ° ...7.) fz., -8 cd Y y 4-6' U 12, ° (1.) .28 up 0 a CL., • .;:,, I, CA = ,..M 0 u V) Cip Ci) Cd Cd "' C/D C..) ,, \ Cd I..) C/) = • A-A -.4 "Ci = = •('-' CI... '4,_, el., ... ..1 o . CZ1 • '-' Cd."' . 9.-) C/) C;I 8 8 . R, O n t 0 f2L, -d . c-E1 ,2, (1) up • •- 0 O up ....,Cl.) . 41 Ct c 8 „a, 0 . . 0 ,..„. ct C.) 7—: › • ,-, 0 c.) C7Ei ce, c.) 0 a) cd c.) ,Sn t•4c 14 ci) 0 ., 0 'a'- 8 •- > —,..) _rz -a — a u ai) 2 ar ,----, 0 --, ....- > a) _ .4 .`,-1-,' Z 0 ,--, a) jz . ,..,2,c 0 e, a) c_) . • 1-) •› › ;-., , a..) ,._ > Q) o •:o 4::) ,..c I-) 0 (1)j.- 0 .t.4 ,.0>1.) 4, I..) • ''''' o --S 5 * c.75(/D 7:9 -iou 7:2 '..- -= d.) _ — 41,-) -c' 12 > o °-) •,.. ,..a c, .,., cd E co --5 (1)›"2, b. ,.) m cl , É,,J) ..• -,— • - , ; (i., _.- -6 -c) N 1-' a) 7), cs. —0 q 0 ,e -1010 g tb, ,crs `i) ,C .0 C.F-1 0 C/) a di ,z; 00 I; .. . cd es_,, ,c) a) 0 cA cA cA cd • --4 '..' • ''' .-E 7:%'., 72, g' ,-- -d vp"' ;72, s. (1) ,e ça, ' • Cd • 1-_, ci.)''' "t 0 P-' CIA (") 1 Q Cd ,.C, .0 ..0 00 :;-0 (1.) Cd ...i 0 CZI Q) 1..) (1.) (I.) 3-1 75' 6 -' 'A z ô P-, C4 1A4 c/ PIU1C 21Ç.:11) 12IWCD CL) .-. CN CA.) .71- ‘,-) VD r-- ,-, ,d- vs, ‘0 c--- co 0. .— .4

•

n 4 71- ri) ....., •,' .,.-- :-..„_,n 0 __. (.5 "Cl 1-.4 cl.) Cr. ,,, -o ct c;D" zz,' .2 -5 o- ,-, o 0 - o - ,d-' c,1 75 -0 ,-, E bo•- c+. • — - 5 - • 1.) 2 › . ac4 0 t1 .5. 0 ctE •-. ,„ ,..• -,-':, ,.. :. 1-, C \ . •-1 '••• '4-1 __, C) (I) •-r-i v)C) ;—,cip 4-...., •1-4 ,—, 0 0 CL, ,..4, (,) ...- ,- cl) c•i) - IDI 0 C) .a izq 0 ,- 0 (A' a..) ,-, 0 0 -5 ••• }E.,' 0 3 8 ;-. F- .'" (4 (.) -ci r-u n •› ,t: I.4' CL) 4_, (i) 0 CL) , ,' , L) ;._, a) ct •- ct c, • 5. " 0 8 • -c•,.t, ,.., . -ci a-) 1-) a .t>u • — - — u r,1 oa..) 0 E f=14-8 m ..?. ,,, -6., k; .,- E czt ..a E (''' ,1 *- ._. 13 8 2 ›, L>'-'ce--J-0 le) „ci. o ,. _s: .4 = aJ E .,,.- '' ti at -411 a.) 0 Z ,,, z co 1.) .9. -o -5 4. g "C) 0 4,7i a 4-, 4c771 -7_,/ 2 CL, 7",_1.) 0 --t',. h ccl 0 a, el c.) .a ••--, 0 --. E "Cl r- , C) cA --,„• ô E 0 '-â 9.q.. P., F cd 4E c7, ar cd o - - cL.) o ,_, -- s. e4-+- 70 -cd cr) 4-4 cl 1- -' = ,4 $-' (4--, E ,,, o , .. 1..) o .-0 o ,„ - q..) 0 ,,,, , bi) E 0 ....." _1,..., g c9 uto 2 .1; .7)0 _ O cl.c....) g .Ect to-° 2 ?_:':'-,'.c. ,. m, „.., ta) (4 › '.7,-,' 'ILI Ct ,..''' a.) •'n:-.. -o 15 c1/1 S--n ,.4 0 -„, ci.) '-2_,- ' 4-, (4. = ,.. .0 = .,_, • - 0 — • - ,L., 41 (/) V) 0 0 4.-n M czt •._ 8 (Li cA (1) O Z 'a. 0 P4 F, v. z q.) , 2 _ <6' - 0 G.) cct0.\ 0-) - V) P, ,- •',--, = .,., .2 E ô ,-. -., E c."'D -c' Li.24 Ô p4 ra., = - '6., czs '-- c. 0 m ,,,-• ,,..) •- c,s ct - -.1 m F, E v,.., 0 ,. (-) ,_. o c:,.. -0 .4 tb a - , 0 , 4 b q.) ›, 4- C.) ....4 ccl up^ t C, ,..C• cf) C.j .2 .4. ' "Cl -a) s_," t1 •gf-_, -, _ ,—, e4 •,, "" E 04` ,„„2_ tr) a) O c) (i ?4 cA ,-2 )-I ° ''17'.0 B 0 c.) bz '`'-'›,c...) 7)' 77, " ,,,_1 `Auctj,--,cd•-'') -0 o E '' •.°-tdc41,S.,),a • a) c.,') Q a cnD › " 1.) ,-, ,c..,; w ,.rzm e,06.),=It Ec,,s ›,, ›., 0 2.-,) 0 .r. - .7,-,, F.., ct ,, 0 ,_ ,-c -6 acd .-. (:)c-i--, . 0(1)0 a.) ,- cu "ci •'-' s..• ct .- 0 Z 4—' C'7>t I) "e 0 "Z 4-, cu (1.) ,-, „,^ -= -e cu 0 0 tf) cz: • -• vs = > ,e ,... as, = tot+, E e.0E.= 6 E.t,2, o 0 te z-.,„ o 0 - ,.., .1,,, ,- .- ,•7: = . „, -0 0 (3) (4 "F--, .5 E <,12, 1,( ',-,-,;),-E ,,,., -8 ! ,L) . .— c_ -c) ;-. - ›, c'' ô -.B ' c..1 772 I-' ---, I.) c . 1-..2, P-, ''' ,..)' .4 a,' (I) 8 tin - 5 c‘_: , O ,„ _ _ m ,..- - o ..= 2 cd 0 t. .;_,- .”--= 3 = . - ' k. 5 . E 4 a ) 3 0 „1,7; » 4<7; Jz, u a) -= — ,4 3 ,E, 0 ,.ct I-)• o . .,/ E ,J0 o a., 0 -ii „ '61 .-- *= E 0 = ,--,-0 8 c.) .2-2. O t g i Ir; = 2,, g Lz_.1,..,,-.7d ..,5_, ..O 0 , ,4_, L., ,i,..3. - (1-) -F, „u .0 a , :,., -ci .,..) E 1-) c,_. e>-5 „..0 ct = ,.= tip— 0 c:3 ,112, '77"1 0 ,- = m 0 2 17' "ic c„ .y .- ,,t) ,..1--_-_,, . _O . 5 .(:) t.),_ 0-ci ,i.) 0 ,-, 0 - - • tu), i., .,,,, . ,c, a) , (i 1..) ›, "ci 4.4_C5' cr> ,_,t1:) 4: "..\ 0 a 4.4 zct j . b% .771,_, .s--i-> 69 g a%.) .Zci) eq.-f cuLd CD---1 4-1)', c/)71-' '1 cucip E 0 C'A 3 O '-, '-"1 (I.) • - - (1) ct (1.) _a ,z) -c) ....., A--, ci-) ,4 (7 .2 1.) . - s...- 0 .4 3 a -,-. = .-,- ra., 0 f.L.) 4., v, c.)" a) o ,4 0 0 f:)-'I )-1 ,--• -n -n C..) "El cd 0 49, = cd az, 7,=:), >. A a cl-) 1.) cit ,,D 1-) C-,• -( 6 4.i szi-, t '4 . - 5' •-z up › at -4 cv2 " • 5- . 1 c '7; ai) (1) 1.) (L) ai) f:' c-1- 4.•J ^ 0 t ' li " (L) (1) '21 '-CI . 2 ,_ a 0- , ,--;s i: 8 .2 - ',, -= ° "' 6 ' c9-1 ct .'-' 1 0 y.' -- u c.,. (-) 2 ci, b.0 E , ,_., › ;...1.- 4 -c 1.) ,-,'D ,.-., 0 t'A .4--, . czt ••-• eu:/ ... C) 0:1 4.1 1Z)..4.0 ..A- '•n__• cd (-) Tt, s--• t,) 0 ;--, :,-,, ,', _. „il-) eL-4 - E E go 7c:,:i -0 4" -, (.)- 2 0 'e ,.8 sa% „,

:1 ''-' 0 8 :8 -' Tj "g o 0

• -

kr) 00 Cc)

•".. /.1 ••--, ,.0 C--- f c-1 ...... ,.., /7, ,,-^. .--•), n •••---. CS) ••- ,_, s /-••• .-. V-1 - C"-- (---- tr) ,----. t--- - ••••••n - d' ,--. )r) .-.. •-•. •-•-• )./-) •-: (-1 d' d" •-. cl. — ,—. - t----7 --.. in v:, N t-- d- tn r--- .. N 4' -: r- tn ,n in - _ s„...... , , ,-- ....., ...... _, ...... ,

0 0

CL 0. 0, CL, Ia, 0., ;-. '•-n ceicelcelct MM 0 0 O cl.)' a4 1:4 C4 124 (;, 0 71: - O 0 0 a.) a.) 1..) cl.) .,..., ., cet-0 O cll a.) a.) cl) a) 0., 0, C) tc,',3 .'4 rii 0 C:.) (:3 -.)" d c!:) cl:) O 0 ,.) 0 , .., , a4 • --, ›., 1.-im ct celcetcelcetcelcett) 0 • ri 0 z 04. 1c4 1241)00 00 0 0 ci.) u cet cet cet • —,,.,s 0 1, ,... — V) V) V) C/) C/) V) s., s. C...) C..) ;_, C.) O u a) cl.) cL) ci.) a.) a) C....) L) z z 2 L tx0 i:4) , › › › › › › ç:,., ca, „ „ Q) › 0 ti- -cs = o o o o o o cc: czi 4.) -in -in -in -° -c -c) o o Q.7 Q) (7 4:', P1 1:4 E--1 < < < < <: d 00 u0

00 (.) up cd0 u0 cd0 C.) 00 GO GO C00 GO on oo oo cdo L) GO GO Co0 00 (.) :D :D :D :D :D 1:-1 a) a) a.). 1:21 :D :D :D )- n- Ce Pe )- I- .4 Ce ae

• Cl '4.) 77, 71- cc on ch vn vn le) Vs) le) Ln Cl Cl CV CA CA CA Cl 6 Cfn 4 Cr; Ir; oc; 6 6 6 6 6 CD ---, -4 -4

CA 00 CD C7\ CDC 7t VD 7.,71 Q° CV Cl CA CA Cl Cl ,t ,r „f ch ch ch k0 nC) kID7t ‹) CD CD "7:h r- ,r-D ,rD- ,r- rc)- , r- cD ,1 cq c,1 0 0 Cm clc,1 CA CA Cl N Cl Cl Cl Cl

00 71- ,, CA CA CA Cq 7h CD 71- CA Cq Cl cn oc 00 m Oh ch ch ch ch Ch CA CA CA r- r- co oo oo CA 00 00 00 00 00 CA CA CA ON CA CA CA CA ON CA CA CA ON CA CA CA CA CA CA CA CA CA CA CA --I 81 23 23 4 25 N OC vn r- N vn r- on CA N Cr) V") CD CV Cc) Cl on o c,1 (NI CD CD CD cc) 23 C..1 C---1 "in-1 n(.0 00 o 23 36 23 23 23 23 23 ;6

m vn on on on on on on on • c,1 c.,1 cl cq c,1 cl cl cl oq CA CV CA CACA CA CA CACA CA CA CA CA CA CA CA CA

CA CD 00 00 Cc) CV CV CV CA Cc) 0000CV CD CD CD Cl Cq CD CD C) C) vv>00 00 kr) kr) oo r- r- co oo oo cc oc

CA 00 4 v r- oo V:, 00 CA CA C) V -) V-) 00 00 on c-,1 on on on on 0 0 CD on 7t N cq cq oq on on on on on on on on on on on

,---. ,---. h ,•—n h. ,---. n N .... ,JD M ,40 -.. 0.. t...... , ..,--1.. 1--- v") r--- - -1(- ,.._ __, y.: ,,---,,,1n 4 ,, ,..,..7 ---. -:, s r- 4 4 .71-- t--: ,--. • • ,--. Lr, — — — 4 — — — — --. ,- r- v-, r- If:, 1---. ,--, ,—.., ,..—, ,...... " ,...../ s. --, "....., n—...' '... —.., `...... "

-0 -0 O LID CID o cd ct >, fa, 114-• ct 7:3 7:3 -zi -ci zzi 72, o o 5 cz9 v9 cAu v9 (..) .4 4 ,4 -,, 'al, 'FliaL'ill' ca: c E E II 4. • " •- LID ,,, ,,, •- (7:1 m Ct ct ct C O O ..." 1..) ct ct czt ,.51 ,_<.2 ,42 c9"1:4 P4 (24 a4 4 4 z --" =u cd' ct"" czt cu cu cu ci) cu ci.) ,- - = a) a a a c.,: .,.., .T._, ,,, ci, c., 0 72 ; ; ,..) 0 0 C,..) 0 (L) <1.) Cli O t" 4.-, tn Ln tr) Ir) 5f) --, Cl) C/) Cl) 4: (. 1 (- 1 (-) 4 d: d: d: d: d: d` 4:: 4) (1) (I)Ctctct < Cl CNI Cl N (-NI N Cl N N N -‹: -‹ LL, P:i › › ›

00 C.,) 00 00 00 00 GO GO U0 DO GO DO GO DO DO 00 00 Cd0 00 00 GO 00 D

r- r- 71. m Ln ,, un un un un \ZD \C Cl Cl CD d' \C) \SD CN .6 .6 '.0 00 C -4 -4 -2 4 4 4 d: d d: d: 0000 6 6 -2, -4 -4 cl N ci cl N cl N N Cl CA Cl cc) cc) cc) cc) cc)

CD CN 71- -71. 'd- d' 71- -" r- r- r- r- ,t ,r (7-) (7) 4-) Cl Ce) '-' C7) C7\ C7' kn N Ca N r- r- un un r- r- qe. \.1:D r- \C) Cl Cl CA CA) r- r- r- r- ci cl „ „.„1

C) d- CA cA N cc) .0cn cc) C) cA cc) N N C C 09 00 CT 00 00 CT CN 00 CT C CT CN CT CP1 CT CT Ch CN CT \ \ \ C \ Ch CN CT \ •-n .... 1—n1 1-1 n •••n "`n •- -•- Z) Ir) N Cc)— 00 00 '.000 ci") 00 00 CO CO Cr) N 00 00 CO

NN oo 25 'c'f) CD 23 ô el5 C: Cl /-1 /—nI 1-1

cc) cc)on on on on on on cc) cn cc) oc) oc) oc) cc) cc) cl N CA CA CA N CA Cl CA Cl CA CA CA CA CN CN CN CN CN CN CN CN CN CN CN CN cr, (E7, 4 4 4 1..?) 'Cr.5 \ID \CD r- r- r- r- oo po CD CD CDCD CD CD CD CD CD CDCD CD CD CD CD oo oo oo 00oo 00 po oo ;(3. ocoo 00 00 oo oo oo

00 CD CN Cl on r- 00 C) ,t v-) -Tt un un Z:J nID r- r- on on on onon on on on on onon on on on

- ▪ ▪•

00 Cc)

c,-) C,1 ,--, --., .,1:- ,----, „-----, ,---, c, ) c•-) N N ,.... -- _-- ..--- ,, "--- , 4 '1:' ,--. rn n... r-- r, ,---. cl. r-- r-- N — — r-- N: t--: r--:. t-:. N 1--- N c--:: r-S r--: N: N ...... _., ...... -....- -._. • --.., ,...... "

-7)

1-)

(1) a4 (4 (4 (4 (4 (4 ,4 o 0 124 00000000U-7 > • > > > 7ict U O 00000= o uououaatat ct 0 qq,f) RRRRR U C-) c_3" O ,00000 1) 0— (-)U(....)UL.) a) at at azt • • • • •-n • fal:;_j cdri 1' )"•'•)- > > > > o ct at at a2, E!: Et: Et: Et: G; G; 4

uo o oo cdo c)o on uo cdn GO GO 00 GO 00 DGO 00 C.) DO Do Do oo DO (...) 121 D

n--11-i (:41:4'.)24a4C4-1- a4P4" e)-4» ‘» n-4' 1-4

00 00 V-) vn vn vn vn vn vn vn vn vn vn r- r- vn vn vn vn vn vn vn 7.1"

• Cr.; Cr; CC; Cr; 3 kr; cn rn 71- t .Tt 7i- 7h 7t 71- '.7t. 71- 71 - vn

vn vn rn rn cY r_ raas, r at„ C) 17,;1 ,4 en7.; 00 r- r- r- oo oo oo oo oo oo CN ch r- r- r- r- r- CN CN CN CT CT CN r, rq NN n1D N N r- r- r- r- r-

cY cl cq cY rn nID Cl ec) 'Tr CD 7h CD lD cn 71- CN CN r- r- co oo r- r- oo oo oo CN CN CN▪ CN 00 CT CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CT CT CN CN CT CN CT CN CT ,„ ,„ ,„ , „ CN CN 00 00 Cc) oo r- 00 vn r- r- oo VD CA C) CC cq cl Cl CD CD Cl „ „ Cl cl Cl CD CD CD \XD nID 00 CD CD Cc) CT C) (= 00 (c) CN CD C- c-F5 n 1n1 /-0 •—( 1•• -1 1-1 /-1 1-1

rn cn rn rn rn rn rn rn rn cc) rn rn (-NI cv cY c,) cN) cl cl c,) (NI cY cl cl c")n c".), c. c:)n cn cyn c:)-n -,,,cn c:IN ,-)c7N c:),-, co,-, -, ,-) ,, -, -, -, -, „ -_-,_ ,_ - , -. S5••• oo oo c'.> c) -, -, Cr) (-r) 7i - •—n ,-( ,-1 .--1 ,....4 ,-.4 .--1 ,-) ,„C>C:;) ,--(,„ ,--I,„ ,„ ,„ ,, -.., S-- 00 00 00 00 00 00 00 00 00 00 00 00

00 CD n•C) Cl VD r- oo N r- 00 00 CD CD CD CD C) CD CD CD cr) cc) cc) cc) cc) cc) cc) 71- 71- 7h 71" ▪

71- ,..--, ..--, vD ,-----, — m ,--... ,---, N m ,s r-:: •--. .---, ,----,,----. „---, ,---, Ln ,--, d- • N .-..--• N ,---, N cv cv r- ''—'• - '—` - 'd. '71- ,- \ '';' N. V-: N.: '.1. -- CA - I-. .-1 '-' r r - ,C) r"` ,--, N..' ,--. ' ‹; r-: ,-. ,.., C4 r`: r.: 4 r-:' r-:' h: 1-7 h h 4 -1- . h h . h ., h h h ,‘, , ,--;' , , —, u-,, , —, —, h, —, h, —, — —

7:1 7i 7j •C SZL ,

6or . 5' 4 . rELl' ' 7:—,3 ct CL) CL) Cll m ,-- 4 = = m m s:L., cu 1.) cL) ci.) --â) a CL) (24 8 •0..4 •0 . p4 lx 72 JD e.. ej c;.:)' al = ;LD, 4c74 qc,V-C, 7:) a.) 1.) ,9"' Û 77i 7i ,.c 7i 77,) • . • N.I o (..) 1-, ... •- • _, 0 • -, • _ o c.) c.) 75 75 75 sn., pu, Q) o o F7? c9" g g g-',, ic4 (1) 4, cL,

q -c9(1)(1) (=4C4,11,1t "c'Ec)b134:4)t'n---4çn'''''"tc)mm -, 8 8 .a gc) rP! _12! L) 0g ,, =1y) ,=='v), ,..1.) u. u -c) m ,, . . z : . , . (i) to to to ID ID 0 C ri 0 anQ) (1.) 0 0 o 0 ,4 1.-:', On c:,-, -ci.) a) u 2 ,4 C) CD CD 4 4 4 4 cf) N O.., 1021 XI (4 4, 4, = < U U

u Do GO 00 GO GO GO UP 00 UP GO UP 00 U° C4 00 °0 °0 °0 °0 °° C:) D a) a)

o4 o4 ).-5 4' 4' 4" a a' o4 4' a a' 4' 4' 4' p4

• gD 7t 7h r- r- oo Lin co 4 4 oo r- CN CV CD CD r- 7. cc) Ir; h: 4 4 lei tri oc; 7r oc; oC cs; Ir; tr; \x3 oc; oc; gD r- r- r- r- r- r- oo oo oc oo oo oo oc oo CN Ch CN CN

az, .71- V0 n.(D cc) cc) VD nID 7. .. CD CN cc) 7. Cq CD CN cc) cc) VD CN O CD CD CD CN CN V-) V-) CN 00 cp CN CD -. r- 7t vn cq vn vn CD .. Ir) CD Ch tr) V0 CN V-) 00 r- r- tr) r-) r- rq cq rq Cq CA Cg CA Cg Cl 7. Cg CA .. 7. Cg Cg CA

.---1 ,--1 •—( ,-1 0") 77 7t 7r CC) Cl 7r Cl Cl") 7r CD Cl 7r a:, 7'7 7. ,-.0 ,--( CN CN CN CN 00 CN r- 00 CN CN CN CN CN CN CN CN CN CN Ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN .CN .CN . . ,—, . . -... . ,-.1 ,—I --.1 . . ,--1 ...... -.„ --. -,, --„. --„ .--. --. ------,.. -,.. --. --, --, --. --.. --, -- oc) 7t c- 00 VD 00 cc) r- cc) 7r r-- Cc) 7r l:) nC) 7. 7t ,..1 . . . N n—n c.... --, N (0) 0 , —.1 N c:::) N r-1 0 N —1 —1 —4 —1

- 25 C3 . 33 c.1 -. r-) e) cq m -. e) cn -. C3 C3 C76- 4 213--, 215--, „, ,, --,25 --,23 „, ,4 .4 ,-,

rn rn rn rn ron rn rn rn rn rn rn rn rn rn rn rn rn rn rn rn cl cl cl cl cl cl Cg CV CA Cg Cg Cg CV cl cq cq cq cN1 cl cq CN CN CN CN CN CN CN Ch Ch CN Ch CN CN CN CN CN CN CN Ch CN 23 „„,„ vn vn cc) V.D ‹.) 00 00 gD 00 CN CD CD .. CV Cg Cg CA CA Cg Cg cq cq 2.72 27, oô 36 oô 3:6 '36 OC e)76 oo oc oo oo oo oo oo oo cc oc oo oc oo

4 --I 71 CN 00 -. CV Cg CN 7r r"-- 00 CN CD -. CN CD -. Cg rnn C ,- rn rn rn 4 'Tt 'Tt v- v ti") ..‘D 7r 7t 7t 7h 71- 7h 7h 71- 7t 7h 7h 7h 7h 71- 7r 7r 7h 7t 7h

•

CSn 00 Cc)

,----, ,---. ..---, N 71- ,C) ,'",, ...". 4 ...... N r-i r".., r.... r''.: ,45. RS s ,0_ „,----- s ,I- - ,-, , --, 4 ,--, - N t-7 --. s. - - s ,i- - ,i- s , o: l---: 'r: kr; c•-n c-,:' (---: t---7 --: t-: --:. 4 „--n 1 --. —. f-- --. --, --n ._, so --. ,i- ,--, r-- i n-..-, s....., ,....—, s....." ...... ‘ ...... , ,._...... , ...." ,....., s....., — ,..—•

lj 7j 7j 7j 7j 7j 7j 7j •.—. • • • • • .•-n • • O000 r24 1:4 ° ( C CCS L) V P..9 ?..° UUUU c Q.J CL) Cl.) CL) (1..) CL) C.) u CC #c7_,S E - -taa) '2a) Ea) -taa) -sa) a) u Jaa) c c CC O 000 .g ZZ ZZ 87i (7

c) uo oo coo oo Do uo 00 GO 00 U 00 00 Do 00 00 00 00 00 00 00 (21 121 (2! (2 D (2(

cq on on rn CN CN CA VD VD VD 00 cA CA cA CA CA CA ci on cq 06 O cr oei 71 N N N N od 6 N N N N N N N N tr; CN CN CN CN CD CD CD CD C CD C) CD v-n1 1-1 1.-1 1-1 1—q r-..1 4-1 1-1 1-n1 1-n1 1-1

71- 00 fq CA CA CA CD CN CN C) CD 00 00 ":t CN nCD 00 cq "n cq 7 ,f 71r - 7t AD AD 00 00 (r) cc) tr) tr) CN CA 71- g 5?) r- r— CA In In ln In cr) c") C- r- on c,1 r- r- vn cq vn cq cq cl ".° oq cq oq cil N cg -, cq N CA CA cl

-, oq c) on -, cv on cv on on on on ch ch ch r- ch ch ch ch c, ch CN CN CN CN CN CN CN CN Cr\ CN CN CN CN CN CN CN CD\ CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN cc

23 .743 CD In o0 AD 0 00 CD 00 r- r— tr) CD r- vn r- CD ce, CA CACD CA CA CA CA CD 00 -C1 23 e) cy e) 23 23 --N 00 00 23 m m 23 m

on on on on on on on on on on on on on on on on on cq oq csi N cq cq oq cq N N cq N cq N oq cq cq O\ ch ch CS\ CN CS\ C C CN CN CN Ch CN Ch CN CN CN --I CD CA i":-A cn 71- 71- 71- Û"'"i on on on on c.) c) c) C) CD C) C) CD CD

06- ;6 ;6 ;6 2is ch 2is ch ch ch 2J-;* ch ch ch c, CN CN

cy r- ch c) un AD CD cc) r— CA cc) 71- r- oo oo 00 00 CN CN CN CN CN CN AD AD r- r- r- oo —4- —4. -+ —4- —4- —4- CD /—.., ..---. ,. ....---, 0 "•—•.. ,----. I-- " r... ,..-.. N N ,--..... h •:t N .. .—. ,....-... Lc) v-., N —: r---. .-1 n••••n ,.0 " ,.0 .. . 7r _ t•-.: N r--- d- --.. N.. r--: N NN N— I----. —: —. — v-, ..r-, .._...,`...... — `...... • - `...... / s...... , ...._. ,.._." ,...... - ...... ,._...,

QQQQ -c) QQQQ ôôôôô ô ge) -F;73.'73.' (1 cut,EE (ID c40 a4Ct c:j;

(J4_a Qt IL iL LI LI (I.. 1 C571 L)It LI

tll a..) ct) . . • . • • • • • 1-1 (ID () 1)// UC (C/2 (C (C2 ceL) N N C.7 C.7 CD 4 4 '4 2 2 2 2 2 2 2 2 m

23 c'3 2 23 23 23 23 23 23 23 23 23 23 23 23 S Ed2

N N N CA V) V') kr) oo oo oo oo oo oo oo ao vn If 0000 on kri tri kr; n6 .6 4 4 4 4 -4 4 4 4 .6

1-1 1n1 11 .n1 v-1 1.n1 11 1n1 ..... 1n••1

c)0 71" (C \C) CN CS\ CD 0 CD --, m m ch chNN ch r, 71" 7)- 71- 71- 00 00 CN 0\ CN CN CN CI\ CN CN v.) c.,1 c c 11-) V.) n n cq („1 („. CA Cl Cl cq CA SD \ID CD cl cq cq

CA CA Cc) CA Cl Cl \.0 Cc) Cl Cl Cl Cl CP\ C\ CS\ CN r- CN r- co 00 CN r- ch ch CN CN CN r- ch ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN -„ -„ N r- CA s.ID 00 CA CN 0000 Cc) Cl Cl CD Cl 0 Cl CA CA Cl Cl CA Cl Cl

' C.) Cc) C"") 23 r- C--f5 CN C) 00 23 23 23 23 3Z5 cq cq oc

cr)cr) Mtnt-c) cc) on on on on on on on rn NN NNCN1 N NN cq cq cq cq ci cl ch chCN OCC C:rc.n 0\ CN CN CN

SS. C.7.! 1‘) ttn tin tin n.c \.0 N C C

c) c> • c. oh 0\ 0\ 0\ CT 27-ç ch eis o, g

CI cr) 7tN rc) 00 CT 0 Cl .CD CC CD C) CD Cl Cc) V) II Lr) U") V.) V) V.) tr) -

Cr)

,...--, ,,-..“--- ....--, ,C) •-n ,C) /.... ,..--. kr- If) C-• ,__, „_„ ,__, '1',__, kr'; CA - , r": e.". l'..- Q"" tri r-- ,r) C--. r."- t ,-. ,•O "....,,..../ ,..J nJ‘..." ,.../"...-/I ,...... "

CA VD V) C4 CA 0000

--duuuuuuuu . 5%,,:.,0 ouu O 22222_uuuuuuu(..) 0 (-)uuuu,o).040.040M ct ,,„,,-mmimmmmmm u mmmm 0 00 u uuuu-a000000.›-››.%, 0 u uuuu.mmmmmmimmin.mmmm„omm ::)21a)(211=OZ4

uu(Juucnc4c/c4cnUcnuUcnci)U

cua) mmmmmokninintnkr

mmmmmm ,71-7 ,71-71-7,

<4 <4 cq at) CD -o 00 00 00 00 00 C) r- r- Q N oo on oo C) r- r, r- r- cl „ „ „ cn on v) 01 01 cr) CD Cq COO NN NNN NNNNNNNN-o--,

0101NN000101C--00000101010101 01 0n 01 01 01 01 01 1 01 01 01 0101010n 0101 01 010101010101010101ChCh0 - ...... 4. 4. „ -.-„ „ „ .-- 00 r-- -‘zr 00 CD Ch ce) cn In CD n1) 1 -1 —INNOONNmm N—IONNN-oNNN- -.- 0- C) oo oo C) oo oo Cr) 00 on C) Cr) Cr) .(:5. Cc71-- 1-1

mm mmm mmmmmmmmmm NN NNN Nc1NNNNNNNN (TON Crn C'n ()C7\c,NCr) -o -o -o -o —(-omm ,71- 7tmkr 1-1

aC di- .- 27'-,

-oNM 71- lind- Ch-oNtr)N ,40r- vnvnivn vn vn \.4) n.1) r- r- r- oo oo 00 vnvnif”) Ir) In In V") kn kn in 'In In V")

N - ,---, ,...--, ,---. r ,---,

n.n r-S v-, r-- r- 71- r- r- r-- r r- - N r-- v-) _ .. .1- d- ,--, • r- NN n - - n 0 ,- v ) : r------. r's ^ ,—, r-- - ....., •._.., __ 1 ..... ,..—, s...., ..—, ....., ....., ,.._.., ,-_-, ,....,r-- ..._.,,--. .—.., . ".n._..,

77 77 al., ai., m m a4 4 (4 (4 g g g 8 _ 7...__,,, ,....._, .— a,.., -0 -0 .-0 _ a a a a a a = = sa, sa-, szi- 0 0 0 0 0 . 2 (.)., ',D (,)., -5, 4-, 4-,›, ›, ›,•.- •,-, cd cd cd cd 3 •,-, • -, 3 cgu (..) g ci,t > c,.s (.7s. (,,, sz4 s:4 0 "1Q ''''''c...)' "1-'1 ".`"U U ''-c_)1 cl U =ct c71= (71 ct cn cn cA ) C.2 ct cd -, --1 ,- ,.., -d 0 0 0 0 Ou Ou Ce ) C-a.)) ",---a, -8 c, 1 ' ' - a ) - c' D 1 - - ) - - 7d 771 7d 7. ci Z ,c.) $..., a.) c.)i. s.c.) 1.) 1..) w w w w ,,, = '5 — ,. (-) ,) >->ct,t cd ct,.,,...),....,..)uu, 0 0 , Ç,) -',.-i. •FE1 0 A -5 ..c) ,..0 m m> m> m> —6 ,r:?) .0' 4'. JD u)4:, 4D ch z '65 c) › < < ,_. ,_. ,_ .a x c) c) CD CD C) C) -. -. -. a.,

