Fire Management today Volume 63 • No. 3 • Summer 2003

WILDLAND FIRE BEHAVIOR CASE STUDIES AND ANALYSES: PART 1

United States Department of Agriculture Forest Service DEDICATION

This special issue of Fire Management Today is dedicated to the memory of Paul M. Gleason (1946–2003).

His passion for wildland firefighter safety and his deep professional interest in wildland fire behavior will be sorely missed.

Martin E. Alexander and David A. Thomas Issue Coordinators

Editor’s note: This issue of Fire Management Today reprints articles from early editions of the journal, some of them decades old. Although the articles appear in today’s format, the text is reprinted largely ver­ batim and therefore reflects the style and usage of the time. We made minor wording changes for clarity, added intertitles and metric conversions where needed, and occasionally broke up paragraphs or broke out sidebars to improve readability. All illustrations are taken from the original articles.

Erratum In the Spring 2003 issue of Fire Management Today, the table of contents showed an incorrect title for Stephen J. Pyne’s article (volume 63[2], page 17). The correct title is “Firestop II.”

Fire Management Today is published by the Forest Service of the U.S. Department of Agriculture, Washington, DC. The Secretary of Agriculture has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department.

Fire Management Today is for sale by the Superintendent of Documents, U.S. Government Printing Office, at: Internet: bookstore.gpo.gov Phone: 202-512-1800 Fax: 202-512-2250 Mail: Stop SSOP, Washington, DC 20402-0001

Fire Management Today is available on the World Wide Web at .

Ann M. Veneman, Secretary April J. Baily U.S. Department of Agriculture General Manager

Dale Bosworth, Chief Robert H. “Hutch” Brown, Ph.D. Forest Service Managing Editor

Jerry Williams, Director Madelyn Dillon Fire and Aviation Management Editor

Carol LoSapio Guest Editor

Martin E. Alexander, Ph.D., and David A. Thomas Issue Coordinators

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, sex, religion, age, disability, political beliefs, sexual orientation, or marital or family sta­ tus. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD).

To file a complaint of discrimination, write USDA, Director, Office of Civil Rights, Room 326-W, Whitten Building, 1400 Independence Avenue, SW, Washington, DC 20250-9410 or call (202) 720-5964 (voice and TDD). USDA is an equal opportunity provider and employer.

Disclaimer: The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are responsible for the technical accuracy of the material presented in Fire Management Today. Fire today Management Volume 63 • No. 3 • Summer 2003

On the Cover: CONTENTS Wildland Fire Behavior Case Studies and Analyses: Value, Approaches, and Practical Uses ...... 4 M.E. Alexander and D.A. Thomas Blackwater Fire on the Shoshone ...... 9 Division of Fire Control The Factors and Circumstances That Led to the Blackwater Fire Tragedy ...... 11 A.A. Brown Lessons From Larger Fires on National Forests, 1938 ...... 15 Roy Headley Lessons From Larger Fires on National Forests, 1939 ...... 23 Historical photo from USDA Roy Headley Forest Service files showing Lessons of the McVey Fire, Black Hills National Forest ...... 25 A.A. Brown the Wheeler Fire in 1948 An Analysis of the Honey Fire ...... 29 “working down Bear Canyon C.F. Olsen toward Wheeler Gorge Camp” The Bower Cave Fire ...... 42 on the Los Padres National Leon R. Thomas Forest, CA. Photo: Forest The Possible Relation of Air Turbulence to Erratic Fire Service Photograph Behavior in the Southeast ...... 46 Collection, USDA Forest George M. Byram and Ralph M. Nelson The Pinyon–Juniper Fuel Type Can Really Burn ...... 52 Service, Washington Office, Dwight A. Hester Washington, DC (no. 451594; A Firewhirl of Tornadic Violence ...... 54 F.E. Dunham, 1948). Howard E. Graham Rate of Spread on a Washington Fern Fire ...... 56 William G. Morris Fire-Whirlwind Formation as Favored by Topography and Upper Winds ...... 59 Howard E. Graham Relationship of Weather Factors to Rate of Spread of the Robie Creek Fire ...... 63 The FIRE 21 symbol (shown below and on the R.T. Small cover) stands for the safe and effective use of A Key to Blowup Conditions in the Southwest? ...... 68 wildland fire, now and throughout the 21st century. Its shape represents the fire triangle Robert W. Bates (oxygen, heat, and fuel). The three outer red A Fire-Whirlwind in Alabama ...... 71 triangles represent the basic functions of Gordon Powell wildland fire organizations (planning, operations, and aviation management), and the three critical The Forest Fires of April 1963 in New Jersey Point aspects of wildland fire management (prevention, the Way to Better Protection and Management ...... 74 suppression, and prescription). The black interior represents land affected by fire; the emerging Wayne G. Banks and Silas Little green points symbolize the growth, restoration, The Harrogate Fire—March 15, 1964 ...... 79 and sustainability associated with fire-adapted B.J. Graham ecosystems. The flame represents fire itself as an ever-present force in nature. For more informa­ The Fire Behavior Team in Action: The Coyote Fire, 1964 . . . . . 81 tion on FIRE 21 and the science, research, and John D. Dell innovative thinking behind it, contact Mike Apicello, National Interagency Fire Center, “Gleason Complex” Puts Up Huge “Plume”: 208-387-5460. A Tribute to Paul Gleason ...... 85 Paul Keller Interview With Paul Gleason ...... 91 Jim Cook and Angela Tom