C.) coo coo oo coo coo oo oo oo oo Cd0 DO GO 00 00 GO 00 00 c") GO Cd0 D r:4 P4"

Ir) CD v-) d- CD lD CD 00 m oo tc) kc) oo oo oo vD O -11 o6 406 c; C1 C; O C1 N 4 d: 6 Cf• C) Cfn 4 d: d: 0,6 AD N N N r- r- r- r- r- r- r- oc oc oc Ch 00 00 00 Ch Ch Ch Ch --I •-n

rn Cr) 6-N S,'. 2? S ;1 (13 c,z a a a p N r- r- cc-) rn kr) Ce) r..... r, r, r, r, C7) C't 'cl CI '4) '4) cq -. cl c,1 cq cq cl

- n CD ,--( --I CA 'rt Ce) Cl.) CA 71- CD --1 --I , ( -- ce) CA Cr) Cr) cl cc) d- CD Ch Ch Ch Ch Ch Ch Ch r- ch Ch Ch Ch Ch Ch oo r- 00 Ch r- oo ch ch ch ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ---1 -.0 ,--1 -.1 ,-4 ---( ,-4 --4 -- n -.4 ,-.1 --n --4 .-/ ,-.4 -1 --( --n ,-1 --4 -0 --I -_-:-.... ------, -- -- -, -- -, -- --, 'i--,-.1 --, -, -, -- -, --. -, -, --, CD 00 rn CA --( -4 CA CA Cl II-) --.4 CA 14-) ,---( 4 CA i rn cl cl cl ....„---4 --I ,, CD ,„ CD,, --1,_ -CD , 25cl g Cl g g (1.'-,0,.1 g .7,.., ,..ci, 2 ;;1 p --1 CD QD Cq rn rn rn r- rn cy c) . —, 1-1 r--1 T-1 .--I ,-1 1-1 1-1 1-1 1—.1 ....4

rn rn rn rn rn rn rn rn rn rn rn rn rn cc rn Cr) Cr) CA CA CA CA CA Cl CA CA Cl CA CA CA CA Cl CA CA CA cq Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch --4 ,-.1 ....1 ,..-1 --1 ,-.1 --i --I --n --1 -..n --1 --4 --, ,-.4 --/ ....4 ...-1 - , , ....„ ,, ..„ -.„ ...., -, -„. ..„ ...„ ..., -, ...„ ..„„ - VD r- oo 00 00 00 00 00 oo oo oo rq cq cq m m tr, kr, ,.., . --, --, ,-, ,.., ,.., ....-, --,-, . N cq CA CA CA -,Cl cl Ch Ch Ch Ch as.--.,-Ch ai..., ai-, ai,„ aFs, Char--.,- ai-, ai-, -,--- Ch------Ch------Ch Ch

Ch CA Cn VD CD 4 or) u...) n47 Ch CA ce) r- lD 00 Ch C7 o0 Ch Ch Ch Ch CD CD 00 CD CD CD CA CA CA Cl CA CA rn • tr-) te-) tr) 14-) VD 00 VD VD \(7 nsD lD V0 VD VD •

Cr) Cr)

,--, C.: ,---. ,--, ,----, ,..-.n ,...... ,----, .-, ..--, C-- Lc-, ,--, ,,---, ,--, _ .. ..--, .----. ,---, • r--• Lr) I-- r-- r- r- t--- t---- r- r-- r-- t-- v-, r- h: -.. r--7 t--- .. h r- „----, r--: h: r-: h : - - s- - tri r•-n ,--. h ,--n -r-: - t --7 ...-, •-..., ...-, .... tr: s' - ...... N.-.., ,-, ,-- ,...... , .....- ..-- ....." ,_, ,--, .._. ,.._, ...... "

a a O 0 a a of) bio Q Q Q Q Q Q Q Q Q Q Q = = U U :7) QQQQQQQ QQQQ . " .E cip cn SZL., n 1--, ›-, >», -- >-, ›-n >, ›., ›.., ..., -- -$.?.., -- .-; 1 t).0 b.13 ct Q a Q Q 4, ça, ca., = a4 G Q Q Q Q Q Q c't E cr) cn •, -., • L-_,. m 't C`" ``3 c`i c7i m ct czt czt n n (::- a a. Q = sfa.,- ' sa..) uuUUUUUUUUU ou ,2.2 cu 1..),2„),:.;_ci)c/D. t).13 tel b.0 tiO iD40 tO a a a .,„ .-, .- ...- •-, cl, Q a a a a a czs czt czs CC e" :1., .1.) a) ..-, • • = ct cl a E E ,.., ct ,_4 ;_ Q 2a" ..= —4 Dn c‘n DO cf) C00 Cn 1-, )-, 0-1 1-1 C'.5 6' (_ 0 CD ai a. E-1 E--( c.

c/(..)c/Dc4 c4 cn cnvpU c4cAUUUUUUcnc/c4Uc4

a4' a4" P4 12' 1:4' P4 1:4 0-'‘ f:4 n-1

71- "1" 71- 71. r.".• r".. 00 00 CT CN CT CN CeN CN CD CD lin un cin 4 4 4 4 4 4 V3 V3 V3 V3 V3 o6 o6 oc; 06 o6 oc) cri cci tri c 0 c) c) c) c) c) o c) c) c) c) c) Lci cl cq cl cl cq c-1 cl cl Cl Cl Cl cq cl cl Cl cl CA cl cl cq Cl cl

• cl cl (7) v) .7t le..) cl nO %.(D cn m Cq c>00 000 ' c71--1 cNr- rr-- rr-- 7t. Clr- Clr- r,Cl Clr, rCl_ Clr_ 71- 71" r, clm CDr, N N Cl cl

Cl 71- cl m 71- cl m m m r- c r- ch o. cs. cs, r- C r- 00 00 r- oc oc oCN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CT CN CN CN CN CN ON CT „. „, „, „, „, „, „, „, „, cn cn cfn r- mcnrnmcn te) q.r.) lD Cr) O cl CD Cl ,„ CA cl cl CD CA Cl C) cl Cl CD Cq Cl Cl_ Cl CD oo 00 CD 00 cl 0 00 C71 o0 o0 oo oo CD m

cc) Cr) Cr) cn Cr) Cr) cin cin cn cn Cr) m m m cl') m Cl cl CA cl CA CA Cl cl Cl cl cl Cl Cq CA CA Cl CN CN CN CN CN CN ON CN CN CN ON CN CN CN CN CN . ,-, ,-, ,-.1 . , . -, . ,-, -4 „, . ,--, . . --N --r- --r- -172: --r- F-lzr- -- i:r- -- oo oc Q ers: (.1:5 (T5 2:3 N cl cq ci (NI cl c.1 c.1 cl c.1 -,cl c.1c,..) m rn m n CN Cis. CN ON CN CN CN CN 27- CN ON CN CN CN CN ON

Cq Cr) 71- V-) lD r- co o\ le..) lD r- oc CD cl (-r) (-r) Cr) Cr) Cr) cn Cr) Cr) 71. 7t ,71- 71. V) Lr) V-) \C lD vD %%ID V.") \C lD \.1D lD y:D

71" Cr)

kt-) ..--. N .---, 4 tr• ‘.0 ,-. .-• .---. . ,-----. .---. NNN N sc) ,r) " NN N. --: N7 r: v-; t-: N N N N " .--. Lr) •-• — •-•. kr, C/) '....—..-,,,-, .-...-, ,....,,--,

Z * 0 =0 =0 =0 0 0 0 = • Z = = M = = .'-'>, J-, >, >, >-, >, ..-1C:"7-) C) _9,5?,C) ,Cr2,C •,-..••,-•--,) C":) .9.,' M"6-'ggcggglC'El M CO CO 'c7'CSc473c477J''Ft'l"U 0 U 0 L) L) 0 U 0 (UU(..)(..)UUU80000 c]..") e.3 (1.--i Q) u I) c.) u u 0 (1., -o-to-t 22222222222=== ;' D =c`'.. Q=u u u 0 0 ou c)0(:)EEE N 6 05 6 6 6 6 6 (7.)' c"))) c9 (.4 C,)C,)C C/] Cf) V) CO CO ,--1 Cl Cl Cl CA Cl CA CA "a 42' 421 ,a "a) -t1 "a ".° "a '-' "a .'''' cl cq c,1 cq c,1 (-NI cq c,1 0 CD CD CD CD CD CD CD CD CD CD 4 0Q 011P

oo Do Do oo c) oo oo can oo oo on on oo oo oo oo oo oo oo oo oo oo ;=D

n„,c) (y) -. on on on cl cl on on un un tin ten un un r- • C3 CD CD CD CD C3 CD cl cl cl (-/- or-; o> 4 4 4 4 tri tri Y> cq cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl cl rq • N cl cl Cl CA CA Cl Cl Cl CA CA CA CA CA Cl Cl Cl Cl Cl CA Cl

- Cc) Cr) 71- 71- 71- 71- -" -" cq " CA "-" CACl,1 cz p cD Ln Ir) ,n,( CD CD CD vn CD un r- CD CD CD r- r- cl c c - cl-" cl clr- -"cl cl vcl.) cl"r) -" -" -" clcg") clen r-r- oo oo oo oo oo oo oo

ciciclNNcic-l-l4m-csImmm n Coor-oor-ooc›Chr-oo r--CNCOn C7...... -.,....,,,,0\CYCNCYNCTOCNCNOC .... ------..------,------,-,--,-„-,-„--„-- 400w00000000d- cnNuntNONNr-tr ONNrINNNNOONNNNONO-INNO-, --- - 00 -,-- - 6, Pz (N C-.-1 c.:7-1 cl cl cl CD oo --• ô - . c h ,-1 00 Cr) •-- .-- •-, •--, •-• •-, •-•

on on on on on on on on on on on Cr) Cr) Cr) Cr) N cl cl cl cl cl cl cl cl cl cl cl cl cl cl CN CN 01 CN CN CN CN CN CN CN CN CSN CN CN CN ,-, ,-, .---( ,-1 ...-1 ,-( ,---! ,-, ,-( ---1 ,--1 ,----i ,-, ,--4 -••-.-•-...., ,...... , ...„...... , ...... , ..... ,...--( , . ,-1 ,-.4 ,-( ,-.4 ,--1 ,-, ...... ,--( N --- •—•--- -C") .C.) .(--) Cc) CD CD CD CD CD CD CD CD CD CD CD CD CD CA 27S CD 25 CD CD CD CD CD CD CD C3 cpc, o 8-'\ „. -. „. -. -. -. -. -. -. -.

14-.) vn un

cf-) 00 CN CD --1 cl 4 un tri un le) tel le) un un un un vn \,(D COQ\..0 n.0 VD ,40 ,0 ,.0 ,0 VD \40, • - ▪ ▪ •

tc) Ch cc) ,.. 0

C.) C.) -o "c• N 7:1 r`: PI, a.) " •—n )-• N V^) Ct •—• r-- kr, 1-1 o 00

-0 -0 8 >, — .—— cd cd ct ct r:4 ô C.) G.) ,- a.) a) q ,o ,---! 0 q..) (...) to u u co cs .•-•a a..) '5' c.) a.) a) a.) 0 0 •-' a • -, o 70 G.) a.) a) a) czt 4-' i..., 6 (....) 8 >' a) a.) c...5 .3 u u mE () ct c.) 0 g › .. "c) "cs -,i -t:) H — u • 4-, 0 000 •— a) PZ4 0 (:) o (D o a ). - ri . , 8 o 5 0 7 (4(..) -0 -o C.) 0 o ,..< cdcd n Z5, ,n al ;... (1.) cn 0 .,--.(.-• ap pp d E-. 0 .-•--. -5 cc) a.)cl. ,-, L) = cd G.) , ,,,o -= 0 C/D C/) C/D CID CC) (..) • v..1 1., (7; 4—( -rC1 › ?; (,-1 -.,: C) 7 0 (1.) (..) u) - •,_,-, .a.)"., 4.) c..) "c) )-4 f=4› i24 -5 '-.6 ct o •,-, a..) cu --, -0 -10 ,.. v, 6 -0 -2 .•,: •-= • O • •--,,__, a..) P. P. .,_, o r- r- r- .- . c.)-. 1.-i tri tri tri q3 .,_• up cl cq CA CA CA CA cA Cl cc) 4-) • Li-, ,, a..) Cl CA CA CA CA Cl CA CA Cl CA c) ,..0 e3 1.) E 0.0 0 .• ..---.-1 7, = . --,z • -"'- 0 • ,.. ---, ',. a'' -71 .-(N 'ci!) 0 up =1,4. '-',,, cd a.) - C) 00 CN cn cc) d- CD ,. ,..-0,, r- r- CD CD ' 14-) CA -' VD D VD N tc) VD 71 'ai f 8., '=`-5 (1),z CA CA Cl CA CA CA r- NN r- .S..) a) fm. ,-) 1) bi:) P--, cz: ct = b _oa) c..) , 4) z o ;.., O cl.) i:4) -51-V-~1;_, ,.. ,a) ;_, 'a .,-i . g o' 3 -6 cc,4s m — bi:).m -ci 'te.'" d z ,4 (4 --' (I-) cn on CA on CA Cl CA CA (4 0 › li "" "d (7/ CN CN CN CN CN CN r- CN CN r- C/) Ca$ "7) ci) a) a) 0 '4 CN CN CN CN CN CN CN CN CN CN (Ci --• o c/D o ' 3 g. .> N (.- .-: c_,2'y › 4 . ...j 75 .b.1) : : '20 Cq CA CA CD CA r., (1) ..Z ,. C...) ,,, 0 a.) ,-,,, cr, •C CCC NCOCNCOO ,4 0 )-, .E -)••- L) a) 8 cl 6 ,,, = (71 w = "0 •-• ''' P. r2)., o . .(1_,) cd 3 ...a9 P. 0 - › . o O o •—• vi O C.) CCS 0 cl.) ,..?., c.,a9 . i",71 0 ei..) $1:_j, 0 (/) 4-n on on rn rn G) .sz -- ••-, = m rn rn rn rn •. -I-, v.) (/) 3 a) V = CA CA CA CA Cl Cl CA CA CA > (1..) . .--1 . ,_, S. 0 CN CN CN CN CN CN CN CN CN c) 8 __H/D' 2E, zr -c(i) c.) . di, • '." › 4) 7,5 E • 1'3 n a) 7J n•''' (1) VD VD \-(D VD r- r- OC Cf) 7:1 -Ci cd 0 a.) •- a.) — CD CD CD CD CD CD 0 --I ô 2, c,3 c,: .•,..- .>"" ' (cr.3 "E CD CD CD CD CD CC CC bi) .4.. ^ "d "Ci N O 8 0 ='' cl-) "ti Z cl 0 cl tb ;... 0 0 :0 -rj (..-1 4-, 0 v) 00 c., •-: > . R4. ›, clei c,5‘ 2 .4), up . ,. , ra, s., . i::: iC (1) -. czi o cd >.s ,Jj -0 -0 u) •'111,-, •'"".—. :-.ct . ,..•'---' r- 00 Vn r- G) (1.) G) 1 -. 0 On ct 12 r-i r- 00 00 00 ch Ch 0 d- c.L' .)u (2i 121 w cti G) \I:, \ID \CD \ID \ID r- r- ,---n cv ce-, ,i- t---- 00 oN - ,. • • •

cr) -ci z cz a; bo ,..= z 0 o ›, 'o = (.,-, m -,:i C..) a) ai a) F) o EL, (..7 NI a) = ;... 0 a) .5 O at 0 3 3 -0 r) ,.. ,z) to '5 a) o -d ,.= 0 m 3 (4 E a) ci c4 a.) .) Q - c.) 7,.' a) ;_.1..) c71 ci -c) › ,5 bz = • - czt ct .c ..= cd' o aS = (u" Ôn N q-•• al -Z -• c) = I"' m ›(1) a) c...jo.z.' t.., . Q ..c5 ,L '-z a)E-' Q= cl.) ai) b13 b1).,„ • ,- a) ,4 .F. t W .-, a) 0 a) .SD CI) CL) ,—,s:,., ,.,7zi ,_ cl ,S ct Q 415) 0 (1) "CJ (..) ,-. Ct c}y 4—t = E 0 0 0 'F,,,s' to ,,, = . ....,- •z - •.9u., "cH- '- O 0 Q., 0 -cf „_t to a.) 'F's .,,3 .P ., - a) 0 al., ct-, - ci,a) a) ci -6 cz:;-, ›. a) 0 (i) -(i .z 0cd .2 ,.o to 'a" , • = a) 1=1. .;",4 8 (...7 161' T'j c/9 -4 ct I- (4 ,....) -cri- = 3 v) -cs — ..g . 1,,,) azc 2n 8 .F,'n n ...,b.0 ›, a) ,,,D ..- 0 .- 0 0 a) Z czt c,, g -5, 8 0 C4-1 ..C7' M 0 cA 0 .,- cn Z • '-',._•-_, =: n'/),:,- -..' U cu tt-j.1 ,2,›,. . ›.'-' o-- d) ;-n (I) • ..."7',_ )t 7:1 0 ,.. cl) ,C . 2._.), .'" 0 (I,L,) u Lrj L) , ct E U 1--, "+= = cl.) tc:,..; 01 c.A• .,_,U X .7:ji ct cIn)c E -E =`‘ ,7--1' g z LT..1 7:3 o f:1-, sL., L) -- 0. 1) a) — ci, (7: • - o c7: a., -4 zoo-ic4a4 c4 ci) ,.. - c-N1 , ,i- ,r) ‘.0 r-- GO d ,-.1

• •▪ ▪

O (4,1, c\1 N CT a..) ., cr) 0-• › ,--,- ...c.) (1) <1-) ..9• c, ,5 ,cp •==', I q-'c .,,, -5 ,•-t, ,— ,., ,,, ca ;... 0 a, ct .4 ,__, (..) ;6, __, . tir-) 5 .4 c.) (., ;-' = 4- › o > 2 ,...- .„ •— 0 . 0 c. ,L) , 0 s=1-, .1,,i 0 ••,„ • r• C) Ct ctl ,::.) al, '..' fa, = ct O •{n1 ;,C74 Ct,1 . ,4'' .- ...._ . V) 0., 0n 0 tli) (1) ;31,, CJn C...) • 1• 1 , , ,.2 .›,--, ,.,..- ;` _24_ ---''. 0 . 1. cd a..) —, "0›

54.9. a) ça., ,..--' ',":5 ;-•n `0 7:3c7.: • 34 1.),, 2 - !Do E , 0 ,.= to3-10 s.., -,c-i, `•-4 15

0 (/) t., ct z 4_, 1_, cA = Ô 2 -Ft .,-, ••-n u c7i 4.) 0 $,-, a.) 1.. ,.. = ,4 P 0 C.) 1-1 y., . ,_, O ,-= 2, f:).., 0 ct el, 0 ,L.), cci 0 a.) -0 ac., --, - 0 a) L- ca., •--• ,4-. "a' Z'z, ,:t, , c& 2 . -E. ,.`,1 .t.,1 a) Sa. CD 0 b0 ,,.,L.,0,.,-"' (t)upa)ct 0 u Np—, .0 c.) at E 42 ct ,..a 0 c.,.cd ,. vp a) fm-0-• .- ,-, .= 71'.) ki 1-) ---f.' • ct Cl. LA. = .,,, '›.-' ,--• ,, ;-' • to a) 7'd . R•76 a) ',,-, g z (1.),....,-, ct . L) , ,.. ‘ 9_, o 0 = , ,.> „„ —. A -0 0 = .B.'") 1.... 4- 'c's 4c-A c.)< '-' • - ct ,4 c1) - r: 0 `c/rs', to 2 1.) 0 :-. ' o --, L) 0 V, O 7,,d 0 .,,?-, , -" 1) 't -z "s 0 .z_i ("1 '1-) C-) (1.) '-' ,-41) ()- ••• CL) , a..) tP1.0 ct "0 TD4 .0 ••-n ,.. c+. -z, u 771 = ,+-, 6 ;5 ,.., E 0 — b-0 — o (...) ,--,•, o -,:s - •- - „, — c, ci, 0 — a..) 0 TD, ,.r :)-, ca".3 = .c.,) ,_ ,.. 7=1 c.) = , .4•-• = ,L), up Ç. , ,--:..' °) ci) . q `" "ia% :6 o . 7,-; 3 _ cu .E -cs ,—, o E (.7 73 .- I 2 E 0 0 czs E ,,•• .F, ct • -o ,, -5, ,. -ri .,c,,,) czt, 0 ,•-•- ct VD = Cr) WO _1, z 0 = :4±.1 -"' .• ,- ,i)1-1 (1) • ' 3-1 ,-Z F; 75 1.) ct g 2 c--/-i- ,9-, -) ;,. f54- 8 ::,., _. ,..) c t ..c c..) m bi)..--, VD IR', r, r, ,' j0 Cd a) -5 0 . 7,,,, u ; . - = -c'7, t - — - - 6 »,-A cu ct td) -5 •,-, '3' 0 ..(-) 2 •R u t g, 45 -5 .- .4 ,.. -,- -,t1 .--, - u ct 4-4 4_, ca, o i.., ci.)'F -1. sa, ..c-A ,t5. -63 ---,, ,+-, (I,- =- 2 czt c:14 ,. = 4-, lazz.-,, , a.) rz. , 0 m'-. , tic )... sz).. o c, — . -0 ,.., a) I., 0 Q) czs up ,.._. ,-, cl.) t = Es-, c)•,- ,.....• to;.- 0..) , = •- ...,(5- -1 =nn (1) 4.7.— ,,s -, ct 0 ,A» 0 o O CII t " d" ., < L S al . , '6 , n,' Cl D z 7:0 2 b.1:1 çp..• ›•-, 0 .., •-c) 1-n › f:4 —!--)a)=00 —, a) ...0 ,t, E cd ,,,, ,․.) ..fz2 ,_.-- ,, zd., ,s 0 _E Q) - ci ,,,..) ...0 .- E -,:1 0' - •›. c•-,, -c,i, 00 - ,,, ct -c- d cz, .c ,(1..) z c._).0 ci;44, •_, 0..)0 -- - (4- E`0-' 0 E2.3 .4 z o 3 bo r › t) • › .2 4:, E -',-'d s--, 78 0 .,t 0 "cii r- (1) 2 0 -n-• 2 0 •-cl • ct $_, 0. ,- cip c,,s" ••-n ,4 ct = 77/ a) a) 0 4-' a) M ct M 0 77) ,-, c-)r, Z 2.',) t•O t.)' ..-n up `/D ,0 e-) y, ;-• -0 ,, t,--• -•, •Z czs c+, ( 61 D cc I / t p) § , -a.) ' t:c 1 ).). . ,. . c.1 2) . CD- 1 L) n 1', 4N . e . - d! ) ' - - (. j _t% , C, , ) c 1 . .,...- •-• s., .c,,, 0.) cf) 0 Z -- ..c.n a..) ...."-' ••••-, , ---, Or"--- Q) 1:2., n ct a) 1-.) (1-) Z (.4- a) • t5 ›-, ,-., li 0 '-' s. › 0 b.0 E E ct 0 E, (L) = ., -n••, ,--.;‘ b.0 E .4,-- 0 , ,,, ›•,4 c, .!:-_,, ctQ 2 --,-;‘, c--• c-, 7:::, - <„,-„- ...... - o O .4_,,.,. —„, q. 4_, czt - •_,, .7, — > •- ,__, ,i) -,-- •,-.• • - O I--, ÇA ,. :-• ;., i.) o..) -c) •-i: 0 ÇA 7:i ,, "4--' czi "F•••• "F) -"' (L) I.) cA ^ › • .-, CL) Q) 1) CL) CI, "d " n 0.) 40 ›- $-+ ''''= 7..:1 ci' = ,..p. ' t ,-. a.) "0 ,...c1) N Z '',-;' $.-. E 'a' 2 a.) -1 ct . —, .L..., ,= — " v, .;.:. c'' , › u — ct 0 ,... , ,c, &D.,c.,_, E ,... - . Fi 0 t am O ,, ,. ,,- 0 = u 6E'cip2 ô O a.) c.) a) ,..4 s_, .._, •-c)° 2)^ C) 0 cD— 'c'Z' •-"--' 20 ;-, ct ,. d 0 a) ,,, (7\ —. t •--, ci .C%5 ''F'' ''' -6u 12 r' - n''' - ' C, .:j . . :- t, ".', ' t. :-I -. ,E 1..) ct gn ,—," .0 '-' .5 czt ' — ,st ,,, = m „--.,c, "D b.0 i21 6 ci;-,7) 4 .7;; 0 " 5, c" ct "t' ,—. 0 E 8 -6 74 ,-`' ct •1.0 'Tc') ,— 8 1.) • - .=.' m u — 0 — '0c fz,, u z •— 0 4".' J., r, › C-'7i 0 = › Ct N CC! • ". ;-"' Ct ,:.i ' '-' -.-' -2 4 (2) ;-4., ;., •Zi . . 0 „4 (...) ci.) . ci, "' V) ›, ''' cf) ,,

4 P 0 • inI L.:( .), 0 ›.) S:::), I) (1:: , Ct ,0 c...) h••• W '- al_ 2 0 .:' 2,...', , iL - ) .r ) 2 , • - "- 4 2 - 5, . . , . EV c)c) . cil -01) .---. ,..)M •-• o.) '-' .c , 3 ct (1.) 1.) ci.) F ct .7).., PD ' 5 . (1.),, cr; at 4 8 if ',1.• -.5 -,5 cl. 0 ,,, çz,., t).0 •

,.--, ,---, .---, ,,--. ,-, ln kr) Lr) d- -.. --, ,-- n-n N •r-- trl tr) ,_, ,-. ,-... 0 0 kt-, 4 tr) .• - • v- - 'J)) lf") kr, ,----. ,---... ,----, ,----, ,---.. ,----, ,---. tr) N 1--7 - t--- N r--- --n kt") -. •-n -. Nt- I -- r-- r- NN n - ,....., s....., s...., ,..-, ‘,...., s._., - • ' '...... ' s.J ••n•' ,..j ,..../ `...... / ,.... ,....., ,..n ,..n ,...... " S...."

00 000000 0 0 0 0 0 0 • 0 0

'a "a 'a 'a a' 'a a' C-) C...) CU C.) U 0 0 0 0 0 0 o 0 >1.) (i.") tu a) a) (t) >•• > > • 2 2 2u 21) ( (ui:2.a2.P2.r:2.a.a (...)(...) CU C.) (A (A up (A cd Cd cd cd cd Cd cd LID 8 c) c) c) c) c) c) c) .)8 .) .) aaaaaaa.c9

cAcic/Dc/]v)v)c4c/]c4(..)cn(jUcnUUUUUUUL.)

1-5 .4 .4 .5 .4 ae (24" ae ae 1:4 .4

C) cz) c) 'r? v? 0 0 0 c=> 0 0 0 0 0 0 0 —1

cd C-) > - c\I c.c1 c "ci Cq Cl D r- oo 00 cN '7i- 71- "71- 11-) g") • \..0 un vn r- r- C) C) c-) Ir) v-) "r) v") "cl cq1") N ci cl ci ci ci ci r-

• Cq Cl CN ln Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Chl Cl Cq • CN Cr, 00 CN CN r- r- r- r- r- CN CN 01 01 C11 CN CN r- Cs 0.n cs cs CN CN 01 CN CN CN 01 CN CN CN CN CN CN CN 23 23 25 -(2) cg) cf) ci rn rn ci ci ci ci cl ci ci ci ci ci ci ci ci ci 3 23 2.5 "Z 23 23 oo oo co oo oc oo tr) Ir) tr) v-) v-) v-) vn oo 1nI

ln VD ir) Cl Cl Cl rn ci ,..; un vn un ci ci ci ci ci ci ci N NN h r- r- N r"--- r"-- C"-- I"- t"-- t"-- Cs- I"- Ch Ch 01 0 0 0 0 CO Ch 00 01 Ch 01 00 00 00 CO CO CO CO d ,--i ,.--t ,...-( ,.....4 ,.--n .--.1 ,--1,-...1 ,-. ,...( ,--1 ,-1 ,, ,-.1 ,....4 -. --, ,--( ,„.... .„ ...... „ ....„ ,, ,„ 71- ,--1 11") VD \ID 5___.,—,0 ...... ,--, .....,,—I ci 0 0 VD VD \ID

CD 7r ›, oc oc CD Ch ln D r- h In m 0 m r- Ch VD(.0 \CD ci C) 00 00 O O oc r- r- oc VD N un ci rn rn ci ci ci ci ci rn cl ci ci un rn N 4

a.) :-. ;-. I--, 0 s-, >-, .-, ›-, I-. ;-. I-, I--, I--. I. u-, c-9 (.6 a..) a) cu c7i tu :-, ;-, tu 0 CU 0 CL) 0 0 = •,--1 oi E a) 9 a a 2,, 1.) 1.) i..) a) cu a) a) 0 ,--, g g l,..) 8 CL) z 6, et" C cl.) CL.) A.) ,C1.2 C.) 0 A) 0,:d CU o 0. a;' 6 6' 6' i g> g> cr\CN s„---. cn 4 —s s— s-- —s t---- I-, ,,..., • .. .. ./. n r.-- tr:' ,r, kr) ,...... n ,....., f,-7 N ,-. ..", ,..., .-. .--. ,--. r•••-• ,-_, N. ,-, e.--• • ,...... N N • _ (--..: ,__, ,.._, ,---, ,...--„ •--... .--, 4 4 4 4 ,-1- ' kr) tr) tr) kr) . 2, _ • .. _ .- _., 1--- L.-- kr) n ....- — —, — — ,vu (---- -d- CA ,n -. ,—. — —, tr) 1/ tr", v:-; ,) --n n n n n `,. ' ,...... / ....." `,.. / ,...... , `... • ,..../ ••n.., ,..." `...... , ,..n, ...." ,...... , `...r, ••n•' 'n../ `...... " ,..n' \r' •n.,

Q) Q) ÇJD .4 1 Ct "ri "I:) I..) (1.) go, • • T.., c...) ,.., tn 1:4) 4. 4.-, 4-4 PCI PI:t 7:j ,-c) ••-i ..--, • --I —I 0'4 (24 4 ct 00 0 77 77 77 "Ci "Ci 771 77 77 c-7-i M ct a) > > t) Ot • .q *q ,4 >,., ,,,- s ,,,-s .0 .0 (.9.' c'it'-' c',:4-' c9-' PI' c,4•4 c9, -' g j7 C'T >TIC.) a4 (24 P4 (24 (24 124 a4 o4 4 — 0 $., ;.., =._, s- s-s._, ;._, ,_, w• w f-r.. n=, (...) a) ct cc; ,-. a) a) 1.) a.) I) a) a) a) ,,, ,,, cd czs 0 0 0 0 0 4 4 am to to b1) b0 00 to to to <1.) 0 ' '-' ' 71 ,--, -0 -7) 7:3 -0 70 .-7 -0 77 a) a) R a) 1.) a) a) 4 -a-) -a-) O O'', m ct ct al cl cc: cc; cc; ...laa=000lo4 ,71-moia00000Poopaolaoraol000

n n n n cd, nnnn S

Cl 00 Cl A 07 C) CD '71" CD cr) cc) V-) 00 00 CD CD C) C) CD C) 6 6 6 6 C‘i 4 4 kr) 4 4 4 r": 00 00 00 00 00 00

N r- m r- te.-) r- r- oo CD ClCD C)CD NV. )Cc) In tr.)V") i() 71- 71'0 CT CN_ m mV7 m In71. CD V") CD o 0 r- co co oo r- r- r- r- m m r- r„. r- cq NI cq cq NNNNNN

cq cq cq 71- 7t 7r 7r 71- cNI -. 71- 0 r- r- cN r- cr. ch cN CN CN CN CN r- r- oo crn CN CN CN CN CN CN CN CN CN CN CT\ CN CN CN CN CN CN CN CN CN CN CN CN CS\ CN CN

C--n1 27-:- C.-- 4 4. 4 4 4 4 7r CN C) Cl r- r- Le-) oo oo rq cq cl cq cl rq o o c)cz) -.6. cq oo oo lD Cl Cl Cl Cl Cl Cl Cl 00 \,‘D ,

Cl Cl CN Cl r- r- r- r- r- r- r- c ,1 N m 71- 00 CN r- r- r- r- m r- cr) csn c), CN CT CN V N Cl C1 CN CD Cl Cl CN CN CN 00 00 00 00 00 00 00 00 CN 00 CN 00 00 CN CN CN CN ,„ S.S. \ SS.. \ \ \ \ Cl Cl c— 00 00 00 al- Q c---5 oR0000ccNi cq c) cq CD CD 00 -, NN N 0 N Pz Nco ,4

Q) U 77 CN CN CD 00 Cl Cc) CD 4 V-) 7t N O c,4C•1

CID cn c./) tLA 71.1cn01)01)(lla..) c.) ;en , —,.., -0 -0 • (1) gg§g rcz 7:1 • uo P- Con W rT1

C 71- N r- r- t-- s r- r- r- s s r- s s r- r- s s ... NN ., Lr, - ,. If If tr Lr If Ln. tr: ti Ln tr v-:' v--", tr v-.: v--: Le::

lr) ,--. .-. tri Ir:. lri ln In tr) tri ,ri I() ..ei tri ,ri tri tri IF lri tri n n , , ,...., ,...../ , ./ ,...... , ,....., ,..., ...... / ,...., ',.., ,..../ ,...., S...., ,.n/ "...., ,...... / ....n/ \.../ ,../ .....,

7J 77 77 77 77 77 /J 7J 77 77 7J 7J 1J 7J 7J 7J 7J 1J 1J 7J 7z1, 7:J a, a a a a a cl., a a cl., a a cl., a al., cl., ci, lz., cl., cL, fm., cl., ct czt ms mi ct ct ct el m: cd m cz: cci crt ct ct (71 ct ct z cc ct = 04 04 414 04 =4 g4 c:4 C4 0 C4 g4 =4 =4 =4 ;4 =4 =4 =4 g4 =4 =4 CD 7.. „ 7. 7. 7. „ „ „ 7. I. 7. 7. 7. 7. „ I. 7.. 7. 7. „ • 7. .,.. a) a) a) a) a) a) a) a) a) a) a) a.) a) fl.) CL) (I) I) a) cl.) 1.) cL) cl.> 'c») bi) 00 loA bA bl) t1.0 tO bA th bA tO bA b.0 bh b.0 tO b!) b.0 b-0 t),C) b-0 tIA C-) 77 17 17 17 77 77 77 77 77 77 77 7J 77 77 77 77 77 77 77 7J IJ 77 C) c71 c71 czl czt czt cz: ct c7S czt Mt ct ct cd cd c7) cl ct cc3 ct cci cti ct .-. PO PO PO PO PO PO PO PO PO PO PO PO PO PO ao PO OP PO PO PO CO PO

c/) U ct) c4 ci) c/D c.) c4 cA cn c.) c/D 6- .2

CD CD CD 0 CD 0 CD C7 CD 0 C) C) C) 0 CD C) CD CD C7 CD CD CD oc; 0.6 oc; o6 oc; oc; oe) oc; oc) oc; o6 oc) oc; o6 oc) oc) oc; oc) o6 o6 of) o6

• V7 le) VD V7 r- r- oo Cs cl CD cl Tt cl r- oo rn Tt tin rn 752, CA CA C9 Tt tr) VD ,C) le) 00 00 v.--4 VD • r- C7 CD C7 CD CD CD C7 CD CD CD m r- r- c) c.) c) 0 c cl cl ci cq cl cl cs" cl cq cl cl cl (-1 cl c‘.1 Cl CA CA cq („1

fe) (e) 1-1 1-1 Cs Cs Cs Cs cs Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch „, _„ ,„ „„ _„ _„

Cl .7h 7t 71- re.)\-4:-5 Zr' cq cv cq cp Cl CD CD C7 cl CD C7 CD CD CD C7 cl cl C7 8S. 0 0 C1- 4 25 Tt CD C7 CD Tt CD CD 25 oo oo 4 7r cD 0 1-.1 1-.1

oo -. cl cl cl cl cl Ln nLD vD vD Ch Ch Ch Ch Ch Ch Cl rn m (n Tt 71- V") V") 11") V') V") in tr) ln te) kr) V-) V") le) un vD vD \..0 Cs cs Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ,...-( --1 ,...., --1 ,-( -1 --( ,-( ,-, --, ,-4 -.4 ,...-1 ,-.4 ,-..( ,--4 --1 -. -- n -...( --... - , ,.... ,, -..„ -„. -..... ,, -, -.„ -.„ S. -, ..., ,, -.... -, Ir) ,--1 Cl ).--( C '-0 cc) ,..0 Ch 00 Cl 7r r-- r-- ,--, ,---1 -, ,- (-9,-.1 --, Cc) Cl --0 Cl -4 CA C7 Cl --I --( ....„ ,, -... ''S -.._ -- -, -- -- -, -, -, -, -, -- ES-7: r.":: N r- Cs cs 7.(-) 00 Ch Ch V7 00 VD VD AD VD

cc) 00 CA "Oi ..-: Cc) Ch Cl 00 Cl V-) 5 flIf 52, FA --. 06 Cl CO C? C;) OC CC ,6 00 AD 00 Cl c.; cq 36 vD r- I cx) C7 1,r)

1 0.1 '171 Q) -0 0 2 2, 42 -2 2 3 43 43 ›, ›, ›, O Q0 m, A% `,/,' g "F) m`.8. g g g

71-

/1. Cs- l"--. /"". ., ,... ./1. /1 /1 ,-, -. I"-- /1 I-, —, ,--, i,-) —, --, ----. - ,----, .--, S 1---- in v., i--- • - ,----, ,_,• - /1 /1, /1 /1 in ., N ,-. ,-, .71- ^' kn .71- N ^ '1- .7r 't .71- .71- 't '-.- .ri- .' .. '....., ,.. ..-. 1-, N kl'Ç . •• 4 ^ in —, 4 — —, --, --, — in, in ir- in Vs, Vs) nID , ' V'', n n n C1 ‹) ,-- tr', •- `,..-.., •.-.' ,.../ ,...... ' ,.../ ,..., .... ,...... , ...... , ,...., ,..... ,..../ ,.../ ‘..... ,...., ,...... , ,....., ,.. , s---,'

Q)6‘ -73 -73 -7) -7) -zn P, a 01, C2t, O., fal ...-z azt czt czt ct act al ccS ct ct 1.., 77) "Ci 7n "0 7:1 -0 -0 -0 77J 7n p4 p4 o4 p4 p4 p4 p4 p4 p4 p4 o .,---ti cctmczsctmctmcdcztctct„, (Duuuucuui,,Du r:4P4a4r:44 1-r-, ,?,2ac1 u (L) ei) o o ;- r ) 10 01 ci o .= a.)a..)a.)a)a..)t1.)a.) czs bP bP bp b0 bP bP bP aaaa aaaaa c..) 771 -d -c) "c3 "Ci "Ci "C)li"C) "Ci "Ci "Ci CI CI CI CI CI CI CI CI CI CI Q) MCI MCIC CI ct cicd m:00000000 op o 0 a ,_ CC) PC) PC) PC1 PQ PC) PC1 PL".) PC) c/D DO CA U0 DO 00 00 C1' CID 00 . ai c/) cr) vp L.) c/ cr) v") U (.) c/) c4 )21 4r.) c..) •". „, 0.) cdo .!

o ,... c, a.) 0.) .Q) cD cD cc cD cp cp cp cD c? c"! cl cl. cl. cr.). . . •4 oc; o6 o6 0,6 od oc; oc; oc; oc; o6 o6 o6 al., .c: ct ,... bi) o ;.., '-c5 Q) Q) ,4 ..0 ra, E P12 g g g 2?' S n g!,2 M M 7t 71- 71- ..= 0

1-1 11›, --zt -) 1-1 1-1 1-1 0') CN CD CD CD rn rn rn CN 00 CN 00 CN CN GN CN CN CN CN CN CN 00 CN CN CN CN CN CN CN CN (L.) CN GN CN CN CN CN CN GN CN CN GN CN CN CN CN GN CN CN CN CN CN GN 0 -ctzi Q) Q) rn rn rn rn vn r-- Cl IND 00 00 00 Cl Cl Cl C) R CD CD Cl Cl 7 _1 s11:1) '`' ct rc .j' rnC)S SS SSSS 5555 5555 55 C) Cl E - 5 Cl un 00 kin o0 o0 o0 CN 0 CR. CD oo cl 71- —1 —1 cl ' $2, • › -0 ..(2. cA .- v) (---- Cr) Cr") Cr) Cr) 'cis V) V') Cs-- ts--- -4NNcnce) cn crn N --, .--, n ON s N N N s N N d- ,—( N N cl N N 0 —1 cri':a (.4— '' n n n s•-n ,—' 0 0 Cn On Crn a CP \ CP \ CN CN CN Os CN CX) o0 CT CN CS\ Cr) Ch c7 \ 0‘) 0-0 ,--1 ,—( 1....1 1-1 1-1 1-1 e•n•4 ... n••nI ••n••I ,•-I 1-1 /-I 1-1 n•••nI 1-1 a.) .,,-1., \ \ \ \ \ \ \ \ 00 an a!)S2 \ \an CC) CT \ kr) kr) 00 Vs . ,'—,' 3-1 1•n•1 0 o _ Ea- Cl '. '-' 0 f=14 \ 000000 ,.= u ô , n P., ct N- I CNN—':- \ Q) .9. = —fEL (.1—, a..) 1.., o s—' o ..., (1) N 00 nCD C C rn r- rn rn 71- ,4 CN CN I c.) 1n1 oo • cr) In d- 71- oc E Cl v.) r- c) c) cl oo r c? m kc) CD CD oô r'4 N00- 4.; '-c9 0000 4 eri VD V') If) V') 3 IC) ,,, 7:1 7:3 7, • 4 = a) a) a) a) 4= 0 a) a) w ;), •52 ..2 cA "0 a) -15 a. z 0 sn., a) ci.) a-)z a) ,.:c vs .E a)0-) a) 0 0 0 ci 0 o cl OP CZ hr, hr! hr! nn !\el "!n ▪ • • ▪ ▪ ▪ ▪ ▪ • ▪ ▪ ▪ ▪

71- ,---, ,---, - - ,---, ,---, „--, ,---, _ ,---. ----, , h h kt-, 71- • ,-----, I"- r-- ,-...--. ,--, - m 11", ,-. ,- . ,-. .--, ,---. 1---' - tn d -d: -, c.-, „ tt-, s d' cn d- - r... ,..._ _ ,.., • - Lr, kn ,- -, -, - --.,N r-.- kr).' , . teç kr, r) in v-, tr, - (--,1 N N LI"; ir).' Ci 4 1.,-,, I.r": ,...... " ,...... " •...... , s..-/ ,...-.." ....-." __ ,..-, ,...... , ,...... , ,..- ...... , n._./ ‘..._.., ,..-., __ -4 -c) 7:) "a • ,-1 77) '>1. 6- , a4 Cd :L, 1.) cd ,, cdf .CL' '-* C/) b.0 ,Q ,5) 1..r.., fz4 cd cd :4 = EEE> 0;a m -. .4 ,.4 ,4 .4 .4 -S--)4 P4 (-5, "5 '"c4 o o o ›- (I cl) ,,, (,) ,4 ,D - -, (:) (i) CI Ct CI 0 C) › ›. › ,--i 774 -.." CCi Ct Ct CC; CC: C C : ›..,.,,_, tO '''. ›. ›. >., ›., c t cz I o 8 t at C.;) 9 .,-E1 -'Jd-' 1.) o o o o a) —1 (..) ,.9-,, z = = U m rn = s.-1 ;-• - (-)7g 7g "g = o 73, g t L'il o5 2 E I -45 4 g -a' -FI oô 00 00 oô oô oô 75 0 cd :a :a ...a c) o o o ;,. o 2 uo oo --, --, -, -, ,-, •-• P:1 Z (...) oo oo cd() uo uo 24 (24 a4 z PC1 PI-,

GO DO L) C„) L) 00 00 CA L) U0 Do Do Do Do 00 C4 (.)

CD CCCCC cCO 00000000 7r 7r od od oeS o6 06 o6 o6 c5 00 00000.6 -4 cNi cl cl -4 n6 cc N ci ci cl ci co ro ro rn rn on on 7t

Cl CD Cq 71- Cl --' 000 n n N ..CD CD C) CD nID nC, Cl N r.„ rnr_ rrn_ CDro ro rC,„. \CDr_ r0_ __ 00LcS vo tr) CD CD C) r- o r- co 71- D r- r- \.4D ,JD „ID 00 Cl Cl M Cl Cl rn on ci ci cq cl cq

rn on cn on 71- cn 7t cn -4 CD 00 Cl Cl Cl 71- 7t 71- CD GN CT CN CT Ch CN CT CN CN CN CN nC) r- r- es. cN r- r- oo CN CN CN CN CN CN CN CN CN CN CN CT CN CN CN CN CT CN CN CN CN CT CT CN CN 1-1 1.-.^.1 1-1 1-".1 1^.-1 1-1 1-4 1-1 1—.1 1-1 1-1 1..^1 1-1 1-1 n, n, n„ n, cl cq vo• cq vo vo cq rn o oo ro ro cn oo r- 7t Cl Cl Cl NN Cl Cl Cl Cl CD Cl CNNCl N Cl Cl Cl NN Cl Cl Cl Cl Cl N N Cl cq 00 CD n.0 00 ci Cl n.0 M 00 Cq Cl --I

00 00 r- r- ro r-o cf-) cn on M 1.r.) CD 00 CD 00 00 r- ro -, r- .71- CD N N N N C\CDCDCD CDCDCDCCD00 CDCDCD C7 CT CT

CD CT • 00 00 c- M h- 71- 00 00 CD C C C) C NC CD Cl 00 00 r- oo 00 000000

• vo cq cq cn ,4 CD r- cn r- cq vo • c‘i n47) CD CD O 4) - Cl d" r- r- cl ro ro 0 o r- o CD n1D 7h (-) • MN (24 A

Q) Q Z 75 CCS 00 (1) ,..0 CA g g g g -cl al. =-, •-a, E E E E a ) C ,SD CL) 771 i-i 0 CL) CL) (1..) g ( 1 ) '-'0 7-7 o o o o o on ô i a.),, fc2 ri?) c-n ,-,. -.. n. r ) !../1. 77 (n rr. rT . r/i. r-1-1-', 4-1 n••••", C-- C-- C-- r•-• s s . - vp r--: " - - Ln v-, - nr) 4 - - v-, kr, ..71- ..11' ....., ...... _.... -...... `,..'n ,...... , S...... , li EL, cd Ç24 cf) (J) (ip cn cn (f) c/) c/) ,, bo a) a.) cu a) cu a) au a) ..- ...... 4 ....4 ... b.0 = > > > > > > > > a) a.) a.) a) a) cu 1.) • ..-.0000o0o01)(1)(Da) a) a) a) = ,.) c_) ...) ,L,..) c.) .1"..) c.) ) ;-.. 2 1O-, -7) ..E. Û Û Û .-') Û Û a..) cu a) ,.0 M 1=17i 7:) cn a a a a fD, L., (1.) 0 C.) cl.) ell ci.) cl) a..) cd cd cd cd cd cd n Z LID nE ' --, .1-t, .-t.-- .t-M c:1 i4 f:), SD, tD, SD, SZ. ID, ta4 SZI, cl) 1) cl") cl) c.") (t) cl) n n (3..) '7- • . r.: . i.--- 'i.--1 g g g g -6,--,) .c,L.) o o o = c/D=2 c/)° °c/) ° ..., 7_,) F-1 E.-.4 F-( F- , F-1 E-1 F.4 F-1 ...... _. ... = z - ,.. • - c...) ci.) ai a.) u, .5 2 g g g g Li F,, g g g I' g~ (9-' g~ A.A.PAPAPAPIAAPPAPAPCIZZZZL7OZZZOL)L ) L)

cd0 GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO (..) DO DO GO GO

4•-j' :eaef:eae C4' (:e

r- VD VD nCD VD VD CV ce) ce) rn VD cf') n CN cc) cr) cc) cri Cr; CT CT CT CT CT CT C N C•1 N C‘i N N C•1 C,1 C\i 7h 7t 7t n n n n n n 11-) n Z) \.0

C NNN C 00 00 00 00 00 00 Cq 00 00 00 CD CD Cl 71- r"-- Cq Cq Cc) N oc oc oc n n v:) c) s:) r;) C•1 v:) CV Cl`r) Cl --I Cl t-- r-

cc) Cl 00 Cl Cl rn --I --I -.. cq cq ce) 71- Cq Cc) Cl Cr.) 00 00 VD 00 CN Ch Ch CN CN C".-- 00 00 C".. CN CN CN 00 CN r- CN r- GN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN

71- n Cc) in Cl 71" Cq C.-1 CD CD Cl Cl Cl CD CD CD CD CD Cl CD CD CD CD CD nCl, CD 2.5 CD CTVD 00 CD Cl 00 .C.7I 00 CO VD CD 0 cn CD 00 Cq 00 CN n CN 1-1 nn•n•1 1-1 1-1 1-1

cN ca) le) oc cq r- d- Cl Cl n N 00 cc) VD VD CN CN CN CN CN CN CN CN CN 1-1 S„ n, \ \ \ \ \ \ In Cl Cl cc) 00 00 Zr* cc) cl-) - CN1 0 0 (5- 00 00 nO N 2:5

00 25 76 C[: 00 4

O _,73

77') 7r1 .`)

▪ ▪

d- ,---. r-- ...---•. -, r- - ,,---. -, _r- ,---, r N tr) r- ,--, ,—, trç ,410.. ,--. ,—... \o., ,auo.. Ln' - - .._., ,....-, ‘..._.-•r : (--‘1..—, if.._..,) ti,..----; ..—,tr) \....."tr) \.....,,r)

z o m 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (...) Q) C) I-) (1-) 77) 7i 'Ci 77) 7=3 7i 771 771 7:1 "Ci 7:1 "ri 7D "ri "Ci "ri Ct (71 (CS Ct M Ct C71 (71 CI cd CZI CCI (CS ;-n •aaa 00-5-5-5 0 0 00000000 000000000000000 r:-.-j r z7I, 0c...) c4cAcnUUUUL)C_)C.)(JUUL)C_)U(...)U`-' a.) cl cd cl .174 .., .174 .174 .174 . 7.7:1 .?4 174 t'4 C) C) C) C.) n 2nnnnnnnnnnnnnnn n

• . cn . cin. . 7t. 71-. d-. 7h. 7h. d-. 7t . 71-.. d-. 7h. d-. d-. d-. 7t. . v,D. 00. VD. cn. cA N cq ri ce, nCD vD VD n.0 vD vD VD VD vD VD VD VD VD VD VD VD VD n.fD VD

O cn cn cn cn rn rn cn cn V Cq cr) 7t rn Cr) oo oo oc oc oo oo oo co oo oc oc oc oc oo oc cq cq cl In cl n n n n N r- r- CD D vD rD VD VD VD VD ID VD ID NCD clCD CDcl clcq clCD

cl cin o0 cl cl rn 7h oo cl cl rn 7h 00 cl cl cl rn CN r- Ch vD r- oo oo oo r- oo oo oo vD r- r- oc oc CN CN CN CN O'N CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN ,„ Cq 71- 00 CS CD r- oo CN CD r- oc Ch Cr) CD neD Cq CD N CD cl CD cl CD cl cl cl CD CD CD CD CD • CN 'Cr3 23 c) 00 C:11 VD CD CD oo Ch vD oo CD CD cc) oc oc rn oo „,

N cl N r- CN CN Ce) CN CN 00 00 CN CN CN CN CN ,,-1 ,,,--1 -,-,,..-q,...1 ....„ -1 ,-( -4 Ir) Ir) c"- * VD• \.0 C•11 Cq CD CD 23 CD CD oo oo 00 00 —I 00 00

t.> NCr) VD In Ch oc N CN1 00 00 CC N NN NN

C.) cn v) am 1. i. SD4 (1.) (1.) --I ,.,---, g ., ., <1.) 1., r- ) cl-) 1-)

e....., ln "—, N -- _ _ _n—, r-- v-) ,-, _r-- ,--. —N , r- ,--t--——-447,--7-_—. v-) .-1-• ,--,— -1--- . ..---, _ ., _ _ _ Q0 N410. (--. ”•-n ^ te) __, , 7t- ir, 7i- • - - • • _. ,.---. kr, • ,.._,, ,....., ,r),...-,-,------,...... ---...... -,..------,-.--,,,----,,,,--,, -- --tr,v)inir,v-, ,JD,r),-, v-,

-ci -cl 77) 2 -ci -ci -ci -ci -ci -ci -ci UU CL, CL, CL, CL, CL, CL, CL, CL, CL, CL, CL, CL, CI, CL, u cd cd cd cd cd cct CC: cd cd cd cd c) o4 o4 1:4 o4 a4 o4 c4 o4 1:4 o4 a4 a4 o4 a4 (-) (..) 2 = = = = 1:1 CdD 00 o o o o o o o o o o o o o gg gggggggggggg c..K.Juuuuuuuuuuuutuu.5 o mmmmmmmmmmmmmmu Tc) gcl .4 .4 .4 .4 .4 .4 .4 .4 .4 E--( C)

L) GO 00 (.) (...) C.) c.) c) DO 00 (.) 00 (..) DO L) C.) 00 U0 00

on cl Cl un Un un vn tin On un un un VD VD VD CD r- 00 un C c; kr; Ir; kr; kr; kr; Ir; Ir; kr; kr; kr; kr; 06 N N (-ri tri z; \ci n1D VD VD VD VD VD VD VD VD D nSD nSD C- C- r- r- r- r-

un on on CD CD CD CD CD VD 00 --I m 1/0 cV g g on C- CV Cl on on ce 00 00 oo oo on cl un on on «,d- r- cl tr) sel CD 7h 71- CD CD CD CD CD r- CD M m c Cl Cl Cl cl Cl CV Cl Cl („1

71- 71- 000NNMOMOM-1 -1 0000 Chr-CT\--00o0C7Cn CNCY CNCNCNCT(TCNCNCNCCNCNCNCNC ...... -NN O- NNO NNN ONNON ,-INCANCONCVCCOCCDNCINN

e.."1 1--I

CA Cl Cl CN r- CD --( M m cc) --I --; CSN m m m N r- 7r- tr) on CN 7t r- r- CN CN CN CD Cl Cq Cl ce 00 CN CN CN CN CN CN CN CN CN CN CN CS\ CN CN -, -1 ,, -, -, -0 ,, „, --, -, -, -, -, -, -, -0 -, -, -.... „ „ -- -- un un Un 5:::: CN 7h c"- g.."- --( VD VD VD CV Cl -0cq c5 -, S S CA 00 oa C.) ""'-, .„ '-'- -.... "' 00 oo r- VD ^' r- CN kk 00 o0 00

M 00 -0 pN C cNi 4: cd pp VDkfn 7r • vn cl on CD CN CN r- r- r- c, m '7t m oo oo - 71- ,4 Cl Cq cl 00 r- ,, cl c CD tr) m tr) 71-, CD NN eci 00 o0 00 1--.) VD VD VD 'Ln 140 140 ,-, - 6 :ai v)sk 1..,V) = -0 U0 U u 0 —›,1=1., '0771 70 — —_ ---. 7), '-,-,1 8 8 bo- er.4 ,-ri ›,"u4),»uuurc,„9

▪▪

,-. r- ,--, --, --..v,-. ,-. ,-, ,-. ,-, ...-. ,-. ,-, kr, . I's-. r- /-, ,-. ,, -, NNr- ,-...., r- ., rn . . ,-, 7i' 4 4 4 4 4 — Lr-',' v-" ,-, , --:, —,_, — ,_, Ln ,_, i1 •-/kr) •-/--, ,--.,_, kr),_, kr,,_, ,_,(r) 1r),_, ,_,If ,_,kr) ,_,V-, ,_,(-.1 ,_,kr, tr,

71 7j 77j •14 • • CL) (L) CL) CL) (1..) (L) CD Q CL) CI) C • 0-1, FL, a) a) cu a) a.) a) a..,), c) c) L) L) ç) -6 .3), (1.) (1) Q) cll (1) 1.) a) a.) a) c,7): 'sa., FL, 'FL fa, ( ),,f) a a a " b, " g 0 p4 p4 o4 o4 a4 ddddd a) 0000 O "7) 7=1 7:3 — 0 to to bo to bo to tdD -- to -- g g g g g ,%> (S) ' 171 40 xxxxZ GO 00 DO 00 pa pa P:1 Pc1 Pc) 121:1 1:121 a C)

Do Do L) GO 00 00 00 00 00 00 00 (..) C) Q) C) C) C) 00 00 1C

tr) tf) 00 00 Z) Tt Tt Tt Tt Tt Tt Tt Tt cc

Z). Z:,. Z:). Z). Z). 06 06 06 06 C.:- N N N N N N N N N N N 06 N r- r- r- r- r- r- r- r- oo cc oo oo oo oo oc oo oo oo oo oo oo

• c) CN CD CN V) tr) 00 7r 7r 7r f:-) \c, Cq tr) Tt cc) cc) cc) CN CN CN c c CS\ c N C Tt Tt tin (4-) 14") 1,1") CD CD) CD r- r- r, r- r- -" -" v") oc cx) cq cq cq cq r--- C•1

cq O N C CD 00 N N 71. on 7t on N N on 7r 7r 7t Tt cc) n47) CN CN CN CN CN r- oo r- 00 00 00 CN 00 CN CN r- CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN

(1) CN r- r- vn vn -. on cN rn cq cq ch rn cr., on r- CD cq „„N —4 CD C) oc cr, 2R on C.) (--7-1 on on Q3 00 23 23 'CZ (-;:a C.). -

on on r- on on cq r- on N N Cq CA cc) (NI rq C-- on Z, tf) If) CT Ch CN CN CN 00 CN CN CN ,-..n 1.1 r...I 1-1 1-1 1-1 /—I 1•••••1 1-1 ...... , ...... „ ...., ...... „ ...... , . 2..- ` "....."-C 1--1 /C 00 ,.., ,--( ,--1 0 --4 ,--1 oq -, ,, 0 ,,—1 N 00 00 --I 00 0000oO ,. nID

tin CT on on 7i CN cl) Tt -0 F, on Ch 00 7h 71- un 0 4

O czt 1.)E E çz. (1) o o

...... , .."-", ,..--, ..-`, N r- N .v-, ..-. — I-- —, -- ,— ,-, 71- --. 71:' 71: __, • .--. ,__, 7t" 7t 7f tr, tr, ,t ' • • ,t y-, . . ..c., -- --' —.. 71 71:. :7T- ' .7 n t--- — s. (--. —. v2; v----, in -- -- — ...,, -- In.' .1 r: Vi ,l'i tr) ,...... , ,...../ \n, ,..., ,...../ '....n, ,..../ ,..../ ,..../ ,...n, ,.n/ "...... , ,..n, ..n,

ct cd cd cd cd cd a') ;-m P4 4 4 o4 p4 c4 p4 a4 a4 o4 C) C)

1 1 1! 1! --(c! "! c.:56'6'zxzmoo u zuç)ç)(...)c.)u

UC...)(jc4c4c/Dc4c4c4c/DUc4cnc/DUUL)UUUC..)v)<4 <4 <4

• rr) tr) kr) y:D nc) vD n47) D CD CD CD CD rin on con cn cn cin °C. CD CD cc; cc; cc; rri rri rri cci rri cri tri tri tri tir; D6 o6 oe5 o6 oc; o6 "n CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CD

cl oo nc, cv cq r- r- r- vn CD Cl 00 en Z) r- r- \CD 00 r- CD CD en 71- un vn 7i- rr) 71- ,_, Ce) 0-) Tt \CD \,1D nq CD CD un rn 7t 7t cA cl \.0 \CD r- 00 000000 Z) 00N r- rn *Tt cl cl rn rn cl cl --I --I --I cl cl cl cl cl cl cl cl Cl CA CA cl

rn rn cl 0000000 rn rn rn Tf "1- Tt re) TI" 0-) re) 77 CD lD CN CN CN CN lD CN \ID oo CN CN CN CN CN CN CN CN CN CN CN CN oo CN CN CN CN CN CN CN CN CN C7N CN CN CN CN CN CN CN CN CN CN CN CN

NN 23 "s5- - 25 Z) el:: 4 4- 4- F-7.2 P--2 25 'c.?) ,115 cl cl c: rn cA cl cl cl CD CD CD cl CD cl cl CD 0 Cn1 -Fn1 ;6 2iS rn cq cq rn rn rn cl rn c•I NOON 7t 25 rn-1

N CD 00 cl cl cl cl 00 CN CN CN 00 r-- r- r- 00 \ID \O rn LrcN71- 71- N \O rn CD CD CD rn \ID \.0 4-) \ID \.0 CN CN CN oo oo CN CN CN CN CN CN CN CN CN CN CN CN -. rn rn CN \.0 CN CN rn rn \.0 \.0 --I Cl CA Cl CD CD Cl Cl cl CD CD CD Cl CD CD Cl rn N r- cA oo oo r- r- N ezi vn un vn

\ CD 01 rn • cq r- r- rn 71- oo ,4 ,4 ,4 rn \CD r-- rn CD r- N NNOrncncnc0 N \CD \CD 71- cici cl•• c.) gZ• 00 00 cl cl 44 CD CD CD .44: rn cn 0 rn c-A c.ci P4 e‘i

•n••••I 1.•n•n1 bA 0.0 C/] rd 0 Ti ›•, 1-1 .4 at5 73 ..4 '1<:11 7:2.11) C4,) ce, ›- G.) 4 = a.) I-) -.(;), 0 0 0-2, •a"3 r=4 =UUUU W ocl

•

OC 71-

,--", / -n ,_ r- __, -a -. NN .---, ,----. e---, •-n ,-, in in 71- 4 „---, 4 ir, — — — _ .,r--- ,.._ ----.. ,..-.",,..., ,r) .4 4 . . . 4 r-- _ ., ----, ---...--. ., . ,- — N N N N N N —.. —.. --:. r--.. t--.. r-.. t'-' ‘,-;• '''''. N tr) N 4. •-, N - Ln n kr , -- r- r- r- -, -, -. --, ,e) aa -, sa-, n .._..., - .._., ,...." •-_-, __.., s...., ,...., .._., ._.., s..._-, ,a_a, ,.._-, ....., ,...., ..._., ..._..., ,...... ,

E E 0 ct cu a.) c7s 1.) a.) a.) a.) -7) 7) -5, -6 g ct2, (.131 (.63 Ç,;:j L) aa aa p4ct a a g 2T:1,— a - fa:au cl °o . ° E E E E EE,D EE — cd cd cd m cd cd cd cd na cd c-) 'E ic4c4a4 ° cn a 0 0 v) c/B .rô `,/, C.) Cl) z C,) C/) C/) CID . . cd c Q CCS Cd Cd V) :d :a :a ..a Cip ct ct cd cd ,o d P4 DO PQ PQ PQ PQ QC GO GO GO DO PQ PQ PQ PQ PQ OQ oo PQ PQ

c.) 00 Do oo oo u oo (..) u u 00 oo oo Q) oo C) oo 00 Cd° '1=D n=.D 1=-D

• c) r- r- oo oo CD C) CD CD CD CD rn rn rn ,t ,t cd oc; cd cd cd cd cd cd cd cd cd cd cd cd cd CD CD

cl N 71- 00 it-m; g °° °° °° °° n.1D Cl n4D nD cl Cl r- oo r- r- r- r- r- oo OC ClCD CN Cl Cl ClCN ClCN Cl CNN CA Cl" 1 Cl Cl Cl Cl Cl Cl Cl Cl Cl

Cl Cl 7t 71- 71- d- d- Cl cl n.(D rn d- d' d- d- d- d- d" 7t d- d- CN Ch ON Ch Ch Ch Ch Ch r-- r- r- oo crn Ch Ch Ch Ch Ch Ch Ch Ch Ch CN CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch „„ —1 —4 —4 —4 Ch Ch r- r- r- r- oo r- c) rn ri ion r- r- r- NN r- r- r- oo oo • c)c) CD CD CD rn cl cl cl CD CD CD

d- d- CD un CD CD CD --I N CC r-- 00 I-- 01D CC Cl • \4D CD r- CD CD un CD CD ln CCNten C-- co OC CD CD nID CN CN CN Ch CN Ch CN Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch ch Ch aaa -. -. —4 -. —4 -. „, 4-, -. -, -- ,, -, NN vn--- d ln i i in c,.-- . .,c--. (--1 cc Cl ,., ,..) ...,. o cf) c.) o ....., C.,) C.-) -.7, r- - (:-.±.5 c.) c...) n•.0 II In \ .0 \ C)

.D.c.'111 ,1 ,-, un ,t cl „. 06 . c) --,. cl rn 71- „. . --i -4 C d 7h'4D OCcp ..-r Cri —4 71- __, tin Cri Cri nID • 44 (--- a, VD " PQ 11.) (-- PQ PP PQ tin N N = --4 < cf:2 4.)1 x ° CI) Q ,,, c. c- •," E °;) g E 4;. ,L1.,) c., 124 a4 g '.,.9 -42

▪ ▪

/..... h- ,-.• ....., n ..". e."'", "-----, . .--1 h- h h ....•n Lt) ,-----. N h ,—. ,--. ,--. h ,_- e". - • ,-----. N ,—, ,—, ---, N .4 _q.--', ..-, Lf-, N — N h

. h r-- - h ',,t _. 71- 't d" 1 - '-1-. •71-- v--, -. v-, tr-- ,d- 7. ,-----. (---- . . . . -- — — 40 h n n -- h h h l--- h LI"; t---: kr-7 tr: ir--7 i:' ..,-; ,e7 (-4 •- tr n ,ei v--, v.--, ,45 ...... , ...... _. — ,-...... , ...... , ..... s...., ,-...., ...... , ...., ,---, ...... , ..... — -----, ..-, .... ,-......