SHORT FEATURES Firefighter and public safety is Websites on Fire ...... 80 our first priority. Guidelines for Contributors ...... 95

Volume 63 • No. 3 • Summer 2003 3 WILDLAND FIRE BEHAVIOR CASE STUDIES AND ANALYSES: VALUE, APPROACHES, AND PRACTICAL USES

M.E. Alexander and D.A. Thomas

ince 1936, the Washington Office of the USDA Forest In an effort to unbury the past S Service has published a period­ and to increase both institutional memory ical devoted to articles dealing with and organizational learning within the wildland a very wide range of fire manage­ ment topics. The name of this jour­ fire community, we are reprinting past articles nal has changed through the years, on fire behavior. from Fire Control Notes, to Fire Management, to Fire Management Notes, and finally to Fire Manage­ This special issue of Fire Manage­ ly support this notion and have ment Today.* A good many of the ment Today begins a series of three endeavored to reflect it in our indi­ 243 issues that have been pub­ consecutive issues with articles vidual work areas in fire research lished in the past 67 years have related to fire behavior. This issue and fire management, respectively included a fire-behavior-related contains the first of two install­ (Alexander and Lanoville 1987; article. With the passage of time, ments of articles involving fire Thomas 1991). however, many of these articles behavior case studies and analyses have become “buried,” found only of wildfires; examples pertaining to The idea of relying on wildfires as a by the most intrepid researchers on prescribed fires are not included possible source of data is especially the shelves of major libraries. (e.g., Custer and Thorsen 1996). pertinent to empirically based The 19 case studies and analyses in schemes for quantitative fire In an effort to unbury the past and this issue are presented in chrono­ behavior prediction that rely on to increase both institutional mem­ logical order, from 1937 to 1967. this kind of information in whole ory and organizational learning The third issue in this series will or in part (e.g., Alexander 1985; within the wildland fire communi­ be devoted to aids, guides, and Forestry Canada Fire Danger ty, the authors approached the edi­ knowledge-based protocols Group 1992; Rothermel 1991). torial staff of Fire Management involved in forecasting wildland This fact is especially significant at Today with the idea of republishing fire behavior for safe and effective the extreme end of the fire intensi­ a selection of these past fire-behav­ fire suppression. ty scale, where experimental fires ior-related articles. We are pleased are exceedingly difficult to arrange that they took us up on our sug­ General Value of Case (Alexander and Quintilio 1990; gestion. Studies Cheney and others 1998). The importance of documented case studies or histories of wildland Information gleaned from wildland fires has been repeatedly empha­ fire behavior case studies has also Marty Alexander is a senior fire behavior sized by both fire managers and proved of value in testing and eval­ research officer with the Canadian Forest fire researchers (e.g., Byram 1960; uating various fire models, theo­ Service at the Northern Forestry Centre, ries, decision aids and support sys­ Edmonton, Alberta; and Dave Thomas is Thomas 1994; Turner and others the regional fuels specialist for the USDA 1961). As long-time Forest Service tems, and management guidelines Forest Service, Intermountain Region, wildland fire researcher/adminis­ (e.g., Anderson 1983; Haines and Ogden, UT. trator Craig Chandler (1976) has others 1986; Nelson 1993; Pearce noted, “Time and time again case and Alexander 1994). For example, * For more on the history of Fire Management Today, Lindenmuth and Davis (1973) used see Hutch Brown, “How Did Fire Control Notes histories have proven their value as Become Fire Management Today?” Fire Management training aids and as sources of an observation of the initial run of Today 60(1) [Winter 2000]: 8–14. research data.” The authors strong- the Battle Fire, a 28,400-acre