LID 5(ID 7= 7i 412, :1-2 A1. ) 77 77 7D 7D E ni ni c,1, CcdL, al . C L) ° z ct cz: F.;, ce.) or) O4 a4 p4 124 ke 'Ep 8 L) 000000001)o-0 -00---uu 0 W E 5 5 5 5 5 5 z z <1.) 4,) (1.) LID `ct :5 :5 .:5 :5 :5 :a cu IN O • oo 00 GO GO UP U0ddOO UP CD 1=1:1 1=1:1 I

• U'D C1' C4 C.) 00 Cd0 DO DO GO GO GO L) Q) UP 00 (.) (.) CA CA (...) < r:4 fx" f:4' .4‘ rx" (x' a4

• 1") kr) r- r- 00 CA CA cn C) 00 CD cn 00 00 CN CN ,C) 06 oeS oc; oci 06 o6 oc; od N vr3 N tri tri CD CD CD CD CD CD CD CD N Cl Cl Cl Cl Cl Cl Cl • m-1 1-1 1-1 1-1 1-1 1-1 ,n1 1—nI 1-1 1-1 1-1 1n1

• le) nID r- CN CD .1) cn 71" C) Cl 0) If') r-- 00 CN C >< (X) Cl Cl Cl Cl CD Cl C) 0-1 CD 1r) cn cn nC) 00000000 CN 7t cn r- 00 00 00 00 00 CN 00 CD Cl 0 Cl N cc) 7t 7t cra, 00 CA cl cq CA CA CA CA Cl cn Cl cn CA CA CA CA --I CA N cl

71- 7r d" 7t 771" ''t" 71- 0-1 Cr) Cc) Cc) CD CD CD CDce) ..7t 771- CN Ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CSN —1 n, n, 4 n.„ n,

OC 00 00 IDO r•••• CD CD CD CD CD CD CD CD CD CD CD ,„ CD CD CD CD CD CD CD cn cn cn cn cn cn en cn cl CA CA CA cc) cc) cn

00 CD CD ,--1 CD CD CD CD Cl ce) 00 cl cc) cl -) cc) CD CD CD CD Cl CD 1---- CD CD CD CD CD CD CD 7f' 0-- Cr) (--- r- r- r- cN c., cy. ce. CN CN CN CN CN CN CS\ CN CN CN CN CN CN 00 CN CN CN CN CN CN CN CN CN n --1 --( ,--1 ,--...... 1 ,...... 4 ,--1 ,--4 ,--1 ,- —1 ,--1 ,-. ,--1 ,-. .--4-, ,--( „ ---( -, --c ,-- ,- -__ ...„ ,, -..„ -..„ 4 cd d d d d ,„i 71- 00 vT 71- 71" ",7:1- le) ,--, r- • (..) V L) L) U c..) NN cD-.„ c) c c) C C c) i--z r- c),. c., cv es- .c. 1 C-7-1 el cl

.ID CD cc) ,--i • 00 • ,--t ,--( 7) (-NI 75 'd CN S "6 kg P (74_ Cl 76 76 7:i CNC) 7fi( CC) ,--n CD CD CC 4 ,4 0 0 4 4 2 d d d ,I - , f - ,i- — . -7 - , S ,`)° - ,1 Cfi 71- 7t d- — I ,..1 4 N -cC = = = (1) (1) CL) tll u., ci) ci) 7:9 cA ck C/] 0 77) ,,,1. ni •-t7; -0 a) CD 1) II) C.) z 5EF CL) '4 CU " ' C ' ° CU CU r, ct c,s— , ,s5E:7:11)'5:-1.)1) ,L)_---- a, , ' 4 =

▪ •

71-

,---. ,.-.. ./.... Vi 1.-- C-- 1.--- ,--n ..---, ,. .. ,--. ,--. "71* n ..---, h 4 tr v-, 'd- v-, _., N irl,.._, ' , ,...—,r7 tri...... , tr,,.._, QD•-_, ,.....,t--- Lr __o CA in

a a a a O c) .0,0, 77) .2ctcztcdczsctctm . a.,' 'c j a4 c4 P4 Ç24 a4 c)(..)c) () (..) m ct c., c+. c., c., c., ,., c., •14 ,-- ,-- ,-- o o 0 ,2,-(5ç2_2tt--,(tttt't c,,L3 g.),. . .. 000000c,. • ,...... q..) . . . $. ij P, li 7J 7J 7J "O "cf '0 (.) C) (...) (..) a) c, a a a a a a a Q 1 y c.) y c.) ;_ c.) a.) a-) I-) I-) ai () "CDa-) cl) (1) 1.-) 1-) , = = ci) ci) c/) C/D (1) "C; ,/ 4' 4 4 4 4 4, = 44, apv-CD fa, a) 4 Z Z = Z CD O 0 C.) CD CD CD CD on PC) C) 2 2 2 2 2 12 2 on on on on

Do 00 (..) L) uo Do Do (..) un Gn GO GO un on con con on C.) C.) C) 2 2 2,2 a) a) 2 121 2 n-1^ C24 C:4 f=4 P4 P4 f:4 (24 tz4" 1=4

\CD VD \.C) VD VD 0 11-1 CT VD vo vo r- oo oo oo oo N -, -4 -4 -, -4 -4 -4 -4 CA Cl Cl Cl Cq Cl Cl Cl CA Cc) on on on on on on on on on on on on

on on un un vn un 7t un n1c, r- r- r- r- 00 00 un In CN CN Cc) Cc) Cl Cq Cc) Cc) Cc) Cc) Cq CD CN CN CN CN CN CN CN CN CD CD 71- 7t cc) cc) cc) on VD CD tr) 71- 't• ,f ,t •71- '7r '1- CA CA CA CA CA Cl CA on cNI --, -.

on 71- on 71- 71- 71- on on —0 00 71- CD Cc) 00 CD 00 CD 00 CD CN CN CN 01 CN CN 0\ CT 0\ CN CN VD CN CN VD CN VD 01 VD CN Ch 01 01 0\ CN CN CN CN CN CN CT CN 01 CT CN CN CN CN CN 01 01 CT

cs.n r- cc) on oo on c) oc c) r- 0 oo o c) cl o o cl Cl CD CD Cl CD CA CD Cl CD • cn cc) cn 25 m 23 C75 ;','") 23 cl on on cNI on CN Cl CT Cl CN Cl

1-. 0 Cc) Cc) (-- ce cc) r- CA CA CA CA On CT (--- r- N cc) r.- N r--- r- r- r.- CN CN CN CT CN CT CT CT 00 00 00 00 ,—i ,—n /1 ....I n I 14 1nI 1.1 Il /1 ,...1 1.nI ...... '.n.. ''...... , CN1 E.'"'•C'-' 7t cs- AD VD VD ,-i 0 r--::: N -t---... C C 0 0 -S- -, ....„ oc cl r- a \ O\

gD cn un 7r cc,) VD cc) V) Cc) Cl Cc) Cl CD r- 00 CN CN 71- ,--, N 00 oo 00 00 —,

0 "o0 I Q) 8 (CL)1-) -o,..c) •=(L) 1

▪ ▪ ▪

r- ,....., nnr- r- _, ...-. Cd r- ,-. ,-, ,-. . ,-. ,t ..-, If r- r-r- - e•—n v -, 7t u-, v--, n . --.... C, 1 . . . . . 13' I1 tri -) If cr, v..) ,1D .4---, .4--, — ,.....- - s....." ,...... , ,....., ,--...,

cu C.) C/) V) cl) cl) V) c/) cl)

8 '8 8 8 8 8 8 8 8 8 0 0 0 0 0 " I- " "9"°),..u2 2uuuuuuuuuu uuuu ..c) 40 a:, ,f2, " „ CCi C7; RI OC Ct C7; 0 0 0 0 0 a.) 0 0 0 0 0 0 0 0 0 CL) 0 0 0 (L) cl 4 Ct4 CCS4 Ct4 4 CCI4 CCS4 4 4 4

L) C.) 00 co0 Cd0 GO C.) C.) C.) (...) L) C) oo 00 00 00 00 Cd0 DO on C.) D

CN CN 00 CD CD 00 rn cen rn rn rn vn k4D n1D

cn 71: 4 1:). \C.) n12; Cr; Cr; Cr; Cr) Cr; Cr; Cr; Cr) Cr) Cr) on on rn on rn on rn rn on rn rn rn 7t 7t 7t 7t 7t

n .c) 00 Ln 1,r) r- Al) CT kr) kr) vn C7N °° \JO cfn cr) CT CT 0 cr/ c'r) cc) C:.) Cl CT 0, 0, 0, cen cc) CT cT CCN cc r-- (7' (" (7' C7N 0 C.) 00 r-- c) o o c) ch cA cq Cq Z) Cl rn Cl ('') Cl Cq Cl cl <)

• on rn cl cl 00 cl rn rn —. con 71- 00 cl CD 4 —. rn CN CN CN oo CN CN %.0 N 00 CN CN CN CN CN C r— C O CN CN Cr\ C- CN CN CN 00 CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN

cn cn 4 00 CN vn Cl 00 4 cc) CD kin cl Cq vn CD 7t N 0 Cl Cl Cl Cl C) Cl CD CD Cl CD Cl --, --I CD 0 CD 00 oo cn CD Cl Cl CN r— CD cin 00 cn rn ch o- rn vn oc CD on r- 1-1

r- cq cq -, c.,' cl cl r- r- r- r- r- cs, cN oo oo --, --,oo00 --, --, --, --, oo-...... - 'Z' 23 23 ô" ô „--_-_,- cl rn1 e—I Tn1 1-1 r- C7.; ...... S,,,

vn -" gi El; Crlc \ 00 ,6 cr) \ C 4

SZ1 0 5'. = ) UP-ZXIZ

-

,---. 1--- --

,,---. ,---. ,---, ..---. ,---, ,---. ti-, v-, kr, tr, kr) kr, rn - c,") ..- •- • -• •-n ON -, "'"" - - - "--", ,--", -- -, - - -, ,- (---- y-, kr, • • • d- • • _ • kr) kt-) tr) tn tf-) - v-, tr-', kr7, ,- .- kr, kr) N ,-. N r-- I-- V-7: `...-.., ,..--, `,....., `,.. --, •-.., `,....." n--.., n-...., ,..-.., •.--, n-.." ,..-.., '...-.., s...... , ,...... / ,..--, ,...... , `,..-.., s...../

cu (D -) y ;.,1 (:.- c.) 13 3 ,) c„ — e.: m ct ,,,I ct 1..) cu cl) (i) cL) Y 1.) 1.) I..) (1) (1) CO (1) (1) ci) 0 1..) a) IL) 70 2 I) o.) 1..) (L.) eu 0 CL) 0 (1.) 1.) cU CU 0 0 CL) CL.) 0 ,.,OU C:-,1 (.:3 e.".")' c.:..1- (17.; c.:3 C c.:3 (:".") c.'3 el) (..31 c:3 --.3' u E (-8 L) gD „CD an gn ,Jn 4= an gn gn 4n 4n 70 4D aD 4D 4D 70 4: 0 ,.1 v) cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd Lk-, › g g g g g g g g g g cg g g g g 'm 'm 'm m c. 0 cl 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

L) GO 00 C.) GO C.) C.) C.) () C.) GO GO C.) C.) Q) uo 00 DO U0 00 00 00

r4r24' r:4" 4.' 1=4f24' 1:4 " C:4f24rilZ4ai ‘ a4r24.» C4' i24' (=4,—,—

AD V") 1/0 Ve) ln AD AD AD AD AD AD AD AD AD AD AD CC CD Ch Ch CD re; rr; 0-; 06 7t de d' d- d' d' d- 1' d- 71- 71. d- d- d- d- d- d' V0 V-) If.) 1-1 r. 1-1 1.-1 1-1 •-n1 1-1 v-1 vn

Ch CA ce) d" V-) CA AD AD AD AD 71- 000 Cl ;:.9 AD 140 CA --I V0 c0 V0 00 V0 Ch Ch Ch Ch Ch Ch Ch CO 71- 71- CD CD ce) ce) tic.) CA AD AD AD Ve) CD COONCD Cs- n1'. CD CD CD r;t, V0 V") AD 00 00 AD C\ICACA'CA-'CACAC\IN

tr) Cf-) Ce) c0 cr) - 1- CO CA —1 CD CD CA 1c cn oo o, Ch Ch Ch Ch rs 00 Ch Ch Ch r- Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch • /-1 1-1 1.-1 n•n-1 q•-n1 1-1 ......

d - d- Ch 00 V0 V."-) •:ct VO VO Ch Ch V0 NOOO- NNN ,--1,-, 0^-4 OCOCs, Ch cn on on cn 00 CO CD 00 00 r- 00 00 00 00 CA CA Cl cc) Ce) ce)

Cq Cq Cq Cq Cq Cq Cq 1- Cr) d- 1' Ch Ch r, 1.1,-, r- . r- r- r- r- r- r- r- r- nin VO AD AD CD CD C") 00 71- ,, oo oo oo 00 Ch 00 00 00 00 Ch Ch Ch Ch Ch Ch Ch 00 Ch Ch ,, ,, ,, ,, ,, ,, ,, ,-n ,, ,, ,, ,, •-n ••••,---,--.±.5 .-- - , ------Ch n.13 ,,c"--- V0 CD 23 , ...,cv • ...„ ...„, ,, 0 0 ,, ....„CA Cc) ,, kr-) r- le) 17:i ,..., ,_, -, N N Cis ,, ,,

V? V?Ch V0 AD 00 Ch ,--I 1- rs- Ch „ci 7:71 \D V0 CA CA CA AD AD V") CA CA 66 c-,4 00 > c.-1;- CNICr) d Qrn/ ,) r- o4 r- r- -~ -~ ô - oo \.0 7cnI-

, n1 v-1 C.) 1.) Q C/D = to 4 cd "To 7 7. 17i.) Q) a q (1), C ▪

.--, ,--, ,--, ,--, ,-, ,---, r- N r- r- r-- N e", •-n .--, •-n —'..,--, r-- ‘.0 "" r--- ,-. kr) r.: ,__, 4- ,„ r-- , ‹; d'' ,-. \NN- ,--. VD ,,---• ,---, ,,---. • ., kr) Q.-. in ) s s kr) tr) tr) - s s s s ,r) It ..-, ...... -‘ ,_..- ...... s-.., s„... .._., ..-‘ ,.._, ...... , ,...... , ,...... - - ‘_....

8 (/) ta , ta, cd Cc a4 o4 o4 c4 t-1 — cz, a a s2t., < a„, 2L.4 L., -7."z: )-4 8 8 8 z z z ,— 0 — > 0 0 Q mt c4 c)000'4-i -ciLd 0 ;:-,Ç,I=,—aaaa z "cD o o o z CD E. E-. E. Z o4 F-1

C) C.) oo uo 00 oo uo uo oo C) C...) O0 DO UO GO 00 00 121

vO C oo V L( 1f y) 0 tr) tri 0 0 c) c) m m m m m 6 4 4 4 4 cic; od od 06 c); ctn te-) v-) VD VD V.") n1D C r- N r- N N r- r- r- r- r- r- r- r- 1-1 1-1 4-1 1-1 1-4

cN oo r_ r, cq r- r- ,n r„ r- r- r, r- r- r- r- c) o cD cn V-) In cq vD cp cp cp Clr- t-, r- lD lD VD lD vD vD vD cq cl cl on cn ,n cqCD („1 cl CN Ch CT CT Ch CT Ch

d- 71- cf) N 0, „5„ C7) Ch cD'N CnC N ChCr C h 00 CT CT CS) r- If LÇ) • CN Ch a O c\ a. d d ch cN 1-1 1-1 ..... 1-.n1 1•n••1 1-1 1-1 •-n1

•n•1 • 1n1 If) kr) kr) If)V)In Cr C C c- -4-—(00000N N cr) mmmmmmm CD 00 00 00 00 \..ID cn

co cn cr) 4 4 4 7t CD Ch cr) CA CA cl cn cA Cl CN Ch CT cr.) cc) co cn If CD cl r- r- r- VD r- r- CN CN CN CN CN CN Ch CT Ch oo Oo oo CN 00 00 -.a, -. „. -. -. -. .. ,-. - ,. -. -. „. .. .. ,. -... ------_, -._ -.5 -.„ ,, ,... -... -...... •..„ --( --) --I -. 71- r-- 00 ,....C'''' vD CN on cn cr) cr) cl cl --1 00 -. -. 55555555N N clS-55 - Ch CN Ch I::: E-- -- r- In et-:. 4 4

a) m ac, n1D O N 7t cgs) Zi 1- VD In i Cl CD VD Z ---i In cc) • y) r- 4 4 4 4 ,:t. 7a- 71- 4 V:D CA ---1 r , ,...: oo cr) t-- nc) cA cv cl cr) VD vD cl d ‹) :=)' '6 r--- ((;:,3 ,,It • N co cn cr) VD nC VD (,. •... CD CD c) r- 1n4 n•n1 v•-n1 CD cn s..., I.. f=1., •_, a) ct a) a) a.) ct v, v) a . 0 8 8 8 3 t t `', e'c'.;) = -', 0 0 0 .-. •-,1 i P5' h,.8 _LL) ” 7.7 = 8 ,,_ _Q) = = 4-) 4. 4-) 4-( O4 O -1 OIO Z V) LT. W 1:4 o••—•n 71- 4 /-••n ,c) If) II) tr)

:-.cs 77 7,74 17 17 77 "0 77 "Ci 4_.

czt ct cd cz$ ct a)-1 ct cd C4 1:4 a4 IC4 cs:En a4 f24 P. P. c.4 (.4 (/) (.4 c/D ci)

-câ -câ -A -A -A -câ -A al 8 P. 4. 4. 4* 4. 4. LI. 4. 4* >cdcdcdcdcdcdcdcdcdct > > o o cd cd czs cd cd cd c.c! cd cd ct 00

cf) C/D C4 Ci) C/D V) c4 cf) L.) v) cl) n21 1f d .21 a4 P4 ri P4 124 124

on on on rn on on on on on on on rn on on on on on vn (7; crn cr c; cD: cs: r- r- r- r- r- r- r- 4-1 1-1 1-1 1—.1 1-1 1-1 1-1 1-1 1-.-1 1.-1 1-1 1—.4

ln r- r- 00 7r 7r ,t7t cDkr) r...... 7SRc> CN 7r ,„1D n,c) n4D r_ c)

N g g g .; cqt e7it Ri( cN cN (c72) ci on

n.CD CN m N 7r r") on oo n un un cN CN CN eh CN CN C) • „, CN Cr) Cr) CD 7r Cr) CN CN CN CN CN CN Ch CN CN ch „. CN In ch r- CN CN CN CN oc „„ CN CN CN CN CN CN CN CN ch CN 00 m 00 ch CA CD N 00 c— c— Z CD cN CD 25 25 C 4 oo on N oo on onon CN CD N n, m 00 00 nC) m m m

N cl cl cq NN N cl CD CN cN N r- r- r- NN r— r— CN CD CD oo oo oo 00 00 co 00 00 CN CN Cr, ,,„-, ,, ,-, „ -, -S-- -..... -.... ,...... „„. CD -. VD CN CN I17-2 C-5 ,-.4 ci c) --( . . -S-- ...., „, „„ 7r 7r 4 't cq :11 S11

n n 71- 0 N Cr) n.17 r- ln ..0 00 CN CD O r- Cr) --I CD ,-.1 --I n CD 7 nC) 00 7t ‹) CN N Ci 71- VD on VD '71- 0 N - 0Cr n ,-.( on nO Z, D kn CN CD CD m CN Ln nJD 0 CN --I LDI 00 N N 00 ,, 100 VD tr)

• QQ 1, s.(/) c./D a) a) a) e,t) • c) 0 0 IF•( ZC/)U(..) • t

kr) .71"

V-1 ,--. d- ,..,_,-----71------,---, _ _ - - - ,-----• - - 7t- 7t- .6 - . n v:i ri.' ,--, v:5 ,J: .6 v5 — v:5 Q:;' ,. 15 iS"- .d...- —- ,..c5- ‘A' trî ir)-- Ln. tri' .7.,-, tri' tr; ,r`, tr,', i---.' tr: t.-: tr,' kr`, C,---,' Lr-, ,...., ‘..... n n n n ir, Nr, tr, ,r1 ‘,.. ••' ‘...." ,• •, ‘,...,' ,.. .., ',....., ,..... ‘,..." ••nn/' ,....' •nn." '...., ,..n ,....., ,....., ,....., ‘...., ,..n., ,....."

77 77 77 77 77 77 7n 77 77 77 77 77 77 77 77 77 77 77 77 77 77

44 4 A4 P4 a4 o4 a4 a4 a4 p4 c4 c4 c9 e4 A4 a4 g4 A4 A4 P4 A4 o4 a4 o4 Çf) cip (/) L) v) ci) ci) (,) c4 cf) cn ct)

71' ------'8 8 '8 8 8 8 8 8 8 15 'a -8 "a -a -8 -8 - - - - - 4. 8 8 8 8 8 ctmctctmmmmctm9,' ctmmmmmcimctm cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd CC cd cd cd cd • .4 4 .4 .4 AL4 .4 .4 .4 .4 .4 .4 .4

DO 00 GO (.) GO GO (..) (...) 00 00 00 Ç..) 00 C.) CdO (..) GO Q..) (.) (.) Q) C.)

7t on on on on -d- d- 71- on on on on on on 7t 71- 71- Cn on ▪ crn crn c;n cs.n csn cn crn cn r-• r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- 10..0 1..nI 10.n1 1-1 1-.0 1-1 1-0 1.•n1

Cl Ch on cl on g, 0,%,), ,, ken 00 r, cl crn -. g 7t Ch ZD Ch Cs- Cs- nSD In In Ir) cel rl" '71" '71" vo \‘') 'D II-D NNr- 0 0CN cy cq cq cl cq cl cq cq CA CA CA CA Cl CA cq cl CA

On On d- CD 71' on ce) on on on --4 ,- 71" Ce) Ce) re) CA ce) on on 7t ---( 4-1 c> Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch Ch CN Ch Ch Ch ..7t Ch ch ch ch ch O 'F,4 Ch Ch Ch Ch CN Ch Ch Ch Ch On Ch Ch ch ,c) ch ch ch ch ch ch c:N ch ,..,-... ,...... --1 ,.. ,-. .--i ,-.1 .-- ,- ,--+ ,-- ,--1 ,... ,... ,--4 ,-.1 ,-.. ...-1 ,-1 -. „ Cis „ —.-„ ------7›, ,-7 ay- -. m -. m . Ch Ch CN oo ,4-, .i3 4 36 36 23 --,3 36 23 23 23 -.....--.0 CD -.....--I ,.....--1 ,--.1 CD CD CD CD CD z"..” z7! z7.... ç.:.2 5;.2 z.._7, Q .ç,72 z..7., zz, —, Q rin on on cl cn re) cn cn on on 00 00 on on rn cin r- on ren rin (-:•,>) cl

NN r- oo CD CD CD Ncq N NNr- cl cq cq on on on on on on on 7i- 7t 71- 7t Ch CN Ch Ch Ch Ch Ch Ch CN Ch Ch Ch Ch Ch Ch Ch Ch Ch s,• s, s, s, 7-, pz s, CA CA CA 27" 7:4 7.7.1 7.7Dn n.0 Ch ts- Irs- Cs- CD CD CD cr) on on on N cl cl cl cl oo oo oo r- r- r- r- 00 00 00 r- r- r- r- r-

n1D cq on cl on oo 71- un 00 CD d- d" rs. d' nC ,--1 00 lin ,Z, Ch on on Ch Ch • Vn lin vn • • ,---i CA Vn on ce..; -4 cp Cs- Cl '7i ''T 4 a4 o4 o4 ''' 66 .,CD:i -.-'.6.'l -.)c-) In OÔ OÔ ,....1-1 NNcc.,;,-1 M -. c..) c.) c.) m cD oô n17 n1D 71- c.:7 n f\i 4i 4 4D un un s kr) -Ft 71- s -, -. (1) o 1.) ,z1ei. •—, •—, •,- •,- T.),,_— - 0„,.., $.'i'4 , m,.. m, — — — 7.0 -,,, ,, '—' g › — — _ •- ° ° ° ' f,.'. . (.1....' . c..' ',...) 0 0 0 ''a ''a ''' 4 4 4 W W P> 4.1 4 -1 -( 4* C.) C) C) CD C.7 C) .p •

...... , In .--e , .1. e.".....-",,...., - ./-n •-.-^. /...,, ,,-^, 71- .71- .71- '7t. ' 1. • 'Ct n Tt .1- 71- 71- ,-1 ,-4 •- r.r.., .--o •-• •-• •-n n-i .-1 .6 .6 .6 .6 .6 .6 .6 .6 .6 .6 q:;' .6 .6 . . .. . , krç kr:. v> ,f;' vi Ir,' ,. r, tr) v--, ,s-) kr) u-)..r-, •..-, ,....,,-..,.._., ,...... 0, ,..-.,...-,,-.,,--, ,..-',..--

7:3 77 7:J 77; zi 7i "cl "cl -0 77 "cl '1=1 7:1 faq CL CL CL, CL CL CL CL P., P., •CL CL cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd a4 a4 g4 c4 c4 a4 a4 o4 a4 a4 a4 a4 a4 a4 a4 a4 CI) CI) Cl) Cl) C/) C/D C/) Cl) CID Cf) C. ) C/) C/) ""c ""c ""c ""c ""c 4. 4. 40 4. 4. 4. 4. 4. 40 4. 4. 4. 4. 4. V. 4. cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd > > > > > > > > > > > > > cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd

Q) Q) C.) un () () (.) C) 00 oo 00 oo oo cdo •

on on on on on on on on on rn cn on on on rn on on n 0 0)C\ cs; c); cD; c); N(D r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- r- n 1-1 enI 1-n1 1n.1 1-1 1.-1 1-1 v -1 v-1 1-1

\D \I) Ln 4n OC cr) ClC) 00 7t cc) cc) cc) on on 7f 00 cDcl 00 r- 00 N VD VD CD CD CD CD CD cD cp In 00 ,,,-, N r- CD 000000000000 00 OC „ „ te) „ CA CNI CN1 CA'

71- on %.0 d' S2 CN CN CN r- r- CN \•0 CN CN CN CN 0) 0) CN CN CN CN CN CTn 0C)0) 0) 0)

at ‘40 Cc) CD ‹D V-) cc) 7J- 'SD CD -D 23 23 23 Cl CA CA ,..--( Cl Cl CA CA Cl ,„ --- CD --- OC cc) on C) on -0000 cq cl cl cl c-si

N r- r- CN 0) 00CD 7t 71- 7r V) V) CN CN CN CN CN ON CN

- N C) C) Cl Cq Cl Cl Cl n N r- r- r- r- *Z- 13

Cl 00 00 00 00 71` 71"

•

—/.-", — ,..-... — ,--,—, ,..-n vi ..6 <5„,—, ,._, ._,kr- ,—, -.,...... , v.: ,—.,Lr:' v-,,.._..,

77 7J 7J 7J 1J 77 7:) 771 77 15 7:3 -5 721 -5 7:3 -15 77 77 /7 • 'E.' a% • a% FEL%* mmmmal.• o; fa; p4 p4 P4 p4 p4 p4 p4 p4 a4 p4 a4 p4 a4 p4 a4 a4 p4 a4 a4 a4 cl) V) V) V) V) V) C/) (I) c/) (/) c/) V) CI) C/D cl) c/) C/) c/) c/) C/)

-- - -I- dcc CC! CC: • > > > > > > > > > > > > > > > > > > CC; CZ: Ct C7i C71 CZ1 CCI C3 CZ: .4 .4 .4CfS W CC;W .4 W .4 .4 .4C7S

DO C) cdo C.) C.) 00 00 CID 00 00 00 00 00 00 DO GO GO 00 U0 DO D0 .•t: a a) (: a

.1,_)24- rx- c:4):4124- a4p4a4rx- r:4-- p4a‘ p4a4pe

M M M 71- MMMMMMMMMMMMM M M M c7; N - ..... 1-n1 1.•••(

0., CA CN g CN CA CA OA CN CA CN OA OA CN CN CN CN 00 00 00 CD C'n 71- 71- nJD n17 l7 nID ‹) D l7 tr) CD CC r- r- cN CA CN CN CN CA CN CN CN CA CN CN CN CN Ln Cl Cl

cq Cl cn Z, 00 M CN in (r) 00 In nC) CA CN CN CN VD l7 r- tin cs vn V”) Ch tr) cn tn cr, CN CN CA CA CN CA Ch CN CN CA CN CN Ch Ch \4:D cN cN . (...71 -(Z) C •—• 25 25 m Cis 2FR 4 25 '-O N '-O C17: c- Cl c- tr) CN Cl Cl CD CD C) N --f 25 --f 23 23 .0 25 23 c- Cl Cl CD - S SS SS SSS SS SS Cl C9 c O 00 ‹) 71- --I 7t nC) --I —4 11-) cn cl cn

C) Cl Cl n47 \.0 l7 l7 n17 CN CN CA CA CN CN cc) 25 25 4 4 Lc) N N Cl 'O NN0000 CN

I-, 4 4 m -1- oo In tr) oc 7t- N vo vn c-A cc) o ci Ln 71- , 00 Ln Ln ..c, Q.) ... 0,;:) _,_ in Cl , (-NI 4 cf ) r- 00 In —4 LC ) CD Cl 4 ,..,.: Cl Cl Y) 00 00 (r) C) 00 N 00 0 cn ,...4 cn '-O0 tri CA - ' dn c:11. 00= QD cl 00 00 A0 71- 00 (,... CA 4 CA IL -) U t--- ON ,•— 1— 71- '71- I:4 '7r i— M 2 .--4 Zs.0 VD C r-- AD nCD 4 71- m ,40 U a) 04 04 Ca., 0 C E -0 te ct 0 m m 1:4b - 0 ''''' q 0 2 :8 1.— ,.?„, 5 1-- m 17; 0 ca., A-',7-f_531c4,74 ..--1= — 7J •,--o 74 1.) - :7---(1 7J —6 7:j ,4 M 'h.-5 -6 -8 -6 0 '6 0 IDA) P4 a4 P0 00 = 04 = o4 ,. 04 W 01 .4 d - -

00 71"

,--. ,,---, ,----, - ,--, ,,---, v-) ,----, ,---, ,---, ,---, d- d- •71- d- -:I- ,- ,i- -71- rt •:I- ,-. n ,...1 71' '7 v5 \6. v5, 6 \ 6 ,-, nC; '''' '..-' ,-, ,-, ,6 \ 6 .6.' n15 trl. kr iri' lr) v-•n n n nIn tn - ,- ,r, .--) ,r, in ,r) ,...-, ,.._. s...-, ',.._., ....., _.., ..-, ,..-,, s...... , ..../ ,....., ....-, ‘.._.., .....

"te "e 73 7:) 73 7:3 77) 7i 7i "Te 7:3 "te "te

cd cd cd cd cd cd cd cd cd cd cd d d dcd Cd d d d C4 c4 o4 a4 c4 a4 a4 a4 a4 a4 o4 a4 o4 o4 a4 o4 c4 a4 a4 c.) (/) C/D CI) V) V) II) C/D (ID C/) (/DC/) CID VO VD CID CID CID

czs ct ct ri al cri CCi cri cri cri Ct C.Z; CZ: cri Ct CC: Ct

CZ:• CZ: CC: CC: Ct Cri Cri Cd CC: CC: CCS CC: C.C: Ct cd cd

on UcAc4c4c/DUUL)Uc4c/DUL)vDcAvDc4c/Dv) 121 (.21 P4 .4

Cr) cc) Cn Cin On On 71- Cr-) Cc) Cr) 71- Cc) Cn Cn Cc) On Cn eR cR eR C C c)R cR cR

If) 7t CA Cl CD CD VD r- r------r- on on C) CD on m m r r r r r CD 7t CN CN - 00 CD CD ',CD ,..CD CD cl 4-) 4-1 occ, r- r- c„..) co OC CN CN CN CN CN CN N " CA cil

g8- 77, CN vn on on on 't* 71- vD on -. on on 7t (7., c.\ CN V-) ,..fD CN CN CN CN CN CN CN CN CN CN CN CN r- z7( CN CN CN CN • crn CN CN CN CN CN CN CN CN CN CN -,4-4 N 7.7.1 ce-1 CA N C c cc o DZ., Cl Cl 5.;.2 zzi nap‘ cq c) rn 7t 00 vn 00 cA on on on on on on on on on CN

on7I- vn un vn vn r- CN CN r- r- rn CD n0 \.0 VD VD VD VD vD VD r- r- vD cn VD CN CN CN CN CN CN--( CN --. CN ,-. CN--4 --4 ,--(CN CN--n CN --4 CN,--( ,--4 -,,.- --._--1 -....--c -, .....„ ,_ -... -„. -... -,.. un N --n --4 --. --1 0 c- c- vn --4 CA ,--ICA NNNcn---(-1:: N 0 CN 13.1-) C-n1" -(11 vn v) 4 3E5

un 1 00cq nC '7t CA rn vD 'd' ,--.1 76 „zi 2 ,..... < (..., n _ cel ,--( N C\ ‹.) cs CO oog c:, 4 cs, (...) r---- d- CA ,_ Ir) vDc'l (.....) Z vD ICI CD fP. tO ;- c.) up En = = ct o >, -.4 -A •,-,cn •.cn •,-, •,-, •-, 1.) N 1-4= -1=i TI - c9D 7 hC D ° (1.) 124- ..--'› ;1'ZZ ZZr:1Eif:1(21 ; (1=1) g, Z

n--, ,----, ,--. ,----. ,---, h h Ln --, -- - --,h ,---, .,__,1- ,---.. ,--. ,,-, , ,. h . h ,-.. ---, ,---. h .--, ,---, ,---. ,---. ,__,v-:: 4_ ,_., v-- -- c,-, -- tr, 4 h 'i- If) ‘I) .---, -- Ln .. . . Ii- : 4 _..— '-' ,--, ..., V) ,r7; r--- Ir, -. , ,, ,- . , ,r) t- .._.., s.._.., ,....., ,....." ,-..-, ,-.., ...... , .._..,d- •-.., ..." ,....., ,...... , \ --, ,...... ,-- ,- --

cu a.) a.) (L) el) a)

72, Û c1-3 01., 1:3 7 721, •,,-,- ::.-' ,.g ,.. ,e . 2_ -0 (ID CC: 7:2 "0 ,.. . 'FI " ' CS - ' C' '7 r 4 - 1 Ac C(/Ci C C S 0 p4 c4 c4 czf ), M ,-- .-- .S4 ., ,... a4 r:1-' C) 2 8 CL)1) (ID cd cip (4 C'i '-' '-' -' 4-' - 1) a. a., a., a., a., 0.) - •- •- -0 c9 9c-)c =1),Du d) ,D,,,i allf21 p 4-, 4-, 3 -a -a -5 -5 to .17.: .4.7,' - .t.-- .- ,. , - ,.,ct ct..„. m o ) cd c,s ..5., , g 9 9 9 (9 (4 4,5 0 0 E cil C.-7s c,ci 7 5' ft ctS 1, :, ; , ;-, i. .7,.: ,==,r. 00 ,-. -.1 ,. = DL., al.., a, a., cq Q7 Q7 CD CD CD ,- N

00 Do 00 Do U7 GO 00 GO U7 DO GO U7 UP (..) 00 UP (-) C'° UP UP (21

cc) cn tr) nC) nCD ‘,0 71- V-) CN CN CN CD CD CA CD CA ln CS\ Crn C C 00 06 06 06 4 kr; 00 06 06 0; cs: N N Ir; tri Ir; Ir; N N r- r- CN CN CN CN CD CD 0 CD C) CD CD --I CA NN Cl CA CA CA CA CA CA CA CA CA Cl CA Cl CA

cd -0 m 4D oo r- r- 00 00 r- r- ch ch CD 00 00 00 CN 71" tf (:). cl r- r- ch oo Ch ch Ch Ch d - Ch ch CD 71- 7t oo oo on 00 VD kr) CN cD CD CD cD r- r- CD CD r- r- vn r- r- oo 7t CN Ch CA k.0 Cl Ci Cl Cl Cl Cl Cl

cc) cc) cc) Cr) CN Cl on cv oo oo oo d- ---4 on oo 71- on Cr) Cr) CD C) (--- Ch CN C-- CN CN Ch 00 CN CN CA 00 00 Ch CN CN Ch Ch Ch CN CN CN CN Ch --( --( •.--4 CN Ch CN --4 CN Ch CN CN CN CN Ch CN Ch Ch ,, Ch„ ,, ,, - -- -, ,„ „„ „ --„ ,, „ __, „ „ r, „, _„ „.., „ -- .---1 --(--, ---,--1 -, S- ---4 „ ---.. „ --... CD 00 00 r0 C) CD CD D r- cN1 cD kt) --I on on -Tt d- C) CD tc) ---n --I ,-1 C) CD --- S-- -- cl --( ,---1 -- CA ,--( --4 ,---, ,..-.[ --I CA CA --, S-, -, „ ,„ ,, „, SS „ -, --- .--... -- JD 00 cc) i::::: ,---( --( ,-1 00 00 cc) --+ 00 ---4 cc) cc) -C-:1 cc) ce) on on

c9 on on on on -71- r- cn cA cq CN CN on on on on r- cq oo oo oo r- on oo r- 71- CD CD oo po oo oo 00 CN CN Ch CA Ch CN CN CN CN CN CN Ch 00 00 00 „ 00 00 c•-• c-• c-• c-• CN 00 ,_c- Cl CA CA CNI CI „ N 4 eiz,

a.) n nID 00 c/B 4-1cA Ch7i • CA • ID CA M vn c.)00 c..) c.) kr) N ONcl 71- tn CD CN VD "7-/ cn • • • r) kr) d- d- c.) r-- C ce) kr) tri vn c 00 c cl un r- cq r- 4 4

CP) (1) C1) c4 cc c C) c4 (4 • • • (,---1) • ,1)'-( 4'o ôôô 4 4 ti) bA tt tt ! il

• ▪

r- v•-• — 'Ti- m r-r- ,-. ,-. oz; ,,-;. — — „--. v.--, ... ,-, -, r- a.) -- „...; „..; r- — N — r: .6... (...,i' , . — — r- 6 .6 .6 rz r, .6 C-7" G" v--, 4 11-1' 4 4 V-1' -CD ,....., 'N../ ,...../ ,..-/ ,....., \-../ ,-, ,.../ ,....., n-, ,-, ,...../ ,...., ,-,/ ,....., "...../ ,....., ,...,

-5 77 mi 72 77 72 v• U0 U0 coz$ M 64 Ct C7; Ct Ct a a I a a - 44 1.) ell (1) (1) 0.) cri 0.)ccEca.) 1.) CL) 0.) CU I) 0.) cll CL) CL) 0.) CD Ob CD 0.0 Of) 0-0 CL) 0.) CL) I) 0-1 CU 0 (.111 (13' :)4 ej.2, -2, • (13

-ri -ri-ri-ri-ri-ri al-, -10 71 "CI Mf V) ri V) Cl) V) 10 1 1.1.1.1.1.111 a.) 121 121 ICI ICI (4 CL0 CLI 121 121 r4

0 -0 Do Co0 DO C) VO Do Ci] v) Vi

r-1) O C-) L.) a.)

c+-n 0 14-) lD r- r- r- r- r- r- r- r- cp CD CD CD CD ;-n ▪ a.) tf- Ir; tri 'tri tri Ir; 1c; In Ir; vD vD D \.,cS CA CA CA Cl Cl CA CA CA CA Cl Cl CA CA CA Cl CA CN1 CA P-n .à:1") CA CA CA CA Cl CA CA CA CA CA CA CA CA CA CA CA Cl Cl t1.0 0 M „C) c7; CN \CD C N CN 00 00 \C) Cr1 nC) C h-. CD co cc) 7t r- oo oo oo oo oo oo oo r- r- oo r- r- r- cl Cq a° r- r- r- r- r- r- r- r- r- r- r- r- r- oo r- r- cq c,1 c,1 cl cq cq cl c,) cl cq c‘) c,1 c‘i cl

C) ›,71 CA CA 71- ,t 7h 7h 7i" 'Mt 71- 71- 7t 7t 71. 71- "z1- r- r- ch ch ch CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN CN n •—• —4 •— ,„ ▪ 14-1 VD r- r- r- CD ,---' VD vD CD lD• lD• vD tc) cl Cl cl ---' CA "..-1 Cl Cl cq -.-.• 000000 00 cr"1 on rn rn rn on cn rn on rn on rn rn• rn• rn rn ›s'

ci) CANCnCnCnCnCnCnCnCnClOMMCI)CnCnCr) C) 00 00 00 00 00000000 00 CD 00 00 00 00 00 00 CN CN 00 00 00 00 00 00 00 00 CN CN 00 00 00 00 00 00 •-1 •-4 •—n v•-n ..... •—• •••.„ 0 • C.-• c•-• c•-• c•-• C- • c•-,„• C-• -d 4 4 4 4 4. cz: cL) c.) CD ,4 ,4 CD CD CN CD r- ,:i 77i CN CN 7h C) 7h VD CD CT• CT d d d d ' CT tr) 3 v-) " 71- Ir) In

z z ri O O 00000 C..> . 0 C.) 0 C.) .(,) 000

' '0 S744 'g g g g g g g g

• •

-c) ce o a.) y to cd li y ,.. ,.. io

13• 0 N a cz . b•O▪ 1) 13 13 13 y ..--›,»* ;-, (.,.7 ,_. o EL, ,,, •› N (L) ..-I cd cd __, a) s-*0 a) ,4Z (1?-'1-- ..0: ^C) oz: 0 tg 0 -C3 1.3 I--n s. g-t :E 0 13 'd cn ›,. c-') •,--,› s-, a.) a)) cl'S (24 › 1- a) O .... t-i-c a) ---' "d rZ4 (:) .5 › ;..,m o ° E -cl -cl cd cd 1.) Q)< .2 s. -do up ,b ct z a> Uo• . 7: , ,A › te a) CL, = a) ....,,,, 5 m ,..= (._) 4c) .. cii ,,, a) o c.,/ y o .--,_ = -E ,4 g ,. c., . ,.,-, ',..4 4,1 o (4-1o $., ,„ — o N 13 cn ,.. 44--• 1-, ^ri = (..) • -, a.) 1..) •,•-• ;_, .,, to c..) = (1) 2 cz9' u.. t •— „cy a)ct 2 ri) '3ri) . 2 . b.0 -CD -cf ,. O 0. ,.. 4_, -c) › 4_,'ctd' ,_ e" • r•-n CL) a) B —,'ai f=1-n .',72, = 0 to . ,5 = -o .,.<1,z,) .-' ° C3 13 -a . . ,.> ,.., c„ ---, › (,) ,.., , ' cA •-ii „I=. , Q.) , „.., cn 0) 7::), .a.2, -.E o ,4 = .':8 5 c),a) › -o o 5' = ". 8 .b.0 0 0 „. 0 7=) "Ci 1.Z 0 (1.> . 7 o--1 , ,. „_, —. 0 7E, •0 ..'''.), 0 CA 'Tt' (/) >, '41 -^ y '''' S' el .5(1) rC) 0 c) .5 '771c b..0 .-cs •— a.) 0 ,..zi cn — 4 :8 (1) :—, ra ., .4.J E cd cd cl.)h & .2 "ci c, C.) .,„y •sa, Olai) o :. , ,.. y ...„.., : II. bi) eu) m --' 3 3 I•..., › > ;-, 1-, o • u, ct Q)00 1..) ›, ;-n • - ç-, 0 ,- -c) czt .. › • 71 b.0 a4 a) -tc • -, c,z/ a) bA › › = ._, • - -ir.,, m c9 -c) a.) (A 0 -•-•' up 0 › "Cl Z? "d („ • cci 7:7 I.) a) a) 0 - ,c)cl) ,- up -, ci, al ,.= b•O 0 cl) al.,•c,r; ..= = ;Ej - cl.) . 5, . --' " d- --. 0 • --, .. -c:1 aq = I 7:u;,' • ,,,,E 08 .51) .,,,,,,u., -5,1) rcy .,;,.° 8 ..., ._. rz,L., -F) c,9 0 • - (') 45 -ci c.) (1) ,,,— czs <1.) c.) = 3 ;' Zzi. ^c) 5' '' ,.= .... ,.. -ci ....., ••—, •,--, f t a.) c.) --, -= (L) 4,ct •—, c.) • —,0 ,. —. 0 ct fa ct 3 '''' sa .•--, .5 a> .9-, a) = n-' 3 : •,„ •,..• ct „..) 4— / . - ,.= -4 .5%' 3 „°D ';c 0• C3 •,s: g r(8 0 ,... 1-.) 0› c,: ,:) ct._ o ri . —. 4, 0 0 ..-- ct > 'A '--. — ...-- .. - - -7) -0 NI ,-,.) 8 0 — ct c,,- .- .2 ° -z 78 g ,,b Lt r, '- -,9z aa c,, ,,„ az ,-, — 0 c, b-0 rzi t)•0 d, j Ct Ct ,C, .2 .4 •`.r. •, -E‘ 3 z 'il ›, > = —4. =, , ,-- -d 44• (''., fz,.., I-, I-, I.., ,,I., -1-. Ct ,,,,3 7. >, 4 4 4 cn !-,t, *4 '' Clà "Cj E g..Fo, ô flo a.) a.) a.) a.) s.,-, 0 g gl, c...)(a)laiwC..7cdZCA-,a41,V) < L) C') ,:r v-, ‹;, r-00 O\ c) , .(1) - c'l c'-' '71- lr' (:) ts-

• • • ▪

.-, (...) a.) CZ$ CC; C71 C.C; CC: CC; CC1 CZt CZ: CCI Ct ,•-41 ci ,- LT-I 4-1 4-1 4-1 1-T-1 4-1 4-1 4-1 .n T-LLLI. cti 0 c) yCS_, . 0F..) (4 • ';-( 0 .. 4--' ,.(1) -= (7)' E "--' -64 ô ' c'5))-4 0 72) , /I • tff° r) '-' 7:3 '' N d- \.0 re) CC 0 O o E 7:3 '4 › . 4 N tri v6 (-,-; N -4 -4 5 N ,,,, a.) 7t ,75 cal ct 0 1.) o C..) >< E › '7»a: - 6 -:) m . ;... c.) .1). 0 ;L. -5 1.) ....' — •-c a) el.) . 7:) r-c; cip t....c -0 773 6 7.... • I:, 64 - ' o -,•:. --g -0 0 0 (:1:1 ,j'-' ...,_ ct 0 • PC1 ,- i tzt) (-) vn 0 ‹.) rn co oo tin 0 N ct ck ,..,:i :2-4 14 c...) u C5 C5 C5 6 6 c,1 Csn ci) co:-. ,1) 4..)i . to ,,D,4 8 ,E,J,

-c, 00 L.1 • ..„ (IS N 0 U 7 ,4 ' P = - c ) 0 cn . - -, ct .'" a) c.) <+-n cf) , a) c) >, c4 a) O „'-' c.) s-, -.-J 0 ra, cn _E ,,,,, m -d ..c7 E-, cA a);L.- RI 0 a) al; , z ..'0'7) 0 (4 "1:/ 1..., —

- i- cep n 5 ,,„= nj = 2 2n c4n tin oo C) v:) d- ,C) •4.--4, -ci c./D 74 ”,ct 4.-,/,) ;--. ,-..., • .-, C5 C5 -21 •-n 6 C5 6 C5 C5 C5 -4 -4 4.-+ 0 = Zelf)(1) • (L) • 0 2) (L) ,9 c,>, 'T,,,' ,,, a) ".° c.) = '" . 0 4 '' a 2 ct c' ..5 „.,1> o u ,s „c o „ . CA,.., CA • '''' 1) 0 ,-,-. lr) 0 b4 s---' E ,.. ,-, ,.o 1-1 0, -0 >r:,•c;-; 0 a.) = ci.) 4 a) ,4 z • 7-3 —, .4 . c.4 -0 0.., u z •-4 0 ._., CS oo NNN cci c4 -c) › 4 vi "K:J N c-n I n N N 'E C5 —; •- 6 c5 6 0 o0 •-4 -0 c), - -,- ct 0 a) c-' a) ,- ct = `c-fi,) I-, •••• n ... 0 0 In1 1:) (I) 0 (..) --, -0 -0 •-• ct 4- 0 4-4 (71 0 0 >-, 1-. a.) a.) , 0 • Z '-' f2L, U " '" O ct .S..) c,, cai ,- Q.) E c„f:)., u = t), ,.= • -7) a) — E — C5 c4n rn cal ...Y, .›, cz; 'T'zi .-7 Cs N , 71-. 00. -4. co. . . . . n = ..z CD s•-n .4 z (1.) . -,-,;- RI -4--. cq (4 r . n' v ç - C r) 79. , 4 4, - ,' , CS " 1. a.) ,-I-d VD5',=,(D E 7:: c) cd a) 0 c.c (....) 0 , c,) ,.fL st -,1) a E v) 7::=' ° "ci ,_, ,__,• .- c°1 ,.. ›,' ''7,' E ,- .';', ii5 (13 -0 -R1 - o 00 c-1-1• *--'Sn., -ci 0 N•- n77 -5 gp" W.1 i•••4 O ry.-11) . 5 fa, 40 -0••••-i •-•• -0 -0 1:14 VD M = ,^ 4 Sa-, 0, 1-' ,•-•,'-' C.) O 0 z 0C..) 0 - .40 a) • 0 -.1., .:-, z z z ,-. V) ,In .4- 10 .-0 ••C V) 4-- CL) • •-n 1--, li 1,) 4J O c.") V) 5 5 V) VD n • r•-I 42, •.n '0 A) ,..,' = ,i . ) 0 z 7) U 3.- 3--4 -,.J 64. 111. c(),, C t 1) 1'2 ,,__,0 4-4 0 1:4 r • al '-',,,, r ; ,4 cr"t) - - - C V) . r ezi- ,:a '

(1) 0 C4 = • 1., b, 0 (1.) 7:) . 1.., ,t b.!) cd ›. • b.0 C.) '' c) c) . ,„`";D . E3 '45 (1) 12, 0 0 cip et — •?-;up 0 C5 4'• 5 ° vp 4, ,= c.) 3 71 • of) cu &.) ..,^ cn ,_,,= ,__,0 4 o b.o $.4 • -6=___,a.)Eoou,„..s--4 - '4-4 O ,4 o -Qui ci) "„::, O. a ...-.4 0 z '1) .2-4 8 -ti ..c.

0 0 0 0 0 0 0 0 0 0 0 cd cd cd

CD CN r– Cl In 00 CD C) C7 C7 CD CD CD CD C) CD CD Ch CD

• le) \t3 N cci 4 nr3 -4 71: C,1 Cc) V-) 6 Cr.) N cl N cn cn r-

Cl In 00 CD CN D C) C> rn 7t cn tr) CD CD Q-: • ,4 cNi N cNi oc; ktD \.6 4 V-) C:t COn C,1 N

a.) re) 00 00 CD kn 00 In Cl cr) 71- Cl n4D C7 71- Ccl C -4 -4 N N C .4 -4 .4 .4 .4 .4 .4 .4 4 cNi

CD (NI (NI CD CD CD N O CD n4D —1 r– 71- (NI CN 7 —; -4 -4 N -4 c cNi (Ni -4 -4 cNi -4 cNi c‘i -4 cri

CA a)" CD r– 71- \SD nr7 Cl Cl cn r- N N cN kn CD CD n N -4 N NNcNi cri 4NNNNNNcri cri 1-.5 C6

-el • C. 04 04 CL, CL, CL, 0 0 0 0 -00 0 -0 -0-0 lO 77 GO GO DO GO DO GO v) cU v) cll ell cll cll a4 1=1.4)

rn CD 00 CN CN CN ""

a4

kr) cNi nso,

71- a) a) .21 ccs g cg cg g g cg a i4LrLr.LrLT."g •T•-1 LT-n 4-1 4-1 LT-1 4-1 LT.( 4-1 4.4 4-1

C) 0 C) CD \ID CD 0 Cc) CD CD C) 00 \CD 0 00 CD (r) CNi C\1 \.r3 c5 1r) cr; tr). 4 Cr) N N

Cl CD CD Cl lD 00 C) Cf) cr) CD r— CN -4 .6 cri 4 c\i (-Ni 4 C;\ (5 (.1.) cr) tri cri v-, ^-* --I 71- C•1

. 00 Cl Cc) CI') CD CN CN . . Cq. Cc). cc). CN. CN. Cl C c\i 6 6 0

Cc) Cc. cc) 71- cc. cc. CS, 14-). 1.0-. \CD 71" 71" CN NcN cri -2 c‘i c\i —4 —4 —4

COCOC) nID O N 7t CD CN 00 in 71- 0 CS\ (r)

4 Cr). C‘i CNi C1-; tr; CNi 4 -4 cNi c\i -4 —N

'7»at • g "c; " d c3 o o 4 cr) (D 4 a) a) cr) a.) a) 4 a) a) a) (1.) a4 r:4 CZ4

CD CN CN CN CN

o (.4 vi

nID 1.1-) OC \CD nXD kr) 71- 0 0 0 cl at cl al al cl cl cl cl cl c71

CD r- c) C) CD C) 00 VD CA CN cc) Cc) lD 00 Cq kr; kr; n cS N C v.; tri c•i 6 6 6 CD „, .16 N Ncq

C) cl 00 m cl 7P V") cc) CD ir) CD V-) Cl CN VD cc)

C V-; tr; V) \.6 06 6 cr; --:, 6 6 6 6 CC

- 00 r— 00 V-) 00 CN V) NN kr) kr). . . . CD. VD. V ). 6 -4 -4 -4 6 6 6 6 6 6 6 cS

-5

1,c) VD N D 00 kr) Cq Cl CN CN CN 00 N tr) 7t -4 -4 -4 -4 c•i 6 -4 -4 -4 -4 6 -4 cS 6 6 CC

• CT 00 r— lD 00 00 VD Cl 00 V-) N 00 lD CD In m m cq -4 -4 NN 6 -4 c\i c•i -4 6 6 6 6

-17) 4 4 4 „, a.) 3 3 3 3 fa, a) a) a.) "a)El -ai ec)aiO ai "0 0. '0 7) "Cl f2). tn. S21, f2-4 (1.) c)..) (1.) (1) tll tll tll cf) c7: CI 04 E"" 04 a4 a4 o4 c4 &- o4 o4 a4 a4 E-4 E- E- E- E-

Cr) CSn 00

cq a.) r- il cS r- =Z Z Z = ZZ Z cd cd cd czc cz; ct crs czt czi cs cf)

C) Cr) Cq C) C) 00 ‹) C) C) rn cS 6 6 6 cS cS cStr; 6 v:5 c; 4 6 6 N N N N NN

0 .. „. tr) CP\ ,4) 00 6 6 -4 6 cS 6 6 6 Cc; NCNOn NNN

71- cr) Cr) 7t rn CA 71- Cr) u.-) cs 4 c) cc—cc 6 —4 -4 6 —4 6 —4 —4 —4 -4 6 ,4

7t '7r Cr) 71- 71- r— oo r— r- oo r— m 6 6 N 6 cS 6 6 c> -4 -4 —4 —4 N -4 -4 N -4 —4

n C5 If) V-) 0 V-) r— un (-NI 11-) C) Cr) CA Ln 00 Cr) C•i Cr1 C•1 Cr) (Ni C\i C,1 6 6 71: CS C5 CS C5 C5

4 4 4 4 4 4 4 4 ai as ct ccs at >.< cu cu cu CL) CL) (L) Cll 0 0 0 0 0 0 0 0 0 0 f=14 1=1, S=1. 1=4 =1, 1=1, 1=1, O cd cd al Cd al cd Cd Cd al 2 2 a al al 2 2 2 2 F-1 E-1 H' H H H Hi

t:4

kr)

0 al al al cc3 czt cci M at al CZS CCt Ct C't aS MS VN .]E0 r.1.-4 40 4-4 4.4 4-1 40 4-4 40 40 40 4-4

Tt Tt 0 CN Cl CN en en cl Tt CD 00 Len en t----: en 6 en v6 c‘i -4 en Vr; kr) 06 cs; N

oo en oo Tt C•1 CD 0 CA cr) VD oo en cl o 4— ee.; -4 6 6 -4 -4 -4 en c‘i crn c‘i eei

î II V -) II aq V') 00 Tt kr) kr) kr) (.0 t-- CN CN CN 00 7f CD V-) NC VD cr) cr) kr) cr) 00 00 -4 c5 6 c5 6 6 6 -4 6 -4 -4 6 -4 -4 - 4 -4 -4 6 6

In un kr) en un 00 Cc) GO cl 00 in 0 n4") cr) 00 CN 71- VD (-NI "4 6 ,4 "4 -4 6 6 "4 „4 -4 ,4 "4 "4 "4 "4 ,4 "4 „4 "4

Vn un 00 Oc —c cq cq un oo oo ,f en cl cy . ,t. . cl. . n.ID. 00 cl"). . 00. "4 (Ni "4 "4 -4 "4 cl cri cq

0 00 c„ » E E 5 c>t- " gca, a) 0., o $a, z c/ c4 18 c';) 78 riD c/D fo4 a4

00

1-1

o

00 a) CC; CZ: CC: Ct CCI C71 Ct CC; CCI CCS C73 C7i LT.- f-T-0 LL1 LLI LL1 W. 4. 4. 4. 4. 4. 4. 4. NNNN

C) CD CD 00 CC' 00 kr) CC CA CD Ch CD CD CD. CD. cc; CNi 4 4 kr) oc; kr; 4 CC; Cc; C,1 \.0 Cl Cl \.(D Cl Cl 4.1

CD 1,c) CI 00 CD Cl Cl Ch CD CD CV CD cc) cl Ch tri 4 C -4 4 „4 -4 -4 ,4 -4 kri 4 06 c'7)' (-A -

00 Cr\ CD 71- oo cr) CD Nr CN Ch 6 -4 ,4 6 ,4 e5 -4 -4 -4 e5 6 -4 6 c5 6 N -4

cq (.1 CD rn n.(D (NI cl cc) kn CC 00 In kr- CD -4 c\i -4 -4 ,4 -4 -4 -4 C\i ,4 ,4 N ,4 c\i c\i

,t 0 71- cq ,t (,) es, cq kr) 00 Nc1 kr) kr) kr) kr) -4 r‘ict -4 -4 -4 c\i -4 -4 -4 71:

> > • > •,,,71 -8 -8 -8 -8 -8 -8 "8 -8 -8 r:). c1 a1,3 la., 3 3 3 3 3 3 3 3 3 7i fzIL 7:3 7::) "0 7i 7:) 77) 77) "Ci "0 '0 1)4 4 c4 4 CA V) VD (24

Ch Ch Ch 00 00 00 CN Ch C' --I --(

a.) Cl ZZ= CZ: al Cd cc: cd 4.1 V. 4 4. 4. 4. 4—n 4.[ 4. W. 4—n 4. 4.

CD CD. C) CD C) CD Z) vn \.1D m m o0 CD CD CD CD rn 4 cO 4 vn C,1 —4 -4 -4 6 —4 -4 r4 r4 CN o0 cl m cl

CD CD CD m CD CD CN oo nrD vn vn rn 7r 7r In 00

• CNi Cr). tr; Cr) 7 Chi 7-, C5 CS CS CS C5 CS CS VS -4 cNi (In N N N •-n

00 Ch 7r 7r Cr) Ch 7r \ID v." CN r— vn \C) rn \JD vn • c‘i -4 6 6 6 6 6 6 C5 6 C5 C5 —4 ,4

7r in Cq 7r In vn r— CN CN \.(D cl r— c) CN cS 6 —4 cr; cri—4 —4 6 -4

• vn n.4D 0 0 00 In. 00. in. . kn. 7r. (-1. rn. cl. vn...... 1-1 1-4 1-1 • 71: Cr). Cr). Cr). Cr).

b > » > > > » > o czt czs czs c4—c 7:1 o o -c3 7:7 0 a) a.) a) a) a) a.) 2 an w, )-••

cln cn C'N C7n c7N CN

1:4

\.0 o0

r...... C> vn \.0

CC: CC: CC: CC: CC: CC: CC: CCS CC1 CC: m C/D cd cd cd ad

CD V) VD CD VD VD 00 CD V) CN CA CD VD CD CD CN

1. - n ; cri 4 c‘i c‘i 06 co 3 a\ 06 --I \.6 (Z) N cq 44

. VD N 71- 71- 0 . . tCo0...... Cl Cl V) --I --I Cr Cl Cr) Cs- Cr) 00 VD Cr) N f N N V) V-) V) kr)

V). c). . . 00. VD. . VD 00 71" VD CN VD CN VD C- —1 0 cp 6 —4 ,-4 6 —4 —4

CD 00 00 r- cyNN kt-) 7t Cq . . . . . Cl. CN. . . . —4 —4 -4 —4 —4 -4 --I cl —1 ce)

oo tr) r— r— tr) 4 71- CD 00 14) Ch Cl CN CN

-4 cNi cNi -4 c-f- cNi -4 NNNNN Cr) Cr) cri

ct ct 3 3 -c) "c:J cC: ct czt czt czt czs a) a)

• Q

-tC -0 cnK 3 -, d,...... ----. . - s. I I 0 a) 0 N -.0

I I a) > U = 0 c0 -0 N s...,03 -.0 0 I I 0

ct (.4 -0 c lr --a 00 H ri 0 -0 - 0 ..t 0 -0 ycd d ,9 t ;...