Fire Management Today 4 (11,500-ha) fire that occurred May 1966; Sackett and DeCoste 1967). historical information to establish 14–20, 1972, on the Prescott This was no doubt due in large part the fire’s chronology and general National Forest, AZ, to assess the to George Byram’s (1960) influ­ behavior. The reports of Haines and performance of their empirically ence. Sando (1969), Stocks and Walker based model for predicting fire (1973), Street and Alexander spread in Arizona’s oak chaparral Some limited documentation has (1980), and Rothermel (1993) are fuel type. also been undertaken by fire man­ good examples of this approach to agers and fire researchers serving case studies. Approaches to Case as fire behavior officers or special­ Studies ists/analysts on various wildland Pragmatic Value of There are many examples in North fire incidents (e.g., Johnson 1964; Case Studies America and elsewhere where fire McCaw, Maher, and Gillen 1992; A practical fire manager, always researchers and fire managers have Norum 1982; Thomas 1991). Fire interested in the control of wild­ attempted to observe and docu­ researchers have also been involved fires and the ignition of prescribed ment the behavior of free-burning in many “after-the-fact” investiga­ fires, might ask: What is the use of fires, using various types of data tions (e.g., Butler and Reynolds historical fire behavior case stud­ collection methods and monitor­ 1997; Fogarty and others 1996; ies? How can old documents help ing equipment, on an ad hoc or a McCaw, Simpson, and Maher fire management personnel more formal basis (e.g., Barney 1992). Van Wagner (1971) has become better managers of forest and others 1978; Barrows 1961; pointed out that “some valuable and range fires, in all their forms? Billing 1986; Schaefer 1957; reference data can be collected by Beyond the recreation of a “good Traylor 1961*). These efforts being in the right place at the right read,” what utility do these articles extend back many years (Gisborne time. It is, in fact, quite feasible to offer? How can old essays become 1929) and continue into the 21st obtain good data by visiting the relevant for a 21st-century fire­ century (e.g., Burrows and others scene of a … fire shortly after it fighter? 2000). has occurred, while its history is still fresh both on the ground and The old articles will only seem Fire behavior researchers are in the mind of the fire boss.” dated if we fail to make use of rarely in the right place at the them. There are two primary rea­ right time to observe and docu­ Byram (1954) made extensive use sons to thoroughly study these fire ment the behavior of forest and of the case study method of indi­ behavior case studies: range fires. While there have, of vidual fires in his research into course, been some exceptions (e.g., blowup fire behavior. As he notes, •To learn from them and thereby Sneeuwjagt 1974; Stocks and “Some of the observations and lessen the chance of making the Flannigan 1987), including escapes details of behavior are written same mistake again; and from outdoor experimental fires down in fire reports, but most of •To prepare ourselves not to be (Alexander and others 1991; Stocks the information is still in the surprised to the point of distrac­ 1987), for the most part fire opera­ memories of men who worked on tion by a fire’s surprising behav­ tions personnel tend to be in the the fires. Fire behavior may, there­ ior in a particular fuel type under best position to make and record fore, be difficult to reconstruct at a given weather condition. key observations. Probably the times, especially on fires which most concerted and systematic occurred a number of years ago. Not making the same mistake effort by fire researchers to observe Usually, however, a surprising twice and being prepared to be sur­ and record actual fire behavior was amount of detail can be obtained prised will go a long way toward made by the Forest Service’s by talking with men who were on creating a highly reliable firefight­ Southern Forest Fire Laboratory in the fires and by going over the fire ing organization where safety truly Macon, GA, from the late 1950s to area with them.” matters. early 1970s (DeCoste and Sackett A final possibility is the hindsight Unless we actively learn from past analysis of major wildland fire inci­ * A summary of this work can be found in R.E. Traylor, wildland fires, then the only way “Correlation of Weather to Fire Spread in Grass and dents in the light of present-day we can gain additional fire behav­ Brush Fuels on the Snake River Plains in Southern knowledge and tools using existing Idaho,” Fire Control Notes 22(4) [Fall 1961]: 118–119. ior knowledge is to actually experi-