.c Q4:q- 4 i i 0., cU a) ...0 -0 '= 0 c) >, 03 0 -0 X N a) 01 ...... x ..= II E 0-4 In cnu c:c- ° 6 d.) 6 ....- $-. ...= a.) . 1) -, 1.... = 7.') a4Ii 3 ° cd m x L.) ,.' 7d `><) i.,u cal = N a-) X 0 -- -0 ---.' c.c 0 • = a) . -- X ,-c,' x ,,-t 2 -0 a) 0 . caL ar*z-4' u -0 Ii -0 0 ri 00 › g cs ta.t.) › 4 L: po x "7' t.) Jo ,, .t. II '1=:) ..- x 3 o > • - 0 S' 6 44' c.) cu I I „c 0 cd - ,_, ,.= .> -' .',-E ad 3 - 3 09 bo up -5'-6 0 0 , _ 0 -6 ...: = = .c.,,175 V) ° i.." 0 . 0 z V) . r., R , 1-,. —, n Q. 0 _c> - -, a..) 4- • - . t> t) = = -=,,,,, — IS -0 -0 t00 :n., . -.., a) •r: v) 0, c) czt cal ct 03 cd 'a Q s, }-n i•-n {-, L. "ti Mi 0 0 0 0 0 a-) 0 e., -, C.--n 4-i 4-r 4-1 I-, CND 7:1 77 "Cl 7:1 70 0 cc! 3.) a) a) a.) a) a = = 0 = = = . - . - . - . - — ca= -t-_,cl" _..1- , Q a) a) a.) a) Q a) a) a) a) al •,- -0 -0 -0 -0 -0 I.) a.) a.) a) a.)I- EEEEE -c) a)cd -0 Q Q Q Q Q .B. 2 c)c)ddoo y › ›. >- › › pc) ,.= Lin .0 r- 4.-n 432

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

ASHFORK WINTER STORMS

1 118 01/17/88 01/18/88 n.a. 146 2 114 03/10/86 03/13/86 n.a. 52 3 97 02/14/80 02/22/80 n.a. 32 4 87 12/08/65 12/19/65 n.a. 23 5 81 11/12/78 11/15/78 n.a. 18 6 75 03/01/78 03/06/78 n.a. 15 7 70 01/17/90 01/19/90 n.a. 12 8 67 02/11/27 02/17/27 n.a. 11 9 65 01/06/87 01/10/87 n.a. 10 10 65 03/21/54 03/26/54 n.a. 9 11 64 03/01/91 03/02/91 n.a. 8 12 62 12/17/78 12/20/78 n.a. 7 13 61 12/28/36 01/02/37 n.a. 7 t14 60 12/03/66 12/07/66 n.a. 6 15 58 03/02/23 03/04/23 n.a. 6 16 56 01/26/57 01/30/57 n.a. 5 17 56 01/03/91 01/05/91 n.a. 5 18 56 11/29/81 11/30/81 n.a. 5 19 55 11/25/85 11/26/85 n.a. 4 20 54 03/19/91 03/23/91 n.a. 4

ASHFORK SUMMER STORMS t 1 151 08/12/90 08/16/90 n.a. 139 2 120 08/18/88 08/23/88 n.a. 50 3 119 07/09/19 07/18/19 n.a. 30 4 112 08/02/48 08/07/48 n.a. 22 5 107 09/23/83 09/25/83 n.a. 17 §6 107 09/03/39 09/07/39 n.a. 14 7 101 08/28/51 08/29/51 n.a. 12 8 96 07/27/88 08/02/88 n.a. 10 9 95 08/25/88 08/31/88 n.a. 9 10 87 09/16/25 09/18/25 n.a. 8 11 85 09/11/69 09/14/69 n.a. 7 12 83 10/17/72 10/20/72 n.a. 7 6 t13 80 08/15/83 08/20/83 n.a. 433

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

14 74 07/24/82 07/30/82 n.a. 6 15 70 07/12/81 07/18/81 n.a. 5 t16 70 08/03/54 08/06/54 n.a. 5 17 69 08/26/53 08/29/53 n.a. 5 18 69 09/20/52 09/21/52 n.a. 4 19 66 07/25/36 07/29/36 n.a. 4 20 63 07/23/64 07/27/64 n.a. 4

ASHFORK WINTER DAILY PRECIPITATION

1 91 n.a. n.a. 01/18/88 146 2 60 n.a. n.a. 02/28/91 52 3 55 n.a. n.a. 01/18/90 32 4 52 n.a. n.a. 02/19/90 23 5 50 n.a. n.a. 12/31/15 18 6 47 n.a. n.a. 01/04/89 15 7 44 n.a. n.a. 11/25/85 12 8 42 n.a. n.a. 03/01/91 11 9 42 n.a. n.a. 03/11/86 10 10 42 n.a. n.a. 11/30/81 9 11 42 n.a. n.a. 11/13/78 8 12 41 n.a. n.a. 03/02/70 7 13 40 n.a. n.a. 11/11/85 7 6 39 n.a. n.a. 03/23/54 14 6 39 n.a. n.a. 03/01/78 15 5 16 39 n.a. n.a. 11/29/33 n.a. na. 02/09/93 5 17 37 5 18 37 n.a. n.a. 01/27/57 37 n.a. n.a. 01/06/89 4 19 4 20 36 n.a. n.a. 11/12/78

ASHFORK SUMMER DAILY PRECIPITATION 139 1 88 n.a. n.a. 09/24/83 82 n.a. n.a. 08/23/88 50 2 30 3 71 n.a. n.a. 09/12/69 434

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

4 61 n.a. n.a. 10/05/25 22 5 60 n.a. n.a. 08/31/85 17 16 56 n.a. n.a. 08/17/83 14 7 54 n.a. n.a. 10/19/72 12 8 53 n.a. n.a. 08/25/20 10 9 51 n.a. n.a. 08/28/51 9 10 50 n.a. n.a. 08/29/51 8 11 49 n.a. n.a. 09/16/25 7 12 48 n.a. n.a. 10/06/16 7 13 48 n.a. n.a. 10/04/72 6 14 47 n.a. n.a. 10/05/40 6 t15 47 n.a. n.a. 08/13/90 5 16 47 n.a. n.a. 07/18/25 5 17 46 n.a. n.a. 10/07/72 5 18 46 n.a. na. 09/21/52 4 19 45 n.a. n.a. 08/09/41 4 20 45 n.a. n.a. 09/17/61 4

BRIGHT ANGEL RANGER STATION WINTER STORMS

1 186 12/29/51 01/02/52 n.a. 57 21 2 183 02/17/80 02/21/80 n.a. 3 171 03/01/78 03/06/78 n.a. 13 4 153 01/06/93 01/11/93 n.a. 9 5 137 01/15/79 01/19/79 n.a. 7 135 12/17/78 12/20/78 n.a. 6 6 5 7 129 03/15/58 03/17/58 n.a. 126 03/05/95 03/06/95 n.a. 4 t8 4 9 126 01/07/80 01/15/80 n.a. 126 03/17/83 03/26/83 n.a. 3 10 3 11 123 01/13/93 01/19/93 n.a. 110 01/04/95 01/08/95 n.a. 3 12 3 13 108 03/20/54 03/25/54 n.a. 104 02/13/95 02/15/95 n.a. 2 t14 2 15 97 01/24/57 01/30/57 n.a. 94 03/01/52 03/02/52 n.a. 2 16 2 17 94 12/25/83 12/28/83 n.a. 435

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

18 94 01/07/52 01/09/52 n.a. 2 19 93 02/10/92 02/14/92 n.a. 2 20 90 12/08/84 12/17/84 n.a. 2

BRIGHT ANGEL RANGER STATION SUMMER STORMS

1 140 08/28/51 08/29/51 n.a. 81 t2 84 08/01/63 08/08/63 n.a. 29 3 82 10/30/92 10/31/92 n.a. 18 4 79 10/08/60 10/11/60 n.a. 13 5 79 08/26/53 08/27/53 n.a. 10 6 70 09/20/52 09/23/52 n.a. 8 t7 68 08/24/87 08/27/87 n.a. 7 t8 66 07/24/83 07/28/83 n.a. 6 9 66 08/10/50 08/13/50 n. a. 5 10 65 10/14/60 10/18/60 n.a. 5 11 65 07/31/51 08/05/51 n.a. 4 12 64 10/20/79 10/21/79 n.a. 4 13 61 10/28/74 10/29/74 n.a. 4 t14 60 08/16/63 08/19/63 n.a. 3 15 58 08/16/62 08/22/62 n.a. 3 16 55 09/05/81 09/10/81 n.a. 3 17 55 09/22/67 09/26/67 n.a. 3 18 54 08/30/92 09/01/92 n.a. 3 19 53 10/24/71 10/26/71 n.a. 2 20 51 09/29/83 09/30/83 n.a. 2

BRIGHT ANGEL RANGER STATION WINTER DAILY PRECIPITATION

1 129 n.a. n.a. 12/30/51 57 2 101 n.a. n.a. 03/01/91 21 3 93 n.a. n.a. 03/15/58 13 4 91 n.a. n.a. 01/07/52 9 *5 89 n.a. n.a. 02/14/95 7 6 81 n.a. n.a. 01/29/80 6 7 79 n.a. n.a. 02/20/80 5 436

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

18 78 n.a. n.a. 03/06/95 4 9 76 n.a. n.a. 12/18/78 4 10 70 n.a. n.a. 01/16/79 3 11 69 n.a. n.a. 03/02/52 3 12 68 n.a. n.a. 03/08/92 3 13 64 n.a. n.a. 12/01/82 3 14 63 n.a. n.a. 11/12/85 2 15 61 n.a. n.a. 03/26/91 2 16 59 n.a. n.a. 12/23/82 2 17 58 n.a. n.a. 03/03/78 2 18 58 n.a. n.a. 12/28/92 2 19 57 n.a. n.a. 01/09/75 2 20 54 n.a. n.a. 01/05/95 2

BRIGHT ANGEL RANGER STATION SUMMER DAILY PRECIPITATION

1 108 n.a. n.a. 08/29/51 81 2 80 n.a. n.a. 10/31/92 29 3 56 n.a. n.a. 08/24/88 18 t4 51 n.a. n.a. 07/25/83 13 5 51 n.a. n.a. 08/27/53 10 6 50 n.a. n.a. 10/20/79 8 7 t7 47 n.a. n.a. 08/17/63 8 45 n.a. n.a. 07/03/61 6 9 44 n.a. n.a. 08/31/92 5 10 41 n.a. n.a. 09/30/83 5 11 39 n.a. n.a. 09/20/52 4 12 39 n.a. n.a. 08/12/50 4 t13 38 n.a. n.a. 08/02/63 4 14 38 n.a. n.a. 10/30/87 3 t15 38 n.a. n.a. 08/27/87 3 16 38 n.a. n.a. 09/09/80 3 17 38 n.a. n.a. 08/12/56 3 18 37 n.a. n.a. 10/09/60 3 2 t19 37 n.a. n.a. 08/18/94 20 36 n.a. n.a. 10/28/74 2 437

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

GRAND CANYON WINTER STORMS

"H. 118 12/03/66 12/07/66 n.a. 157 2 107 11/21/05 11/28/05 n.a. 56 3 104 03/01/70 03/03/70 n.a. 34 4 104 12/26/06 01/01/07 n. a. 25 5 103 12/13/08 12/17/08 n.a. 19 6 95 02/10/27 02/17/27 n. a. 16 7 94 12/01/06 12/05/06 n.a. 13 8 86 02/19/13 02/27/13 n.a. 12 9 86 11/11/78 11/12/78 n.a. 10 10 78 12/29/51 01/01/52 n.a. 9 11 75 03/01/78 03/05/78 n.a. 8 12 72 12/22/45 12/26/45 n.a. 8 13 70 11/04/05 11/08/05 n.a. 7 14 70 11/26/19 11/27/19 n.a. 6 15 68 01/06/93 01/11/93 n.a. 6 16 68 12/17/78 12/20/78 n.a. 6 17 66 11/21/06 11/27/06 n.a. 5 18 65 12/26/36 01/02/37 n.a. 5 19 64 02/17/80 02/21/80 n.a. 5 20 61 01/07/40 01/13/40 n.a. 4

GRAND CANYON SUMMER STORMS

ti 140 09/04/39 09/08/39 n.a. 158 2 115 10/02/07 10/06/07 n.a. 57 3 113 09/02/07 09/05/07 n.a. 34 t4 91 08/17/89 08/19/89 n.a. 25 5 89 07/30/04 07/31/04 n.a. 19 *6 86 07/23/83 07/25/83 n.a. 16 7 81 10/15/72 10/21/72 n.a. 13 8 76 08/25/04 08/31/04 n.a. 12 9 74 10/01/12 10/05/12 n.a. 10 §10 73 09/10/39 09/13/39 n.a. 9 11 72 09/28/37 09/29/37 n.a. 8 438

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

112 69 08/18/84 08/22/84 n. a. 8 13 68 08/28/51 08/29/51 n.a. 7 14 67 09/05/81 09/08/81 n.a. 7 15 66 08/21/07 08/24/07 n.a. 6 16 65 09/16/25 09/19/25 n.a. 6 17 65 08/26/53 08/28/53 n.a. 5 18 64 09/26/11 09/30/11 n.a. 5 19 64 07/20/68 07/28/68 n.a. 5 20 64 07/23/04 07/26/04 n.a. 5

GRAND CANYON WINTER DAILY PRECIPITATION

1 101 n.a. n.a. 03/01/70 157 2 50 n.a. n.a. 11/11/78 56 3 48 n.a. n.a. 12/27/23 34 4 45 n.a. n.a. 02/29/60 25 5 44 n.a. n.a. 11/26/19 19 t6 43 n.a. n.a. 12/05/66 16 7 42 n.a. n.a. 12/30/51 13 8 41 n.a. n.a. 12/18/78 12 9 39 n.a. n.a. 12/22/45 10 10 38 n.a. n.a. 12/06/24 9 11 38 n.a. n.a. 01/17/14 8 12 38 n.a. n.a. 11/20/13 8 113 37 n.a. n.a. 03/06/95 7 14 37 n.a. n.a. 11/12/78 6 15 36 n.a. n.a. 12/25/42 6 16 36 n.a. n.a. 02/19/90 6 17 36 n.a. n.a. 03/14/44 5 18 36 n.a. n.a. 11/21/05 5 19 34 n.a. n.a. 02/20/93 5 20 34 n.a. n.a. 01/31/22 4

GRAND CANYON SUMMER DAILY PRECIPITATION 439

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

fl- 80 n.a. n.a. 07/25/83 157 2 69 n.a. n.a. 09/29/37 57 3 59 n.a. n.a. 09/06/81 34 4 56 n.a. n.a. 08/29/51 25 5 51 n.a. n.a. 10/06/16 19 6 51 n.a. n.a. 09/03/07 16 7 51 n.a. n.a. 07/30/04 13 8 49 n.a. n.a. 09/17/25 12 9 48 n.a. n.a. 09/04/07 10 10 46 n.a. n.a. 10/31/51 9 tu 46 n.a. n.a. 08/19/89 8 t12 43 n.a. n.a. 08/22/92 8 13 42 n.a. n.a. 08/03/74 7 §14 42 n.a. n.a. 09/05/39 7 15 41 n.a. n.a. 07/18/50 6 16 41 n.a. n.a. 08/25/40 6 17 40 n.a. n.a. 10/04/11 5 18 39 n.a. n.a. 07/25/74 5 19 38 n.a. n.a. 10/06/24 5 20 38 n.a. n.a. 07/31/04 5

MOUNT TRUMBULL WINTER STORMS

1 94 12/19/21 12/25/21 n.a. 94 2 71 02/11/27 02/17/27 n.a. 34 t3 57 12/06/66 12/07/66 n.a. 21 4 56 11/12/46 11/15/46 n.a. 15 5 55 02/08/20 02/11/20 n.a. 12 6 52 12/27/36 12/29/36 n.a. 9 7 52 02/08/32 02/10/32 n.a. 8 47 03/01/38 03/05/38 n. a. 7 9 44 02/03/28 02/05/28 n.a. 6 10 42 02/17/71 02/18/71 n.a. 5 11 40 12/04/26 12/08/26 n. a. 5 12 39 03/09/43 03/11/43 n.a. 5 13 39 03/21/54 03/25/54 n.a. 4 14 38 12/09/65 12/13/65 n.a. 4 440

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

15 36 02/11/36 02/16/36 n.a. 4 16 36 12/04/47 12/08/47 n.a. 3 17 32 02/13/54 02/14/54 n.a. 3 18 32 12/29/51 12/31/51 n.a. 3 19 32 12/23/40 12/25/40 n.a. 3 20 31 02/09/76 02/10/76 n.a. 3

MOUNT TRUMBULL SUMMER STORMS

1- 1 112 07/24/55 07/25/55 n.a. 93 §2 98 09/10/39 09/12/39 n.a. 33 §3 94 09/04/39 09/06/39 n.a. 20 4 88 10/04/25 10/06/25 n.a. 15 5 66 07/29/76 07/30/76 n.a. 11 6 66 08/23/61 08/24/61 n.a. 9 *7 62 07/25/39 07/28/39 n.a. 8 8 60 08/05/41 08/06/41 n. a. 7 9 57 08/03/48 08/06/48 n.a. 6 10 56 08/08/30 08/11/30 n.a. 5 11 55 08/11/50 08/12/50 n.a. 5 12 54 08/22/49 08/23/49 n.a. 4 13 51 10/26/74 10/28/74 n.a. 4 14 50 07/23/41 07/24/41 n. a. 4 15 50 07/26/26 07/28/26 n.a. 4 16 49 10/15/72 10/22/72 n.a. 3 17 48 08/13/35 08/15/35 n.a. 3 18 48 07/23/67 07/24/67 n.a. 3 19 47 09/16/40 09/17/40 n.a. 3 20 47 10/06/33 10/11/33 n.a. 3

MOUNT TRUMBULL WINTER DAILY PRECIPITATION

1 38 n.a. n.a. 12/31/71 94

1- 2 38 n.a. n.a. 12/06/66 34 3 37 n.a. n.a. 03/03/38 20 4 36 n.a. n.a. 12/28/36 15 441

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

5 34 n.a. n.a. 01/27/49 11 6 33 n.a. n.a. 01/20/21 9 7 32 n.a. n.a. 11/13/57 8 8 30 n.a. n.a. 12/27/46 7 9 28 n.a. n.a. 12/30/51 6 10 28 n.a. n.a. 12/20/21 5 11 28 n.a. n.a. 03/10/43 5 12 27 n.a. n.a. 03/31/69 5 13 27 n.a. n.a. 12/09/65 4 14 27 n.a. n.a. 11/13/46 4 15 26 n.a. n.a. 02/14/54 4 16 26 n.a. n.a. 02/10/20 3 17 26 n.a. n.a. 02/04/28 3 18 25 n.a. n.a. 02/05/48 3 19 25 n.a. n.a. 02/16/27 3 20 25 n.a. na. 12/19/21 3

MOUNT TRUMBULL SUMMER DAILY PRECIPITATION

ti 111 n.a. n.a. 07/24/55 93 2 73 n.a. n.a. 09/18/20 33 3 62 n.a. n.a. 07/29/76 20 §4 52 n.a. n.a. 09/06/39 15 5 51 n.a. n.a. 08/24/61 11 6 49 n.a. n.a. 07/23/41 9 7 47 n.a. n.a. 08/29/51 8 7 §8 47 n.a. n.a. 09/12/39 9 47 n.a. n.a. 08/12/50 6 10 45 n.a. n.a. 08/01/43 5 11 45 n.a. n.a. 09/17/40 5 12 44 n.a. n.a. 08/23/49 4 13 44 n.a. n.a. 10/10/60 4 14 43 n.a. n.a. 09/06/69 4 15 43 n.a. n.a. 07/26/26 4 16 42 n.a. n.a. 10/05/25 3 17 42 n.a. n.a. 08/18/20 3 §18 42 n.a. n.a. 09/11/39 3 442

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

19 40 n.a. n.a. 10/04/25 3 20 39 n.a. n.a. 10/01/21 3

PHANTOM RANCH WINTER STORMS

1 75 n.a. 11/11/78 11/12/78 55 2 64 n.a. 01/18/80 01/19/80 20 t3 53 n.a. 12/03/66 12/07/66 12 4 52 n.a. 01/06/93 01/11/93 9 5 50 n.a. 02/17/80 02/21/80 7 6 46 n.a. 12/17/78 12/19/78 6 7 39 n.a. 01/04/95 01/08/95 5 8 38 n.a. 03/26/91 03/28/91 4 03/01/95 03/06/95 4 1. 9 38 n.a. 10 36 n.a. 03/01/78 03/05/78 3 11 36 n.a. 11/27/81 11/30/81 3 12 34 n.a. 02/19/93 02/21/93 3 13 33 n.a. 11/25/85 11/30/85 2 14 31 n.a. 01/09/80 01/11/80 2 15 31 n.a. 01/13/93 01/20/93 2 16 31 n.a. 12/05/94 12/07/94 2 17 30 n.a. 02/02/88 02/03/88 2 18 29 n.a. 03/17/83 03/21/83 2 19 28 n.a. 12/28/92 12/30/92 2 20 27 n.a. 12/04/72 12/05/72 2

PHANTOM RANCH SUMMER STORMS t1 62 n.a. 08/18/84 08/23/84 53 t2 62 n.a. 07/21/83 07/26/83 19 3 43 n.a. 10/15/72 10/19/72 12 t4 40 n.a. 08/16/89 08/20/89 8 5 39 n.a. 09/06/81 09/08/81 6 6 39 n.a. 08/10/81 08/11/81 5 7 38 n.a. 10/14/94 10/17/94 5 8 38 n.a. 09/07/75 09/13/75 4 443

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

9 37 n.a. 09/23/76 09/27/76 3 10 35 n.a. 08/23/82 08/26/82 3 11 35 n.a. 07/16/76 07/18/76 3 12 34 n.a. 09/06/67 09/07/67 3 t13 33 n.a. 08/13/90 08/16/90 2 14 33 n.a. 09/11/69 09/12/69 2 15 32 n.a. 10/20/79 10/21/79 2 t16 30 n.a. 08/18/93 08/21/93 2 17 27 n.a. 10/03/72 10/04/72 2 18 27 n.a. 09/09/80 09/10/80 2 19 26 n.a. 10/02/81 10/05/81 2 20 25 n.a. 08/21/92 08/23/92 2

PHANTOM RANCH WINTER DAILY PRECIPITATION

1 47 n.a. n.a. 11/11/78 55 2 35 n.a. n.a. 12/18/78 20 3 33 n.a. n.a. 02/19/80 12 4 33 n.a. n.a. 01/18/80 9 5 33 n.a. n.a. 02/19/90 7 6 30 n.a. n.a. 01/19/80 6 7 28 n.a. n.a. 11/12/78 5 t8 26 n.a. n.a. 12/04/66 4 9 25 n.a. n.a. 03/26/91 4 10 25 n.a. n.a. 01/05/95 3 11 24 n.a. n.a. 12/28/92 3 12 24 n.a. n.a. 02/02/88 3 13 24 n.a. n.a. 02/15/86 2 14 23 n.a. n.a. 03/01/91 2 15 23 n.a. n.a. 11/12/85 2 16 22 n.a. n.a. 03/18/83 2 17 22 n.a. n.a. 02/20/93 2 18 22 n.a. n.a. 11/30/82 2 19 21 n.a. n.a. 01/08/93 2 20 21 n.a. n.a. 12/31/91 2 444

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

PHANTOM RANCH SUMMER DAILY PRECIPITATION t 1 47 n.a. n.a. 07/25/83 53 2 43 n.a. n.a. 09/01/85 19 3 33 n.a. n.a. 09/06/67 12 4 33 n.a. n.a. 09/12/69 8 5 32 n.a. n.a. 10/15/94 6 t6 32 n.a. n.a. 08/18/84 5 7 26 n.a. n.a. 10/05/66 4 t8 25 n.a. n.a. 08/20/93 4 9 25 n.a. n.a. 09/09/80 3 10 23 n.a. n.a. 08/11/81 3 11 23 n.a. n.a. 09/06/81 3 t12 23 n.a. n.a. 08/19/89 3 t13 22 n.a. n.a. 08/20/84 2 14 22 n.a. n.a. 10/06/93 2 15 22 n.a. n.a. 10/04/72 2 16 21 n.a. n.a. 10/02/81 2 17 21 n.a. n.a. 09/08/75 2 18 20 n.a. n.a. 10/11/86 2 19 19 n.a. n.a. 09/25/67 2 20 19 n.a. n.a. 08/25/82 2

SELIGMAN WINTER STORMS

1 99 12/13/67 12/16/67 n.a. 151 2 72 03/07/68 03/09/68 n.a. 54 3 69 02/14/80 02/22/80 n.a. 33 4 69 02/23/44 02/29144 n.a. 24 5 64 11/11/78 11/12/78 n.a. 19 6 64 11/05/05 11/06/05 n.a. 15 7 59 03/01/70 03/02/70 n.a. 13 8 57 03/01/78 03/06/78 n.a. 11 !9 54 02/08/93 02/10/93 n.a. 10 10 49 01/23/49 01/26/49 n.a. 9 11 47 02/15/27 02/17/27 n.a. 8 12 47 12/20/21 12/23/21 n.a. 7 445

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

13 46 02/02/08 02/04/08 n.a. 7 14 45 03/01/91 03/02/91 n.a. 6 15 45 03/22/54 03/25/54 n.a. 6 16 45 01/13/93 01/17/93 n.a. 5 17 44 12/09/65 12/10/65 n.a. 5 18 43 12/17/78 12/20/78 n.a. 5 t19 42 12/05/66 12/07/66 n.a. 5 20 42 01/23/44 01/27/44 n.a. 4

SELIGMAN SUMMER STORMS

1 146 07/19/70 07/22/70 n.a. 147 §2 112 09/03/39 09/06/39 n.a. 53 3 94 07/22/15 07/25/15 n.a. 32 4 91 08/28/51 08/29/51 n.a. 23 5 86 10/01/21 10/02/21 n.a. 18 6 86 08/20/21 08/21/21 n.a. 15 7 85 09/23/83 09/24/83 n.a. 13 8 75 10/04/25 10/05/25 n.a. 11 9 74 09/05/24 09/10/24 n.a. 10 10 72 07/31/64 08/01/64 n.a. 9 11 71 08/06/21 08/07/21 n.a. 8 t12 69 08/23/55 08/25/55 n. a. 7 *13 68 08/03/39 08/04/39 n.a. 7 14 67 07/07/19 07/13/19 n.a. 6 t15 64 08/19/94 08/21/94 n.a. 6 t16 64 09/18/90 09/23/90 n.a. 5 17 64 09/04/07 09/05/07 n.a. 5 18 56 07/11/14 07/15/14 n.a. 5 §19 55 09/12/39 09/13/39 n.a. 4 20 55 08/10/81 08/15/81 n.a. 4

SELIGMAN WINTER DAILY PRECIPITATION

1 71 n.a. n.a. 03/07/68 151 2 55 n.a. n.a. 11/05/05 54 446

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

3 48 n.a. n.a. 11/11/78 33 4 47 n.a. n.a. 03/01/70 24 5 44 n.a. n.a. 12/31/15 19 6 38 n.a. n.a. 11/27/19 15 7 38 n.a. n.a. 02/23/43 13 8 38 n.a. n.a. 03/27/25 11 9 38 n.a. n.a. 01/18/16 10 10 37 n.a. n.a. 02/07/20 9 11 36 n.a. n.a. 12/16/67 8 12 35 n.a. n.a. 12/03/08 7 13 34 n.a. n.a. 12/10/65 7 14 33 n.a. n.a. 02/22/20 6 15 33 n.a. n.a. 11/07/63 6 16 32 n.a. n.a. 03/01/78 5 17 32 n.a. n.a. 12/21/67 5 18 31 n.a. n.a. 12/26/21 5 19 31 n.a. n.a. 03/12/18 5 20 30 n.a. n.a. 12/15/67 4

SELIGMAN SUMMER DAILY PRECIPITATION

1 123 n.a. n.a. 07/21/70 147 2 69 n.a. n.a. 10/05/25 53 3 60 n.a. n.a. 09/23/83 32 4 58 n.a. n.a. 10/02/21 23 5 56 n.a. n.a. 08/21/21 18 t6 56 n.a. n.a. 08/20/94 15 7 52 n.a. n.a. 08/28/51 13 8 51 n.a. n.a. 10/06/16 11 19 50 n.a. n.a. 08/03/39 10 10 48 n.a. n.a. 07/16/30 9 11 48 n.a. n.a. 07/15/86 8

1. 12 47 n.a. n.a. 08/23/55 7 13 46 n.a. n.a. 08/07/21 7 14 45 n.a. n.a. 07/31/64 6 15 44 n.a. n.a. 09/25/19 6 §16 43 n.a. n.a. 09/12/39 5 447

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

17 41 n.a. n.a. 08/31/09 5 t18 39 n.a. n.a. 07/25/39 5 19 39 n.a. n.a. 08/19/06 4 §20 39 n.a. n.a. 09/04/39 4

SUPAI WINTER STORMS t 1 40 12/03/66 12/07/66 n. a. 55 2 33 11/14/78 11/15/78 n.a. 20 3 33 11/04/59 11/05/59 n.a. 12 4 32 03/01/70 03/03/70 n.a. 9 5 29 02/06/65 02/07/65 n.a. 7 6 28 12/09/65 12/15/65 n. a. 6 7 27 12/17/78 12/19/78 n.a. 5 8 25 12/27/84 12/29/84 n.a. 4 9 25 02/07/66 02/08/66 n.a. 4 10 25 11/27/81 11/29/81 n.a. 3 11 25 11/06/81 11/07/81 n.a. 3 12 23 11/01/74 11/03/74 n.a. 3 13 22 01/09/80 01/15/80 n.a. 2 14 22 03/01/78 03/03/78 n.a. 2 15 22 03/21/83 03/25/83 n.a. 2 16 22 01/10/78 01/11/78 n.a. 2 17 21 12/04/72 12/05/72 n.a. 2 18 21 03/26/73 03/29/73 n.a. 2 19 21 03/17/83 03/19/83 n.a. 2 20 20 02/19/71 02/20/71 n.a. 2

SUPAI SUMMER STORMS

1 67 08/09/81 08/14/81 n.a. 55 2 53 07/21/70 07/23/70 n.a. 20 3 47 09/21/67 09/25/67 n.a. 12 t4 44 07/18/84 07/21/84 n.a. 9 5 41 08/21/60 08/22/60 n.a. 7 6 40 08/19/57 08/12/59 n.a. 5 448

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

8 34 10/17/72 10/20/72 n.a. 4 9 32 07/08/70 07/10/70 n.a. 4 10 31 07/30/56 07/31/56 n.a. 3 11 31 10/09/86 10/12/86 n.a. 3 12 30 10/20/79 10/21/79 n.a. 3 13 29 08/05/79 08/06/79 n.a. 2 14 29 08/14/58 08/16/58 n.a. 2 t 15 28 07/23/83 07/25/83 n.a. 2 16 28 09/04/58 09/08/58 n.a. 2 17 27 07/31/68 08/01/68 n.a. 2 18 27 10/01/81 10/03/81 n. a. 2 19 26 09/25/76 09/26/76 n.a. 2 20 26 08/02/64 08/06/64 n.a. 2

SUPAI WINTER DAILY PRECIPITATION

1 41 n.a. n.a. 12/19/67 55 2 40 n.a. n.a. 12/06/86 20 3 34 n.a. n.a. 11/07/63 12 4 28 n.a. n.a. 12/27/79 9 5 26 n.a. n.a. 11/12/85 7 6 25 n.a. n.a. 03/22/58 6 7 24 n.a. n.a. 11/07/81 5 8 23 n.a. n.a. 02/14/87 4 9 23 n.a. n.a. 12/27/84 4 10 22 n.a. n.a. 03/01/70 3 11 20 n.a. n.a. 11/14/78 3 12 19 n.a. n.a. 02/06/65 3 13 18 n.a. n.a. 02/08/66 2 14 18 n.a. n.a. 11/05/59 2 15 17 n.a. n.a. 12/05/72 2 16 17 n.a. n.a. 02/04/58 2 17 17 n.a. n.a. 11/02/74 2 18 17 n.a. n.a. 03/18/83 2 19 16 n.a. n.a. 02/09/76 2 20 16 n.a. n.a. 12/02/78 2 449

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

SUPAI SUMMER DAILY PRECIPITATION

1 60 n.a. n.a. 07/30/76 55 2 40 n.a. n.a. 08/21/60 20 3 37 n.a. n.a. 08/11/59 12 4 36 n.a. n.a. 08/18/69 9 5 34 n.a. n.a. 08/10/81 7 6 32 n.a. n.a. 07/04/75 6 7 31 n.a. n.a. 07/29/69 5 8 29 n.a. n.a. 09/13/64 4 9 28 n.a. n.a. 08/05/79 4 10 27 n.a. n.a. 08/20/57 3 11 27 n.a. n.a. 10/31/57 3 12 27 n.a. n.a. 10/20/79 3 13 25 n.a. n.a. 07/11/85 2 14 25 n.a. n.a. 07/21/70 2 15 25 n.a. n.a. 07/27/56 2 16 24 n.a. n.a. 08/14/80 2 17 23 n.a. n.a. 09/24/67 2 18 23 n.a. n.a. 10/11/86 2 19 22 n.a. n.a. 09/06/81 2 20 22 n.a. n.a. 08/24/80 2

TUWEEP RANGER STATION WINTER STORMS t 1 157 12/06/66 12/07/66 n.a. 63 2 97 12/12/65 12/13/65 n.a. 22 3 65 12/29/51 12/31/51 n.a. 14 4 48 03/21/54 03/25/54 n.a. 10 5 48 02/20/80 02/21/80 n.a. 8 6 45 03/05/78 03/06/78 n.a. 6 7 37 03/09/75 03/11/75 n.a. 5 8 36 11/11/85 11/13/85 n.a. 5 9 35 03/01/78 03/03/78 n.a. 4 in 'u nvi 9/7A (11/72/70 n a A 450

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

12 31 01/25/54 01/26/54 n.a. 3 13 30 12/27/84 12/28/84 n.a. 3 14 30 02/09/76 02/10/76 n.a. 3 15 29 11/27/81 11/28/81 n.a. 2 16 29 02/14/54 02/15/54 n.a. 2 17 28 02/17/80 02/18/80 n.a. 2 18 28 03/01/52 03/02/52 n.a. 2 19 27 01/29/80 01/30/80 n. a. 2 20 27 11/11/78 11/12/78 n.a. 2