Volume 63 • No. 3 • Summer 2003 5 ence a fire’s behavior or to model “Time and time again case histories the fire’s behavior on a computer have proven their value as training aids at our desk. Even the most active fire behavior analyst (FBAN) rarely and as sources of research data.” gets enough near-real-time oppor­ –Craig Chandler (1976) tunities to predict the spread and intensity in every fuel complex or experience does not include know­ • Is there something that we can to complete a prediction enough ledge of all the conditions.” apply to our current situation? times to become good at it Presumably then, case studies can • Have we learned all that this old (Thomas 1994). The best learning help supplement and thereby fire has to teach us? scenario for a practicing fire behav­ strengthen (but never replace) a • Could the same situation occur iorist is a combination of all three person’s experience level. today? learning techniques: actively using • What are we going to do differ­ case studies, getting field experi­ Safety Value of Case ently after reading this case ence, and practicing computer Studies study? modeling. Each is a distinct mode As we read through this chronolog­ of learning and adaptation; when ical selection of articles, especially This process, if faithfully followed combined, they become a powerful the accounts of forest fires where throughout a fire season, would model for continuous learning. firefighters lost their lives or there increase both mindfulness and were near-misses or unforeseen resilience (Weick and Sutcliffe Case study knowledge, coupled blowups, we can ask ourselves and 2001), the two hallmarks of indi­ with experienced judgment and fire our crews whether we have fully viduals and their organizations behavior modeling, is also consid­ grasped the major “lessons determined to do everything they ered an effective operational tech­ learned” from these past fire can to control and use wildland nique or procedure for appraising behavior events. Excellent methods fire safely. fire potential (Brown 1978). of using past fire behavior knowl­ Burrows (1984) maintains that edge from case studies to increase Both authors have used case stud­ most wildland firefighters base wildland firefighter safety in the ies to lead training sessions in the their expectations of how a fire will future are the staff ride (Alexander classroom. One of us (Thomas) has behave largely on experience and, 2002; Thomas and Cook 2002),* also used the technique in the field to a lesser extent, on fire behavior the sand box exercise (Euler 1946), at the site of past fires. In June guides. If this is indeed the case, yearly fire refreshers (e.g., the 2001 1994, a group of FBANs on a visit then it is worth reiterating the USDI Bureau of Land Management to the site of the 1949 Mann Gulch points made by Forest Service fire Fireline Safety Refresher videos), Fire were asked, using existing his­ research pioneer Harry T. Gisborne and weekly tailgate safety meet­ torical case study information as a (1948) about experienced judg­ ings. starting point for a fire behavior ment: “For what is experienced prediction, if they could have pre­ judgment except opinion based on For example, one of these articles vented the firefighter fatalities that knowledge acquired by experience? could be handed out each week to occurred on this infamous fire. If you have fought forest fires in members of an organized fire crew. Using the available historical fire every different fuel type, under all The crew would be given time to information, a similar question possible kinds of weather, and if read and ponder the article. Then, was asked of a large group of fire you have remembered exactly what in a group setting, with the fire management personnel on a staff happened in each of these combi­ foreman (i.e., hotshot superintend­ ride of the 1990 Dude Fire nations, your experienced judg­ ent, smokejumper-in-charge, local (Thomas and Cook 2002). In both ment is probably very good. But if fire management officer, etc.) act­ of these examples, many of the stu­ you have not fought all sizes of ing as facilitator, the crew could be dents said that these “training” fires in all kinds of fuel types under led through a series of questions sessions were some of the best they all kinds of weather then your that the article has inspired. For had ever attended. Using case stud­ example: ies or histories, an “old” fire’s fire * For more on the staff ride technique, see the various behavior came alive. articles on the Dude Fire Staff Ride in Fire Management Today 62(4) [Fall 2001].