TUWEEP RANGER STATION SUMMER STORMS

1 104 08/23/82 08/26/82 n.a. 67 t2 91 07/18/84 07/23/84 n.a. 24 3 73 08/03/48 08/04/48 n.a. 15 4 73 07/18/69 07/19/69 n.a. 11 5 54 07/14/53 07/18/53 n.a. 8 6 51 08/01/51 08/03/51 n.a. 7 7 47 09/25/76 09/27/76 n.a. 6 8 47 07/20/70 07/25/70 n. a. 5 9 46 07/31/68 08/01/68 n.a. 4 10 46 08/19/70 08/20/70 n.a. 4 *11 39 07/24/83 07/25/83 n.a. 4 12 39 08/19/57 08/21/57 n.a. 3 13 36 09/20/52 09/21/52 n.a. 3 14 35 08/11/79 08/13/79 n.a. 3 15 35 08/29/61 08/30/61 n.a. 3 16 34 09/11/69 09/13/69 n.a. 2 17 33 10/20/57 10/21/57 n.a. 2 18 33 08/14/67 08/18/67 n.a. 2 19 33 07/16/59 07/18/59 n.a. 2 20 32 07/24/76 07/25/76 n.a. 2

TUWEEP RANGER STATION WINTER DAILY PRECIPITATION t 1 96 n.a. n.a. 12/06/66 62 451

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (him) From To Total Interval (yrs)

2 88 n.a. n.a. 12/12/65 22 t3 61 n.a. n.a. 12/07/66 14 4 55 n.a. n.a. 12/30/51 10 5 51 n.a. n.a. 12/19/78 8 6 46 n.a. n.a. 12/31/72 6 7 43 n.a. n.a. 01/25/69 5 8 38 n.a. n.a. 03/05/78 5 9 36 n.a. n.a. 02/20/80 4 10 34 n.a. n.a. 03/02/70 4 11 34 n.a. n.a. 01/26/56 3 12 32 n.a. n.a. 03/22/58 3 13 30 n.a. n.a. 03/24/64 3 14 30 n.a. n.a. 01/19/80 3 15 30 n.a. n.a. 02/06/76 2 16 28 n.a. n.a. 02/09/76 2 17 28 n.a. n.a. 03/25/50 2 18 27 n.a. n.a. 11/06/60 2 19 27 n.a. n.a. 11/02/74 2 20 27 n.a. n.a. 01/25/54 2

TUWEEP RANGER STATION SUMMER DAILY PRECIPITATION

1 67 n.a. n.a. 07/19/69 67 2 62 n.a. n.a. 08/29/51 24 3 58 n.a. n.a. 09/24/67 15 t4 53 n.a. n.a. 07/20/84 10 5 46 n.a. n.a. 08/23/82 8 6 46 n.a. n.a. 08/03/48 7 7 44 n.a. n.a. 08/26/82 6 8 42 n.a. n.a. 07/18/50 5 9 39 n.a. n.a. 09/24/58 4 10 39 n.a. n.a. 08/10/81 4 11 38 n.a. n.a. 10/02/81 4 12 38 n.a. n.a. 08/19/70 3 t13 38 n.a. n.a. 07/25/83 3 14 37 n.a. n.a. 07/31/68 3 15 36 n.a. n.a. 10/20/79 3 452

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

16 36 n.a. n.a. 09/25/76 2 17 36 n.a. n.a. 08/26/53 2 18 35 n.a. n.a. 10/31/57 2 119 34 n.a. n.a. 08/25/55 2 20 32 n.a. n.a. 07/18/61 2

WILLIAMS WINTER STORMS

1 183 02/13/80 02/22/80 n.a. 159 2 147 02/19/93 02/21/93 n.a. 57 3 145 11/22/65 11/26/65 n.a. 35 4 136 12/27/36 01/01/37 n.a. 25 5 134 02/11/27 02/16/27 n.a. 19 t6 132 12/03/66 12/07/66 n.a. 16 7 130 02/04/76 02/10/76 n.a. 14 8 121 12/08/65 12/17/65 n.a. 12 9 120 02/19/20 02/22/20 n.a. 10 10 119 01/16/17 01/22/17 n.a. 9 11 110 03/11/18 03/13/18 n.a. 8 12 107 01/16/16 01/20/16 n.a. 8 13 106 02/17/17 02/19/17 n.a. 7 14 105 12/29/51 01/01/52 n.a. 7 15 103 02/07/32 02/11/32 n.a. 6 16 101 11/01/57 11/04/57 n.a. 6 17 100 02/01/05 02/06/05 n.a. 5 18 99 01/09/05 01/12/05 n.a. 5 19 99 01/10/30 01/13/30 n.a. 5 20 98 12/14/08 12/17/08 n.a. 5

WILLIAMS SUMMER STORMS

1 193 07/08/19 07/24/19 n.a. 160 2 143 10/15/72 10/21/72 n.a. 58 §3 142 09/04/39 09/07/39 n.a. 35 4 123 10/01/16 10/06/16 n.a. 25 5 123 07/22/76 07/30/76 n.a. 20 453

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

6 110 08/07/46 08/12/46 n.a. 16 7 103 07/23/68 08/01/68 n. a. 14 t8 101 07/13/55 07/21/55 n.a. 12 9 100 07/31/92 08/06/92 n.a. 10 10 99 08/28/51 08/29/51 n.a. 9 11 98 09/22/83 09/24/83 n.a. 8 12 96 08/08/43 08/15/43 n.a. 8 13 92 09/17/65 09/20/65 n.a. 7 14 92 10/08/60 10/11/60 n.a. 7 t15 84 07/27/89 07/31/89 n.a. 6 16 82 08/04/71 08/06/71 n.a. 6 17 82 08/25/53 08/29/53 n.a. 5 18 81 08/02/48 08/08/48 n.a. 5 19 80 07/25/29 08/06/29 n. a. 5 20 80 09/13/41 09/15/41 n.a. 5

WILLIAMS WINTER DAILY PRECIPITATION

1 89 n.a. n.a. 02/20/93 158 2 84 n.a. n.a. 02/19/18 57 3 75 n.a. n.a. 12/30/51 35 4 70 n.a. n.a. 11/03/57 25 5 70 n.a. n.a. 02/19/17 19 6 70 n.a. n.a. 02/09/32 16 7 69 n.a. n.a. 02/09/76 14 8 66 n.a. n.a. 01/11/30 12 9 64 n.a. n.a. 12/28/92 10 10 63 n.a. n.a. 11/25/65 9 11 63 n.a. n.a. 03/08/18 8 12 62 n.a. n.a. 11/23/65 8 13 62 n.a. n.a. 12/16/08 7 14 61 n.a. n.a. 12/26/23 7 15 59 n.a. n.a. 03/13/18 6 16 58 n.a. n.a. 11/01/87 6 17 56 n.a. n.a. 02/19/93 5 18 56 n.a. n.a. 01/01/10 5 19 56 n.a. n.a. 02/14/80 5 454

APPENDIX H. Rankings and estimated probabilities for the 20 largest multiday storms and daily precipitation events at nine stations in northern Arizona - Continued

Date of Precipitation STORM DATES Highest Daily Recurrence Rank (mm) From To Total Interval (yrs)

20 56 n.a. n.a. 12/31/15 5

WTI ,LIAMS SUMMER DAILY PRECIPITATION

1 69 n.a. n.a. 09/18/65 160 2 63 n.a. n.a. 08/29/51 57 3 58 na. n.a. 10/06/93 35 4 58 n.a. n.a. 09/23/83 25 5 57 n.a. n.a. 07/01/80 20 6 56 n.a. n.a. 07/18/35 16 7 56 n.a. n.a. 10/19/72 14 §8 54 n.a. n.a. 09/05/39 12 9 52 n.a. n.a. 10/04/40 10 t 10 52 n.a. n.a. 08/04/63 9 11 51 n.a. n.a. 07/08/19 8 12 51 n.a. n.a. 07/31/92 8 13 51 n.a. n.a. 08/04/71 7 14 51 n.a. n.a. 07/18/46 7 15 50 n.a. n.a. 10/06/24 6 16 50 n.a. n.a. 07/20/16 6 17 49 n.a. n.a. 10/18/72 5 18 49 n.a. n.a. 10/05/25 5 t19 49 n.a. n.a. 09/18/63 5 20 49 n.a. n.a. 08/10/46 5

t, indicates relationship of precipitation event with debris flow(s) where date(s) is/are well known; t, indicates that precipitation event occurred in the month preceeding debris flow(s) for which the approximate date(s) is/are known; §, indicates that the precipitation event was generally associated with the 1939 debris flow at Prospect Creek. The exact date of the event is not known, but constrained to a period of about a 10 days when three tropical cyclones dissapated over western Grand Canyon; n.a., not applicable. 455

APPENDIX L Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon

Pre- Fan Ratio of 'Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above ume to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol./area) percent) Percentage Percentage  10% of Fan

2.8-R n.d. n.d. 79.9 K/T (59.8) Mz (35.2) 2 5.7-R 9 3.5 71.1 Ka (62.1) C (12.3) 3 7.9-R 11 4.1 80.3 KIT (43.7) C (39.1) 2 7.9-L 3 1.0 81.4 K/T (67.8) C (15.1) 3 8.6-R 14 4.9 n.d. n.d. n.d. n.d. 10.2-L 19 6.5 74.4 KIT (69.8) uk (10.0) 2 11.2-R 14 5.1 73.1 C (41.0) K/T (33.8) 2 11.8-L 11 4.3 87.0 KIT (66.0) C (16.0) 2 12.1-L 17 6.7 74.5 K/T (58.0) C (35.5) 2 12.8-R 19 7.1 n.d. n. d. n.d. n.d. 16.3-L 5 0.7 42.2 K/T (49.0) C (27.0) 3 16.8-R 1 n.d. 37.0 KIT (50.1) uk (22.2) 3 18.0-L 9 2.3 24.1 S (40.8) KIT (27.0) 3 19.0-R 12 4.1 75.5 S (49.0) uk (30.0) 2 19.3-R 12 3.8 65.8 S (52.0) Ka (24.9) 3 19.9-L 2 0.5 80.1 S (65.5) K/T (18.5) 2 20.5-R 2 0.5 73.8 S (68.7) KIT (16.2) 3 21.4-L 12 4.0 38.0 S (43.5) KIT (40.0) 2 21.5-R 11 4.1 n.d. n. d. n.d. n.d. 21.8-R 22 7.0 78.0 KIT (48.5) S (32.5) 3 22.9-L 10 3.6 34.3 S (50.0) C (25.0) 3 23.3-L 10 4.3 n.d. n. d. n . d. n.d. 23.5-L 33 11.2 71.5 C (43.5) KIT (28.5) 3 24.4-L 9 6.6 75.6 S (47.0) uk (29.5) 2 456

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 1 Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above ume to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol./area) percent) Percentage Percentage  10% of Fan

24.7-L 17 5.8 67.3 S (78.8) RIM (9.1) 2 25.0-L 10 3.3 52.5 KIT (43.0) S (24.5) 4 25.3-L 15 5.5 n.d. n. d. n.d. n.d. 25.3-R 13 3.9 n.d. n. d. n.d. n.d. 25.4-R 16 5.6 n.d. n . d. n.d. n.d. 29.2-R 5 1.5 67.6 S (43.5) R/M (39.8) 2 30.5-R 22 7.4 86.5 S (52.4) R/M (28.4) 2 31.6-R 5 1.2 80.4 S (58.3) R/M (23.0) 2 32.6-L 1 0.2 71.0 S (84.0) RIM (13.5) 2 34.2-R 3 1.2 69.5 S (60.0) RIM (23.5) 2 34.7-L 6 1.6 83.6 S (79.6) R/M (15.8) 2 35.2-L 6 2.1 68.1 S (85.6) R/M (9.1) 1 35.6-L 4 0.7 80.8 S (69.4) KIT (25.4) 2 36.0-L 2 0.7 96.0 S (57.5) R/M (36.5) 2 37.4-L 3 0.6 45.5 S (44.5) R/M (42.0) 2 37.7-L 4 0.8 79.5 S (51.5) R/M (23.0) 3 39.0-R 1 n.d. 83.5 R/M (53.5) S (44.5) 2 41.0-R 12 3.3 83.0 S (40.7) RIM (30.1) 3 41.3-R 10 2.7 29.7 R/M (60.5) S (24.4) 2 42.9-L 17 4.6 70.9 S (50.3) R/M (22.1) 2 43.2-L 5 1.3 80.9 S (55.3) K/T (21.7) 3 43.7-L 5 1.3 84.8 S (50.5) RIM (40.4) 2 44.6-L 5 1.1 77.4 S (45.2) R/M (23.5) 3 457

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 'Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above urne to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol./area) percent) Percentage Percentage  10% of Fan

44.8-L 14 4.5 52.5 S (39.0) K/T (18.5) 4 45.8-L 6 1.4 n.d. n.d. n.d. n.d. 46.7-R 9 2.3 n.d. n.d. n.d. n.d. 47.0-R 15 5.1 49.0 R/M (68.1) S (18.8) 2 48.5-R 14 4.6 n.d. n.d. n.d. n.d. 49.4-R 18 6.8 69.9 R/M (68.9) S (25.2) 2 49.6-L 13 4.5 78.4 S (41.0) RIM (27.8) 2 49.8-R 14 5.9 78.0 R/M (59.2) S (29.8) 2 50.8-L n.d. n.d. 61.7 R/M (52.5) S (43.0) 2 51.2-L 18 8.9 n.d. n.d. n.d. n.d. 51.7-R 15 7.4 23.5 R/M (61.0) S (24.3) 2 53.1-R 21 8.3 46.9 R/M (55.0) S (37.0) 2 53.5-R 7 2.4 n.d. n.d. n.d. n.d. 53.8-R 7 2.3 75.0 R/M (71.5) S (23.5) 2 54.0-R 16 5.3 75.5 R/M (67.2) S (24.0) 2 55.0-L n.d. n.d. 60.1 S (46.6) R/M (34.9) 2 54.5-R 9 3.1 n.d. n.d. n.d. n.d. 55.4-R 15 4.4 n.d. n.d. n.d. n.d. 56.0-R 23 9.3 49.2 R/M (47.1) S (29.0) 2 57.5-R 17 5.7 77.9 R/M (57.0) S (16.0) 2 57.8-R 18 7.1 n.d. n. d. n.d. n.d. 58.0-R 18 5.3 72.0 R/M (76.5) S (20.0) 2 58.5-L 16 5.8 n.d. n.d. n.d. n.d. 458

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 'Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above urne to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol./area) percent) Percentage Percentage  10% of Fan

58.8-L 9 2.7 n.d. n.d. n . d. n.d. 59.6-L 15 5.7 44.5 R/M (58.8) S (24.5) 2 59.6-R 3 0.9 84.6 RIM (80.8) S (13.0) 2 60.6-R 10 3.1 84.7 R/M (43.7) S (42.6) 2 61.1-R 9 3.3 91.0 R/M (60.0) S (25.5) 2 61.7-R 19 6.8 n.d. n.d. n . d. n.d. 61.9-L 11 3.9 n.d. n.d. n . d . n.d. 62.6-R 13 4.3 55.2 R/M (49.0) uk (26.0) 2 63.3-R 21 6.8 74.7 R/M (69.4) S (24.3) 2 64.5-L 12 3.5 n.d. n.d. n . d . n.d. 64.6-R 9 2.1 87.1 T (51.0) R/M (27.8) 2 69.6-R 17 4.5 13.3 1R/M (30.8) uk (23.0) 3 75.0-L 7 2.0 43.2 uk (32.4) RIM (28.5) 3 75.5-L 12 2.9 19.8 uk (30.0) S (20.0) 4 76.0-L 7 0.9 95.5 YP (47.7) RIM (31.5) 3 76.7-L 22 8.4 84.5 YP (76.0) R/M (20.5) 2 78.7-L 1 n.d. 31.8 YP (32.4) R/M (22.9) 4 85.7-R 7 2.0 65.6 OP (84.9) RIM (12.8) 2 87.2-L 17 4.7 75.5 OP (75.5) uk (15.0) 2 91.5-R 6 1.0 8.4 R/M (42.5) uk (25.5) 2 92.7-L 5 1.6 50.5 S (55.5) RIM (29.5) 3 93.5-L 12 3.7 84.2 R/M (68.3) S (10.8) 3 94.3-R 6 1.7 68.5 RIM (40.2) S (23.7) 4 459

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 1 Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above ume to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol ./area) percent) Percentage Percentage . .10% of Fan

95.0-L 8 2.3 11.5 R/M (59.4) S (34.8) 2 96.7-L 4 1.1 84.4 R/M (55.8) S (32.3) 2 103.9-R 6 2.2 n.d. n. d. n.d. n.d. 104.6-L 3 1.1 n.d. n. d. n.d. n.d. 108.6-R 8 2.5 81.5 S (44.0) R/M (37.0) 2 112.2-R 4 1.0 n.d. n. d. n.d. n.d. 115.8-R 14 5.5 87.1 RIM (59.5) S (27.5) 2 116.5-L 7 1.5 31.0 RIM (64.5) S (14.0) 2 116.8-L 18 7.1 69.9 RIM (88.0) uk (4.7) 1 119.0-R 6 2.0 88.6 R/M (57.1) T (20.8) 2 120.1-R 9 2.9 60.6 RJM (73.3) S (12.5) 2 121.7-L 5 1.7 86.0 RIM (85.5) S (13.5) 2 122.2-R 10 2.4 74.3 R/M (82.0) uk (10.5) 2 122.3-L 10 3.3 79.5 R/M (68.7) S (26.8) 2 122.7-L 12 4.5 57.7 R/M (67.3) S (15.3) 2 124.4-L n.d. n.d. 83.1 RIM (64.4) S (18.0) 2 125.0-L 17 5.5 80.8 R/M (50.7) S (40.6) 2 126.9-R 8 2.6 78.5 R/M (69.6) T (14.8) 2 129.0-L 5 1.5 70.0 R/M (54.5) S (26.5) 2 130.5-R 4 1.3 56.0 R/M (40.0) OP (28.0) 3 130.9-L 13 4.9 2.0 R/M (42.8) S (24.8) 3 133.0-L 9 3.0 87.5 R/M (35.2) YP (30.4) 3 133.8-R 5 2.2 0.0 uk (35.0) YP (23.0) 4 460

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 1 Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above urne to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol./area) percent) Percentage Percentage  10% of Fan

135.9-L 8 3.1 n.d. n.d. n.d. n.d. 139.1-R 12 4.4 83.7 R/M (73.4) S (17.8) 2 139.5-R 37 13.6 90.1 R/M (47.4) S (33.3) 3 139.9-L 12 4.8 81.9 R/M (91.5) uk (6.5) 1 145.6-L 6 1.5 n.d. n.d. n.d. n.d. 149.7-R 5 1.5 n.d. n.d. n.d. n.d. 164.5-R 4 1.0 13.7 RIM (52.4) S (27.3) 2 166.4-L 6 1.6 76.3 RIM (78.7) S (17.7) 2 168.0-R 6 2.0 75.4 RIM (72.0) S (16.7) 2 171.1-R 7 2.3 85.1 RIM (82.8) S (12.6) 2 171.5-L 9 3.6 n.d. n.d. n.d. n.d. 174.4-R 8 2.9 n.d. n.d. n.d. n.d. 175.9-L 4 1.9 33.5 R/M (64.0) S (30.5) 2 176.4-R 18 6.5 87.9 R/M (92.3) uk (3.6) 1 177.7-L 22 7.3 56.5 R/M (67.0) BA (20.5) 2 179.4-L n.d. n.d. 87.1 B (89.6) RIM (5.2) 1 180.8-L n.d. n.d. 73.0 S (50.0) R/M (47.1) 2 183.1-L 13 5.0 n.d. n.d. n.d. n.d. 188.1-R 7 2.5 28.0 B (37.5) RIM (31.1) 3 189.7-L 13 5.5 65.9 RIM (53.0) S (35.5) 2 192.8-L 14 5.4 64.4 RIM (81.1) S (8.1) 1 198.5-L 13 5.0 64.7 RIM (59.8) S (27.0) 2 202.0-R 6 2.1 70.5 RIM (46.5) 5 (37.5) 2 461

APPENDIX I. Topographic, particle-size and lithologic data collected from a subset of pre-dam, river-reworked alluvial fans between river miles 0 and 225 of the Colorado River in Grand Canyon - Continued

Pre- Fan Ratio of 1Surface Primary Secondary Number of Dam Height Fan Vol- Clasts Clast Clast Bedrock Fan, Above urne to >256 Lithology Lithology Lithologies (Mile- 850 m3/s, Plan Area, mm, (in and and Comprising Side) (in meters) (vol ./area) percent) Percentage Percentage  10% of Fan

205.5-L 10 3.8 92.7 RIM (53.7) S (46.3) 2 211.2-L 9 3.9 n.d. n.d. n.d. n.d. 214.2-R 5 1.7 59.4 RIM (70.2) uk (22.0) 2 215.0-L 9 3.7 78.4 R/M (78.4) T(13.8) 2 217.4-L 14 4.5 80.9 OP (65.5 RIM (23.8) 2 219.4-L 5 1.2 88.1 RIM (89.1) T (5.9) 1 221.3-R 12 3.3 84.0 RIM (96.5) S (2.0) 1 223.5-L 5 1.4 67.0 OP (56.5) T (11.3) 2

1 , based on point-count data collected near the approximate elevation of the 2,800 m 3/s discharge; Mz, undiffernetiated Mesozoic units; KIT, Kaibab and Toroweap Limestones; C, Coconino Sandstone; H, Hermit Shale; S, Supai Group sandstones and shales; RIM, Redwall and Muav Limestones; BA, Bright Angel Shale; T, Tapeats Sandstone; YP, younger precambrian sandstones, shales and basalts; OP, Older Precambrian gneisses and schists; B, basalt; uk, unknown lithologic source; n.d., no data. 462

PLATE 1 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 1: River Miles 0.0 to 11.3 (after Schmidt and Graf, 1990)

11/40

36519

36.40 00'

Detailed Debris Flow Study Site 11110 00- Geomorphically Significant Drainage Area Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994) 2 4 6 8 10 km Colorado River

RM 11.3 River mileage at confluence Transverse Mercator Projection 463

PLATE 2 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 2: River Miles 11.3 to 22.6 (after Schmidt and Graf, 1990)

Salt Water Wash FtM11.81.

House Rock Wash RM16.0it

North Canyon RM 20.5R

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Area

Drainage Area that does not meet criteria for geomorphically significant tributary Neils and others, 1994) 5 10 15 20 km Colorado River mileage at confluence RM 22.6 River Transverse Mercator Projection 464

PLATE 3 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 3: River Miles 22.6 to 35.9 (after Schmidt and Graf, 1990)

111'40 00"

RM 22.9L RA425..*

RA4 26.65 rY.4

,36'30 00'

112 110 00"

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Area Drainage Area that does not meet ctiteria for geomorphically significant tributary (Melis and others, 1994) 5 10 15 20 Lan Colorado River

I 5A4 3.5.9 River mileage at confluence Transverse Mercator Projection 465

PLATE 4 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 4: River Miles 35.9 to 61.5 (after Schmidt and Graf, 1990)

into ow 17110 Or

36154. 0. 305 00' 4.416) mislir __vs Tatahatso Wash Buck farm am RM 37.4L Canyon Af RA4 41.0R P a. Saddle Canyon -„.-. Tatalwysa RM 4 7.0R Wash 011110 RM43.2L

FtM all 52.5L ff#Apt

-TO".

Nankowe_ap Canyon RM 52.2A 36-is 00' 36-15 - a P

4 RM ir. 57.5R .411 Kwagunt Creek Rm RM 56.0R 58.0R 401r411, RM 59.6R 60 Mile Canyon .....s...

Iwo oo- uno oo-

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Area

Drainage Area that does not meet criteria N for geomorphically significant tributary (Melis and others, 1994) 0 2 4 6 8 10 km was Colorado River RM 61.5 River mileage at confluence Transverse Mercator Projection 466

PLATE 5 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 5: River Miles 61.5 to 77.4 (after Schmidt and Graf, 1990)

361000'

Geomorphically Significant Drainage Area Drainage Area that does not meet criteria geomorphically significant tributary 36130 00- for (Melis and others, 1994)

2 4 6 River mileage at confluence Transverse Mercator Projection 467

PLATE 6 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 6: River Miles 77.4 to 115.7 (after Schmidt and Graf, 1990)

112170 OW 111210 OW

3610

- t

Detailed Debris Flow Study Site 1121)0 00' Geomorphically Significant Drainage Area

Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994) 5 10 15 20 km V41a'1113. Colorado River

RM /IS. 7 River mileage at confluence Transverse Mercator Projection 468

PLATE 7 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 7: River Miles 115.7 to 125.5 (after Schmidt and Graf, 1990) 1/2'30 00' 1 12t5 00-

3610 00'

Fossil Canyon RM 125 .OL

Royal Arch Creek RM 116.51.

36'10 00' 3610 00'

Detailed Debris Flow Study Site

Geomorphically Significant Drainage An2a

Drainage Area that does not meet criteria 1L25 00' for geomorphically significant tributary (Melis and others, 1994) 2 4 6 10 km trEat' Colorado River RM 125.5 River mileage at confluence Transverse Mercator Projection 469

PLATE 8 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 8: River Miles 125.5 to 140.0 (after Schmidt and Graf, 1990)

RA4 127 .6L

51.4 127.3L

RM 126.91.

Detailed Debris Flow Study Site

Geornorphically Significant Drainage Area

Drainage Area that does not meet criteria for geornorphically significant tributary (Melis and others, 1994) 0 7 4 6 8 LO Colorado River

RM 140.0 River mileage at confluence Transverse Mercator Projection 470

PLATE 9 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 9: River Miles 14-0.0 to 160.0 (after Schmidt and Graf, 1990)

122 40 111'50 00' 3615 00' 362,5 00'

46'75 00' 3615 00'

21210 00' 12250 00'

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Ann

Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994) o 2 4 6 10 km Colorado River

KM 160.0 River mileage at confluence Transverse Mercator Projection 471

PLATE 10 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 10: River Miles 160.0 to 213.9 (after Schmidt and Graf, 1990)

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Area

Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994) 5 10 15 20 0m Colorado River

RM213.9 River mileage at confluence Transverse Mercator Projection 472

PLATE 11 Colorado River Drainage Areas Grand Canyon National Park and Vicinity, Arizona Geomorphic Reach 11: River Miles 213.9 to 225.8 (after Schmidt and Graf, 1990) 11310 012" 113°30 OW

RM 215.7L Three Springs Trail Canyon RA/ 219.4R

c•en L.3.5.50 00' 220 Mile Canyon RM 220.0R

RM 125.35

113 °30 Oa'

Detailed Debris Flow Study Site

Geomorphically Significant Drainage Area

Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994) 200m 5 10 r Colorado River I RM225.al River mileage at confluence Transverse Mercator Projection Colorado River Tributaries Within Grand Canyon National Park and Vicinity. Arizona

36600'00'

Detailed Debris Flow Study Site Geomorphically Significant Drainage Area Drainage Area that does not meet criteria for geomorphically significant tributary (Melis and others, 1994)

o 10 20 30 40 km

Transverse Mercator Projection

113°30'00' 113°00'00" 122 63000' 112600'00' Drainage Areas that Experienced Debris Flows Between 1872 and 1996, Colorado River: Lees Ferry to Diamond Creek

Geomorphically — Significant Drainage Area in which more than one debris flow has been detected dining the period 1872 through 1996, UTAH Geomorphically — Significant Drdinage Area in which at least one debris flow has been detected during the period 1872 tluough 1996, Geomorphically — Significant Drainage Area with frequency data _ available but no evidence of debris flow since 1872, Geomorphically — Significant Drainage Area with no frequency data Drainage Area that does not meet criteria for geomorphically — significant tributary (see Mehs and others, 1994)

10 211 40 Km

Transverse Mercator Projection

113 630'0(r 113°00'00' 112'30'00' 112600'00' Colorado River Drainage Tributaries Showing The Probability Of Debris How _ Occurrance Per Century (After Griffiths et al., 1996)

Pro 441 .44

Kbeft.31

k. MAR

Wlolmws Now 04 VO.WI,

ILL

P- 06614 04 wor.,

76 a, NIB!

t. tia

iebns Flow Probability = 0-25%

Debris Flow Probability = 25 —50%

Debris Flow Probability = 50-75% 1111 Debris Flow Probability .75 —100% Cr•Mi WAS. River mileage ai confluence

Transverse Mercator Projection

BM"

lufour 17.1kur ,e1.61r 476

REFERENCES

Allen, J.R.L., 1965, A review of the origin and characteristics of recent alluvial sediments: Sedimentology, v. 5, P. 89-191. Andrade, E.R., and Sellers, W.D., 1988, El Niiio and its effect on precipitation in Arizona and western New Mexico: Journal of Climatology, v. 8, p.403-410. Anderson, R.S., 1977, A biography of Clarence Edward Dutton (1841-1912), 19th century geologist and geographer: unpublished masters thesis, 71 p.

Anderson, U.S., and Hussey, K.M., 1962, Alluvial fan development at Franklin Bluffs, Alaska: Iowa Academy of Sciences Proceeding, v. 69, p. 310-322.

Andrews, E.D., 1991, Sediment transport in the Colorado River Basin: in Colorado River ecology and dam management, Proceedings of a Symposium, May 24-25, 1990, Santa Fe, New Mexico, National Academy Press, Washington, DC, p. 54-74.

Anstey, R.L., 1965, Physical characteristics of alluvial fans: United States Army Natick Laboratory, Technical Report ES-20, 109 p.

, 1966, A comparison of alluvial fans in west Pakistan and the United States: Pakistan Geographical Review, v. 21, p. 14-20.

Apmann, R.P., 1973, Estimating discharge from superelevation in bends: Journal of the Hydraulics Division, Proceedings of the American Society of Civil Engineers, HY 1, p. 65-79.

Baker, V.R., and Costa, J.E., 1987, Flood power, in Mayer, L. and Nash, D. (editors), Catastrophic Flooding: London, Allen and Unwin, p. 1-21.

Birdseye, C.H., 1924, Plan and profile of the Colorado River from Lees Ferry, Arizona to Black Canyon, Arizona - Nevada, Virgin River, Nevada: U.S. Geological Survey topographic maps, 21 sheets (A-U) consisting of 14 plans (maps) and 7 profiles; scale 1:31,680.

Birdseye, C.H., and Moore, R.C., 1924, Boat voyage through the Grand Canyon of the Colorado: The Geographical Review, v. XIV, no. 2, p. 177-196.

Beatty, C.B., 1989, Great boulders I have known: Geology, v. 17, p. 349-352.

, 1963, Origin of alluvial fans, White Mountains, California and Nevada: Association of American Geographers Annals., v. 53, p. 516-535.

Beverage, J.P., and Culbertson, J.K., 1964, Hyperconcentrations of suspended sediment: American Society of Civil Engineers, Journal of the Hydraulics Division, v. 90, p. 117-126. 477

Birdseye, C.H., and Moore, R.C., 1924, A boat voyage through the Grand Canyon of the Colorado River: The Geographical Review, v. XIV, no. 2, P. 177-196. Blackwelder, Eliot, 1928, Mudflow as a geologic agent in semiarid mountains: Geological Society of America Bulletin, v. 39, p. 465-494.

Blair, T.C., and McPherson, J.G., 1995a, Alluvial fans and their natural distinction from river based on morphologic, hydraulic processes, sedimentary processes, and facies assemblages - reply: Journal of Sedimentary Research, v. A65, no. 3, p. 581-583. , 1995b, Alluvial fans and their natural distinction from river based on morphologic, hydraulic processes, sedimentary processes, and facies assemblages - reply: Journal of Sedimentary Research, v. A65, no. 4, p. 708-711. , 1994, Alluvial fans and their natural distinction from river based on morphologic, hydraulic processes, sedimentary processes, and facies assemblages: Journal of Sedimentary Research, v. A64, no. 3, p. 450-489.

, 1993, The Trollheim alluvial fan and facies model revisited - reply: Geological Society of America Bulletin, v. 105, p. 564-567.

1992, The Trollheim alluvial fan and facies model revisited: Geological Society of America Bulletin, v. 104, p. 762-769.

Blakey, R.C., 1990, Supai Group and Hermit Formation: in Beus, S.S., and Morales, M., editors, Grand Canyon Geology: New York, Oxford University Press, p. 147-182.

Blinn, D.W., Shannon, J.P., and Stevens, L.E., 1995, Consequences of fluctuating discharge for lotic communities: Journal of the North American Benthological Society, v. 14, p. 233-248.

Blissenbach, E., 1954, Geology of alluvial fans in semiarid regions: Bulletin of the Geological Society of America, v. 65, p. 175-190.

Bonney, C.T.G., 1902, Moraines and mudstreams in the Alps: Geological Magazine, v. 9, p. 8-16. Brunsden, D., and Prior, D.B., editors, 1984, Hillslope processes: New York, John Wiley and Sons, 620 p.

Bull, W.B., 1991, Geomorphic responses to climate change: New York, Oxford Press, 326 p. , 1977, The alluvial-fan environment: Progress in Physical Geography, v. 1, p. 222-270. 478

, 1964, Alluvial fans and near-surface subsidence in Western Fresno County California: Geological Survey Professional Paper 437-A, 71 p.

, 1963, Alluvial fan deposits in western Fresno County, California: Journal of Geology, v. 71, p. 243-251.

Campbell, R.H., 1975, Soil slips, debris flows, and rainstorms in the Santa Monica Mountains and vicinity, Southern California: U.S. Geological Survey Professional Paper #851,31 p.

Cannon, S.H., and Ellen, S.D., 1985, Rainfall conditions for abundant debris avalanches, San Francisco Bay region, California: California Geology, v. 38, no. 12, p. 267- 272.

Carlton, A.M., Carpenter, D.A., and Weser, P.J., 1990, Mechanisms of interannual variability of the southwest United States summer rainfall maximum: Journal of Climate, v. 3, p. 999-1015.

Cayan, D.R., and Webb, R.H., 1992, El Nirio/Southern Oscillation and streamflow in the western United States: p. 29-68, in H. Diaz and V. Markgraf (eds.), El Nino; Historical and Paleoclimatic Aspects of the Southern Oscillation, Cambridge University Press.

Chamberlin, T.C., 1890, The method of multiple working hypotheses: Science, v. 15, p. 92-96.

Chang, H.H., 1979, Minimum stream power and river channel patterns: Journal of Hydrology, v. 41, p. 303-307. Chawner, W.D., 1935, Alluvial fan flooding, the Montrose, California flood of 1934: Geographical Review, v. 25, p. 77-88.

Collier, M., Webb, R.H., and Schmidt, J.C., 1995, Dams and rivers - a primer on the downstream effects of dams: U.S. Geological Survey Circular #1126, 94 p.

Conover, W.J., 1980, Practical nonparametric statistics [2nd edition]: New York, John Wiley and Sons, 493 p.

Cooley, M.E., Aldridge, B.N., and Euler, R.C., 1977, Effects of the catastrophic flood of December, 1966, North Rim area, eastern Grand Canyon, Arizona: U.S. Geological Survey Professional Paper 980, 43 p.

Costa, J.E., 1984, Physical geomorphology of debris flows, in Costa, J.E., and Fleisher, P.J., editors, Developments and applications of geomorphology: Berlin, Springer- Verlag Publishing, p. 268-317.

Costa, J.E., and Jarrett, R.D., 1981, Debris flows in small mountain stream channels of Colorado and their hydrologic implications: Bulletin of the Association of 479

Engineering Geologists, v. 18, P. 309-322. Croft, A.R., 1962, Some sedimentation phenomena along the Wasatch Mountain front: Journal of Geophysical Research, v. 67, p. 1511-1524.

Curry, R.R., 1966, Observation of alpine mudflows in the Tenmile Range, central Colorado: Geological Society of America Bulletin, v.77, p 771-776.

Denny, C.S., 1967, Fans and pediments: American Journal of Science, v. 265, p. 81-105.

, 1965, Alluvial fans in the Death Valley region, California and Nevada: United States Geological Survey Professional Paper 466, 62 p.

Diaz, H., and Markgraf, V., 1992, El Nitio (Historical and Paleoclimatic Aspects of the Southern Oscillation): (eds.), Cambridge University Press, Cambridge, U.K., 476 p

Diaz, H.F., and Pulwarty, R.S., 1994, An analysis of the time scales of variability in centuries-long ENSO-sensitivity records in the last 1000 years: Climate Change, v. 26, p. 317-342.

Dietrich, W.E., Kirchner, J.W., Ikeda, H., and Iseya, F., 1989, Sediment supply and the development of the coarse surface layer in gravel-bedded rivers: Nature, v. 340, p. 215-217.

Dolan, R., Howard, A., and Trimble, D., 1978, Structural control of the rapids and pools of the Colorado River in the Grand Canyon: Science, v. 202, p. 629-631.

Dolan, R., Howard, A., and Gallenson, A., 1974, Man's impact on the Colorado River in the Grand Canyon: American Scientist, v. 62, p. 392-401.

Dutton, C.E., 1882, The Tertiary history of the Grand Cation district; with atlas. U.S. Geological Survey Monograph 2, 264 p.; atlas 23 sheets.

Eckis, R., 1928, Alluvial fans of the Cucamonga District, Southern California: Journal of Geology, v. 36, P. 224-247. Ellen, S.D., and Wieczorek, G.F., 1988, Landslides, floods, and marine effects of the storm January 3-5, 1982, in the San Francisco Bay region, California: U.S. Geological Survey Professional Paper # 1434.

Ely, L.L., Enzel, Y., and Cayan, D.R., 1994, Anomolous North Pacific atmospheric circulation and large winter floods in the Southwestern United States: Journal of Climate, v. 7, no. 6, p. 977-987.

Ely, L.L., Webb, R.H., and Enzel, Yehouda, 1992, Accuracy of post-bomb 137Cs and 14C in dating fluvial deposits: Quaternary Research, v. 38, p. 196-204. 480

Enos, Paul, 1977, Flow regimes in debris flow: Sedimentology, v. 24, p. 133-142.

Enzel, Y., Ely, L.L., House, P.K., Baker, V.R., and Webb, R.H., 1993, Paleoflood evidence for a natural upper bound to flood magnitudes in the Colorado River Basin: Water Resources Research, v. 29, no. 7, p. 2287-2297.

, 1971, Features of coarse-grained, high-concentration fluids and their deposits: Journal of Sedimentary Petrology, v. 41, n. 4, p. 916-927.

Folk, R.L., 1974, Petrology of sedimentary rocks: Austin, Texas, Hemphill Publishing, 182 p.

Ford, T.D., Huntoon. P.W., Billingsley, G.H., Breed, W.J., 1974, Rock movement and mass wastage in the Grand Canyon, in Breed, W.J., and E. Roat (editors), Geology of the Grand Canyon:, Museum of Northern Arizona Press, Flagstaff, p. 116-128.

Gibbons, J.D., 1985, Nonparametric methods for quantitative analysis (2nd edition): American Sciences Press, Inc., Syracuse, New York, 481 p.

Gilbert, G.K., 1876, The Colorado Plateau province as a field for geologic study: American Journal of Science, 3rd Series, v. 12, p. 16-24, 85-103.

Glancy, P.A., and Harmsen, L., 1975, A hydrologic assessment of the September 14, 1974, flood in Eldorado Canyon, Nevada: U.S. Geological Survey Professional Paper 930, p. 28.

Gore, J.A., and Shields, Jr., D.F., 1995, Can large rivers be restored?: BioScience, v. 45, no. 3, p. 142-152.

Graf, W.L., 1979, Rapids in canyon rivers: Journal of Geology, v. 87, p. 533-551. , 1980, The Effect of Dam Closure on Downstream Rapids: Water Resources Research, v. 16, no. 1, p. 129-136.

Graf, J.B., Webb, R.H., and Hereford, R., 1991, Relation of sediment load and floodplain formation to climatic variability, Paria River drainage basin, Utah and Arizona: Geological Society of America Bulletin, v. 103, p. 1405-1415.

Grand Canyon Natural History Association, 1976, Geologic map of the Grand Canyon National Park, Arizona: Grand Canyon, Arizona, Grand Canyon Natural History Association, scale 1:62,500.

Griffiths, P.G., Webb, R.H., and Melis, T.S., 1996, Initiation and frequency of debris flows in Grand Canyon, Arizona: U.S. Geological Survey Open-File Report 96- 491, 35 p.

Grove, A.T., 1953, Account of a mudflow on Bredon Hill, Worcestershire, April, 1951: 481

Geologists' Association Proceedings, v. 64, P. 11-13. Gunderson, L.H., Holling, C.S., and Light, S.S., 1995, Barriers and bridges to the renewal of ecosystems and institutions: Columbia University Press, New York, 593 p.

Hamblin, W.K., and Rigby, J.K., 1968, Guidebook to the Colorado River, Part 1 -- Lees Ferry to Phantom Ranch in Grand Canyon National Park: Provo, Utah, Brigham Young University Geology Studies, v. 15, part 5, 84 p.

Hammack, L., 1994, Debris-fan formation and rapid modification at Warm Springs Rapid, Yampa River, Colorado [M.S. Thesis]: Fort Collins, Colorado State University, 125 p.

Hammack, L., and Whol, E., 1996, Debris-fan formation and rapid modification at Warm Springs Rapid, Yampa River, Colorado: Journal of Geology, v. 104, p. 729-740.

Hampton, M.A., 1975, Competence of fine-grained debris flows: Journal of Sedimentary Petrology, v. 45, no. 4, p. 834-844.

Hastings, J.R., and Turner, R.M., 1965, The changing mile: an ecological study of vegetation change with time in the lower mile of an arid and semiarid region: Tucson, University of Arizona Press, 317 p. in the Grand Hereford, R., and Huntoon, P.W., 1990, Rock movement and mass wastage Geology: New Canyon: in Beus, S.S., and Morales, M., editors, Grand Canyon York, Oxford University Press, p. 443-460. warm-season rainfall, southern Hereford, R. and Webb, R.H., 1992, Historic variation of Colorado Plateau, southwestern U.S.A: Climatic Change, v. 22, p. 239-256. Tributary debris Hereford, R., Thompson, K.S., Burke, K.J., and Fairley, H.C., 1996, late Holocene alluvial chronology of the Colorado River, eastern fans and the no. 1, p. Grand Canyon, Arizona: Geological Society of America Bulletin, v. 108, 3-19 J.F., Day, J.C., and Kallio,R., 1990, The influence of man on Hirsch, R.M., Walker, Surface hydrologic systems, Chapter 13, the Geology of North America, Vol. 0-1, Water Hydrology, the Geological Society of America, p. 329-359. Hirschboeck, K.K., 1985, Hydroclimatology of flow events in the Gila River Basin, dissertation]: Tucson, University of Arizona, central and southern Arizona [Ph.D. 335 p. National Weather , 1991, Hydrology of floods and droughts: Climate and Floods: Summary 1988-89, p. 67-88. Wiley, Holling, C.S., 1978, Adaptive environmental assessment and management, John 482

London.

Hooke, R.L., 1993, The Trollheim alluvial fan and facies model revisited: Discussion: Geological Society of America Bulletin, v. 105, p. 563-564.

, 1967, Processes on arid-region alluvial fans: Journal of Geology, v. 75, p. 438- 460.

Howard, A., 1995, Rate laws for bedrock erosion by debris flows [abs.]: in AGU 1995 Fall Meeting Program and Abstracts. American Geophysical Union EUS Transactions, p. F263.

Howard, A., and Dolan, R., 1979, Changes in the fluvial deposits of the Colorado River in the Grand Canyon caused by Glen Canyon Dam: in Linn, R.M. (editor), Proceedings of the First Conference on Scientific Research in the National Parks, v. II, no. 5, U.S. National Park Service, New Orleans, LA, p. 845-851. , 1981, Geomorphology of the Colorado River in Grand Canyon: Journal of Geology, v. 89, p. 269-297.

Howe, E., 1909, Landslides of the San Juan Mountains, Colorado: U.S. Geological Survey Professional Paper 67, 58 p.

Hunt, C. B., 1969, Geologic history of the Colorado River, in The Colorado River region and John Wesley Powell: U.S. Geological Survey Professional Paper 669-C, p. 59-130. Huntoon, P.W., Billingsley, G.H., Breed, W.J., Sears, W.J., Ford, T.D., Clark, M.D., Babcock, S., Brown, E.H., 1986, Geologic map of the eastern part of the Grand Canyon National Park, Arizona: Grand Canyon Natural History Association Map, scale 1:62,500, 1 sheet.

Iverson, R.M., LaHusen, R.G., Major, J.J., and Zimmerman, C.L., 1994, Debris flow against obstacles and bends: dynamics and deposits: [abs.], EUS, Transactions, American Geophysical Union, v. 75, no. 44, p. 274.

Johnson, A.M., and Rodine, J.R., 1984, Debris flow, in Brunsden, D., and Prior, D.B., editors, Slope instability: New York, John Wiley and Sons, p. 257-361.

Junk, W.J., Bayley, P.B., and Sparks, R.E., 1989, The flood pulse concept in river- floodplain systems in Dodge, D.P., (ed.), Proceedings of the International Large River Symposium, Canadian Special Publication of Fisheries and Aquatic Science, v.106, pp. 110-127.

Kahya, E., and Dracup, J.A., 1993, U.S. Streamflow patterns in relation to the El Nirio/Southern Oscillation: Water Resources Research, v. 29, p. 2491-2503.

Karl, T.R., Knight, R.W., and Plummer, N., 1995, Trends in high-frequency climate 483

variability in the twentieth century: Nature, v. 377, P. 217-220. Kearsley, L. H., Schmidt, J.C., and Warren, K.D., 1994, Effects of Glen Canyon Dam on Colorado River sand deposits used as campsites in Grand Canyon National Park, USA: Regulated Rivers, v. 9, p. 137-149.

Kellerhals, R. and Bray, D.I., 1971, Sampling procedures for coarse fluvial sediments: American Society of Civil Engineers, Journal of the Hydraulics Division, v. 97, p. 1165-1180.

Kieffer, S.W., 1985, The 1983 hydraulic jump in Crystal Rapid -- Implications for river- running and geomorphic evolution in the Grand Canyon: Journal of Geology, v. 93, p. 385-406.

, 1987, The waves and rapids of the Colorado River, Grand Canyon, Arizona: U.S. Geological Survey Open File Report 87-096, 69 p. + Appendices.

, 1990, Hydraulics and geomorphology of the Colorado River in the Grand Canyon, in Beus, S.S., and Morales, M. (editors), Grand Canyon Geology: New York, Oxford University Press, p. 333-383.

Kim, S.B., 1995, Alluvial fans and their natural distinction from rivers based on morphology, hydraulic processes, sedimentary processes, and facies assembledges - Discussion: Journal of Sedimentary Research, v. A65, no. 4, p. 706-708.

Kirkby, M.J., 1977, Maximum sediment efficiency as a criterion for alluvial channels, in Gregory, K.J., (editor), River channel changes: New York, Wiley, p. 429-442. Laenen, Antonious, Scott, K.M., Costa, J.E., and Orzol, L.L., 1987, Hydrologic hazards along Squaw Creek from a hypothetical failure of a glacial moraine impounding Carver Lake near Sisters, Oregon: U.S. Geological Survey Open-File Report 87- 41, 48 p.

Langbein, W.B., and Leopold, L.B., 1964, Quasi-equilibrium states in channel morphology: American Journal of Science, v. 262, p. 782-794.

Lee, K.N., 1993, Compass and Gyroscope: integrating science and politics for the environment: Island Press, Washington, DC.

Leggett, R.F., Brown, R.J.E., and Johnson, G.H., 1966, Alluvial fan formation near Aklavik, Northwest Territories, Canada: Geological Society of America Bulletin, v. 77, p. 15-30.

Leopold, L.B., 1951, Vegetation of southwestern watersheds in the nineteenth century: The Geographical Review, XLI, p. 295-316.

, 1969, The rapids and pools -- Grand Canyon, in The Colorado River Region and John Wesley Powell: U.S. Geological Survey Professional Paper 669, p. 131- 484 145.

Ligon, F.K., Dietrich, W.E., and Trush, W.J., 1995, Downstream ecological effects of dams: a geomorphic perspective: BioScience, v. 45, no. 3, p. 183-192.

Lucchitta, I., and Young R.A., 1986, Structure and geomorphic character of western Colorado Plateau in the Grand Canyon-Lake Mead region: in Nations, J.D., Conway, C.M., and Swann, G.A., editors, Geology of central and northern Arizona. Geological Society of America, Rocky Mountain Section Field Trip Guidebook, p. 159-176.

Major, J.J., and Pierson, T.C., 1992, Debris flow rheology: experimental analysis of fine- grained slurries: Water Resources Research, v. 28, no. 3, p. 841-858.

Mason, A.C., and Foster, H.L., 1956, Extruded mudhills of Nirasaki, Japan: Journal of Geology, v. 64, p. 74-83.

McCarthy, T.S., and Cadle, A.B., 1995, Alluvial fans and their natural distinction from rivers based on morphology, hydraulic processes, sedimentary processes, and facies assembledges - Discussion: Journal of Sedimentary Research, v. A65, no. 3, p. 581-583.

McGee, W.J., 1897, Sheetflood erosion: Bulletin of the Geological Society of America, v. 8, p. 87-112.

McKee, E.D., Crosby, E.J., and Berryhill, H.L., Jr., 1967, Flood deposits, Bijou Creek, Colorado, June 1965: Journal of Sedimentary Petrology, v. 37, p. 829-851.

Melis, T.S., and Webb, R.H., 1993, Debris flows in Grand Canyon National Park, Arizona: magnitude, frequency, and effects on the Colorado River, in Shen, H.W., Su, S.T., and Wen, F., editors, Hydraulic Engineering '93, v. 2: New York, American Society of Civil Engineering, p. 1290-1295.

Melis, T.S., Webb, R.H., Griffiths, P.G., and Wise, T.W., 1994, Magnitude and frequency data for historic debris flows in Grand Canyon National Park and vicinity, Arizona: U.S. Geological Survey, Water Resources Investigation Report 94-4214, 285 p.

Melis, T.S., Webb, R.H., Kaplinski, M.A., Hazel, Jr., J.E., Shannon, J.P., Blinn, D.W., Valdez, R.A., and Ryel, R.P., 1995, Impacts of the 1993 Tanner Canyon debris flow on physical and aquatic resources of the Colorado River in Grand Canyon National Park, Arizona [abs.]: in AGU 1995 Fall Meeting Program and Abstracts. American Geophysical Union EOS Transactions, p. F254.

Melis, T.S., Phillips, W.M., Webb, R.H., and Bills, D.J., 1996, When the blue-green waters turn red: Historical flooding in Havasu Creek, Arizona: U.S. Geological Survey Water-Resources Investigation Report 96-4059, 136 p. 485

Melis, T.S., Webb, R.H., and Griffiths, P.G., 1997, Debris flows in Grand Canyon National Park: peak discharges, flow transformations, and hydrographs: American Society of Civil Engineers, Proceedings of the First International Conference on Debris Flow Hazard, Mitigation, and Prediction, in press.

Mullineaux, D.R., and Crandell, D.R., 1962, Recent lahars from Mount St. Helens, Washington: Geological Society of America Bulletin, v. 73, p. 855-870.

O'Connor, J.E., Ely, L.L., Whol, E.E., Stevens, L.E., Melis, T.S., Kale, V.S., and Baker, V.R., 1994, A 4,500 year record of large floods on the Colorado River in the Grand Canyon, Arizona: Journal of Geology, v. 102, p. 1-9.

Pack, F.J., 1923, Potential of desert waters: Pan American Geologist, v. 40, p. 349-357.

Parasiewics, P., 1996, Estimation of physical habitat characteristics using automation and geodesic-based sampling: Regulated Rivers: Research & Management, V. 12, p. 575-583 Péwé, T.L., 1968, Colorado River Guidebook: A geologic and geographic guide from Lees Ferry to Phantom Ranch: Lebeau Printing Inc., Phoenix, Arizona, 78 p. Pierson, T.C., 1980, Erosion and deposition by debris flows at Mt. Thomas, North Canterbury, New Zealand: Earth Surface Processes, v. 5, p. 227-247.

Pierson, T.C., 1984, Effects of slurry composition on debris flow dynamics, Rudd Canyon, Utah, in Proceedings of the Specialty Conference on delineation of Landslide, Flash Flood, and Debris Flow Hazards in Utah, D. S. Bowles, editor, Utah State University, Logan Utah.

Pierson, T.C., 1985, Initiation and flow behavior of the 1980 Pine Creek and Muddy River lahars, Mount St. Helens, Washington: Geological Society of America Bulletin, v. 96, p. 1056-1069. from Pierson, and Scott, K.M., 1985, Downstream dilution of a lahar: Transition T.C., v. 21, p. debris flow to hyperconcentrated streamflow: Water Resources Research, 1511-1524. sediment- Pierson, T.C., and Costa, J.E., 1987, A rheologic classification of subaerial water flows, in Costa, J.E., and Wieczorek, G.F., editors, Debris flows / avalanches-- Process, recognition, and mitigation: Reviews in Engineering Geology, v. VII, p. 1-12. T.C., Janda, R.J., Thouret, J.C., and Borrero, C.A., 1990, Perturbation and Pierson, del Ruiz, of snow and ice by the 13 November 1985 eruption of Nevado melting Journal of Colombia, and consequent mobilization, flow and deposition of lahars: volcanology and Geothermal Research, v. 41, p. 17-66. Pa: Flood and Vincent, The Powell, J.W., 1895, Canyons the Colorado River: Meadville, 486

Chatauqua-Century Press, 400 p. [facsimile repr. 1961, as The exploration of the Colorado River and its canyons, New York, Dover Publications, Inc., 400 p.

Rice, S., and Church, M., 1996, Sampling surficial fluvial gravels: The precision of size distribution percentile estimates: Journal of Sedimentary Research, v. 66, no. 3, p. 654-665.

Rickmers, W.R., 1913, The Duab of Turkestan: A physiographic sketch and account of some travels: Cambridge, Cambridge University Press, p. 193-200.

Rodine, J.D., and Johnson, A.M., 1976, the ability of debris, heavily freighted with coarse clastic materials, to flow on gentle slopes: Sedimentology, v. 23, p. 213- 234.

Rubin, D.M., Schmidt, J.C., and Moore, J.N., 1990, Origin, structure, and evolution of a reattachment bar, Colorado River, Grand Canyon, Arizona: Journal of Sedimentary Petrology, v. 60, p. 982-991.

SAS Institute, Inc., 1990, SAS/STAT User's Guide, Version 6, (4th edition), v. 2: Cary, N.C., SAS Institute Inc., 846 p.

Schick, A.P., 1988, Hydrologic aspects of floods in extreme arid environments, in Baker, V.R., Kochel, R.C., and Patton, P.C., editors, Flood geomorphology: New York, John Wiley and Sons, p. 189-203.

Schmidt, J.C., 1990, Recirculating flow and sedimentation in the Colorado River in Grand Canyon, Arizona: Journal of Geology, v. 98, p. 709-724.

Schmidt, J.C., and Graf, J.B., 1990, Aggradation and degradation of alluvial sand deposits, 1965 to 1986, Colorado River, Grand Canyon National Park, Arizona: U.S. Geological Survey Professional Paper 1493, 74 p.

Schmidt, J.C., Rubin, D. M., and Ikeda, H., 1993, Flume simulation of recirculating flow and sedimentation: Water Resources Research, v. 29, no. 8, p. 2925-2939.

Schmidt, J.C., Mayers, J.L., and Leschin, M.F., 1994a, Geomorphology of post-Glen Canyon Dam fine-grained alluvial deposits of the Colorado River near Point Hansbrough and Saddle Canyon in Grand Canyon National Park: Bureau of Reclamation, Glen Canyon Environmental Studies administrative report, 16 p.

Schmidt, J.C., Mayers, J.L., Leschin, M.F., and Rubin, D.M., 1994b, [abs.] Geomorphic mapping of post-Glen Canyon Dam fine-grained alluvial deposits of the Colorado River near Point Hansbrough and Saddle Canyon in Grand Canyon, Arizona: Geological Society of America Annual Meeting Proceedings.

Schmidt, J.C. and Rubin, D.M., 1995, Regulated streamflow, fine-grained deposits and effective discharge in canyons with abundant debris fans: American Geophysical Union Monograph 89, 177-195. 487

Schmidt, J.C., Grams, P.E., and Webb, R.H., 1995, Comparison of the magnitude of erosion along two large regulated rivers: Water Resources Bulletin, v. 31, no. 4, p. 617-631.

Schmidt, J.C., Webb, R.H., Valdez, R.A., Marzolf, G.R., and Stevens, L.E., 1997, The role of science and the role of value judgement in restoring the Colorado River in Grand Canyon: Bioscience, (in press).

Schumm, S.A., 1991, To interpret the Earth -- Ten ways to be wrong: Cambridge Universtiy Press, New York, 133 p.

Scott, K.M., 1988, Origins, behavior, and sedimentology of lahars and lahar-runout flows in the Toutle-Cowlitz River system: U.S. Geological Survey Professional Paper 1447-A, 74 p. Scrivenor, J.B., 1929, The mudstreams ("Lahars") of Gunong keloet in Java: Geological Magazine, v. LXVI, no. X, p. 433-434.

Shannon, J.P., Blinn, D.W., Benenati, P.L., and Wilson, K.P., 1996, Organic drift in a regulated desert river: Canadian Journal of Fisheries and Aquatic Science, v. 53, p. 1360-1369.

Sharp, R.P., 1942, Mudflow levees: Journal of Geomorphology, v. 5, p. 222-227.

Sharp, R.P., and Nobles, L.H., 1953, Mudflows of 1941 at Wrightwood, southern California: Geological Society of America Bulletin, v. 64, p. 547-560.

Sharpe, C.F.S., 1938, Landslides and Related Phenomena: New York, Columbia University Press, 137 p.

Shreve, R.L., 1968, The Blackhawk landslide: Geological Society of America Special Paper 108, 47 p.

Simmons, G.C., and Gaskill, D.L., 1969, River runner's guide to the canyons of the Green and Colorado Rivers, with emphasis on geologic features, Volume III, Press, 132 p. Marble Gorge and Grand Canyon: Flagstaff, Arizona, Northland Smith, D.L., and Crampton, C.G., 1987, The Colorado River Survey: Robert B. Stanton and the Denver, Colorado Canyon & Pacific Railroad: Howe Brothers, Salt Lake City, 305 p. of Stanford, J.A., Ward, J.V., and others, 1996, A general protocol for restoration regulated rivers: Regulated Rivers: Research & Management, v. 12, p. 391-413. by Dwight L. Smith, Stanton, R.B., 1965, Down the Colorado: edited with an introduction University of Oklahoma Press, Norman, p. 237. 488

Stephens, H.G., and Shoemaker, E.M., 1987, In the footsteps of John Wesley Powell: Boulder, Colorado, Johnson Books, 286 p.

Stevens, L., 1990, The Colorado River in Grand Canyon, A guide: Flagstaff, Arizona, Red Lake Books, 115 p.

Stevens, L.E., Schmidt, J.C., Ayers, T.J., and Brown, B.T., 1995, Flow regulation, geomorphology and Colorado River marsh development in Grand Canyon, Arizona: Ecological Applications, v. 5, p. 1025-1039.

Thomas, B.E., Hjalmarson, H.W., Waltemeyer, S.D., 1994, Methods for estimating the magnitude and frequency of floods in the southwestern United States: U.S. Geological Survey Open-File Report 93-419, 211 p.

Tolman, C.F., 1909, Erosion and deposition in the southern Arizona bolson region: Journal of Geology, v. 17, p. 136-163.

, 1905-6, Notes on desert processes and desert deposits: Journal of the Proceedings of the Annual Convention of the Arizona Miners Association, p. 15.

Trenberth, K.E., 1996, El Nifio Definition: CLIVAR - Exchanges, v. 1, no. 3, p. 6-8.

Trowbridge, A.C., 1911, The terrestrial deposits of Owens Valley, California: Journal of Geology, v. 19, p. 706-747.

Turner, R.M., and Karpiscak, M.M., 1980, Recent vegetation changes along the Colorado River between Glen Canyon Dam and Lake Mead, Arizona: U.S. Geological Survey Porfessional Paper 1132, 125 p.

U.S. Water Resources Council, 1981, Guidelines for determining flood flow frequency: Hydrology Subcommittee Bulletin 17B, 183 p.

of the Interior, 1988, Glen Canyon Environmental Studies, final report: U.S. Department No. Springfield, Virginia, National Technical Information Service Report NTIS PB88-183348/AS, 84 p. + Appendices.

of the Interior, 1995, Operation of Glen Canyon Dam: final U.S. Department and Public environmental impact statement + Appendices, Executive Summary, Comments, Bureau of Reclamation, 337 p. humpback chub (Gila Valdez, R.A., and Ryel, R.J., 1996, Ecology and status of the in the Colorado River, Grand Canyon, Arizona: final report to Bureau of cypha) Report Reclamation, Salt Lake City, Utah, Contract No. 0-CS-40-09110, Bio/West No. TR-250-08, 286 p. H.H., 1967, Debris flow and erosion control problems caused by ash eruptions Waldron, Bulletin 1241-I, 37 of Irazu volcano, Costa Rica: United States Geological Survey p. 489

Walters, C.J., 1986, Adaptive Management of renewable resources: MacMillan Press, New York.

Ward, J.V., and Stanford, J.A., 1995a, The serial discontinuity concept: extending the model to floodplain rivers: Regulated Rivers: Research & Management, v. 10, p. 159-168.

, 1995b, Ecological connectivity in alluvial river ecosystems and its disruption by flow regulation: Research & Management, v. 11, p. 105-119.

Webb, R.H., 1996, Grand Canyon, a century of change: Rephotography of the 1889-1890 Stanton expedition: Tucson, University of Arizona Press, 290 p.

Webb, R.H., and Melis, T.S., 1995, The 1995 debris flow at Lava Falls Rapid: Nature Notes, Grand Canyon National Park, fall, 1995, V. 11, no. 3, p. 1-4.

Webb, R.H., and Betancourt, J.L., 1992, Climatic variability and flood frequency of the Santa Cruz River, Pima County, Arizona: U.S. Geological Survey Water-Supply Paper 2379, 40 p.

Webb, R.H., Melis, T.S., and Schmidt, J.C., 1991a, Historical Analysis of Debris Flows, Recirculation Zones, and Changes in Sand Bars Along the Colorado River in Grand Canyon, Arizona [abs.]: EOS Supplement, Trans., Amer. Geo. Union, v. 72, no. 44, p. 219.

Webb, R.H., Smith, S.S., and McCord, V.A.S., 199 lb, Historical channel change of Kanab Creek, southern Utah and northern Arizona: Grand Canyon Natural History Association Monograph Number 9, 91 p. of the Webb, R.H., Pringle, P.T., and Rink, G.R., 1987, Debris flows from tributaries Colorado River in Grand Canyon National Park, Arizona: U.S. Geological Survey Open-File Report 87-118, 64 p. rapids , 1988, Monument Creek debris flow, 1984: Implications for formation of on the Colorado River in Grand Canyon National Park: Geology, v. 16, p. 50-54.

in tributaries of the Colorado River in Grand Canyon National , 1989, Debris flows p. Park, Arizona: U.S. Geological Survey Professional Paper 1492, 39 Griffiths, P.G., and Elliott, J.G., 1997, Reworking of Webb, R.H., Melis, T.S., flood on the Colorado River in Grand aggraded debris fans by the 1996 controlled Canyon National Park, Arizona: U.S. Geological Survey Open-File Report 97-16, 36 p. cataract," Webb, R.H., Melis, T.S., Wise, T.W., and Elliott, J.G., 1996, "The great on Lava Falls Rapid, Grand Canyon National Effects of late Holocene debris flows Park, Arizona: U.S. Geological Survey Open-File Report 96-460, 96 p. 490

Wentworth, C.K., 1943, Soil avalanches on Oahu, Hawaii: Geological Society of America Bulletin, v. 54, p. 53-64.

Wiberg, P.L., and Smith, J.D., 1987, Calculations of the critical shear stress for motion of uniform and heterogeneous sediments: Water Resources Research, v.23, p. 1471- 1480.

Wieczorek, G.F., 1987, Effect of rainfall intensity and duration on debris flows on central Santa Cruz Mountains, California: in Debris Flows/Avalanches: process, recognition, and mitigation: Reviews in Engineering Geology, v. VII, Geological Society of America, Boulder, CO, 23-104.

Wilson, R.C., Mark, R.K., and Barbato, G., 1993, Operation of a real-time warning system for debris flows in the San Francisco Bay area, California: in Shen, H.W., Su, S.T., and Wen, F., editors, Hydraulic Engineering '93, v. 2: New York, American Society of Civil Engineering, p. 1908-1913.

Winder, C.G., 1965, Alluvial cone construction by alpine mudflow in a humid temperate region: Canadian Journal of Earth Sciences, v. 2, p. 270-277.

Whol, E.E., and Pearthree, P.P., 1991, Debris flows as geomorphic agents in the Juachuca Mountains of southeastern Arizona: Geomorphology, v. 4, p. 273-292.

Williams, G.P., and Wolman, M.G., 1984, Downstream Effects of Dams on Alluvial Rivers: Geological Survey Professional Paper 1286, 81 p.

Wolman, M.G., 1954, A method of sampling coarse-river-bed material: Transactions of the American Geophysical Union, v. 35, no. 6, p. 951-956.