Fire Management Today 6 “A surprising amount of detail can be obtained MD: Society of American Foresters: 287–304. by talking with men who were on the fires Alexander, M.E. 2002. The staff ride and by going over the fire area with them.” approach to wildland fire behavior and firefighter safety awareness training. Fire –George Byram (1954) Management Today. 62(4): 25–30. Alexander, M.E.; Lanoville, R.A. 1987. Wildfires as a source of fire behavior Another benefit of having these follow, are in keeping with the data: A case study from Northwest Territories, Canada. In: Postprint articles available again is for their ideals and sentiment expressed by Volume, Ninth Conference on Fire and use within fuel specialist reports Roy Headley (1936) in the very first Forest Meteorology; 1987 April 21–24; used in environmental assess­ issue of Fire Control Notes. San Diego, CA. Boston, MA: American ments. Fuel specialists are increas­ Meteorological Society: 86–93. Headley, who cofounded the jour­ Alexander, M.E.; Quintilio, D. 1990. ingly called upon to justify why an nal as the head of the Forest Perspectives on experimental fires in interdisciplinary team recom­ Service’s Division of Fire Control Canadian forestry research. Mathematical mended one fire hazard abatement (the predecessor of today’s Fire and and Computer Modelling. 13(12): 17–26. Alexander, M.E.; Stocks, B.J.; Lawson, B.D. technique over another. These case Aviation Management), called for 1991. Fire behavior in black histories, especially the descrip­ integrating and sharing “the expe­ spruce–lichen woodland: The Porter Lake tions of fire behavior in a given rience, thinking, and experiments” Project. Inf. Rep. NOR–X–310. Edmonton, AB: Forestry Canada, fuel type (e.g., Helms 1979), could of the many people engaged in Northern Forestry Centre. be cited in those reports (or hyper- wildland fire management. To this Anderson, H.E. 1983. Predicting wind-driv­ linked to a main database), saving end, Headley envisioned Fire en wild land fire size and shape. Res. Pap. much analysis time. The fuels spe­ INT–305. Ogden, UT: USDA Forest Control Notes as “a common meet­ Service, Intermountain Forest and Range cialist would not have to explain ing ground, a clearing-house of Experiment Station. how a fire might burn in a given developments.” In this sense, Fire Barney, R.J.; Noste, N.V.; Wilson, R.A. 1978. fuel type, for she or he would have Management Today, by republish­ Rates of spread of wildfire in Alaskan fuels. Res. Note PNW–311. Portland, OR: a published account to cite or ing the past (and thereby reviving USDA Forest Service, Pacific Northwest hyperlink to. it for the future), has rediscovered Forest and Range Experiment Station. its own unique niche. Barrows, J.S. 1961. Natural phenomena exhibited by forest fires. In: Berl, W.G., Learning Contribution ed. Proceedings of the International A learning organization has been Acknowledgments Symposium on the Use of Models in Fire Research; 1960 November 9–10; defined as one that is “skilled at The authors offer their sincerest Washington, DC. Publ. 786. Washington, creating, acquiring, interpreting, heartfelt appreciation to Hutch DC: National Academy of Sciences— transferring, and retaining knowl­ Brown, Madelyn Dillon, and Carol National Research Council. edge, and at purposefully modify­ Billing, P. 1986. Operational aspects of the LoSapio, editors of Fire Manage­ infra-red line scanner. Res. Rep. No. 26. ing its behavior to reflect new ment Today, for their significant Melbourne, VIC: Victoria Department of knowledge insights” (Garvin 2000). contributions to this special issue, Conservation, Forests & Lands, Fire Fire behavior case studies go a and to April Baily, the journal’s Protection Branch. Brown, J.K. 1978. Fuel inventory and long way toward preparing a foun­ general manager, for supporting appraisal. Paper presented at the USDA dation for organizational learning; the concept of these special issues Forest Service National Fire-Danger and in so doing, they follow the true on wildland fire behavior. Their Fire-Weather Seminar; 1972 November 14–16; Missoula, MT. spirit of learning implied in this dedication and outstanding editori­ Burrows, N.D. 1984. Predicting blow-up definition. Simply put, our fire al abilities have brought “life” to fires in the jarrah forest. Tech. Pap. No. management culture, now domi­ many of the articles contained in 12. Perth, WA: Forests Department of nated by a learning pattern of trial Western Australia. this issue that have long been Burrows, N.; Ward, B.; Robinson, A. 2000. and error, would become a learn­ forgotten. Behavior and some impacts of a large ing culture, one in which a sys­ wildfire in the Gnangara maritime pine tematic study of the past through (Pinus pinaster) plantation, Western References Australia. CALMScience. 3: 251–260. the use of case studies would Alexander, M.E. 1985. Estimating the Butler, B.W.; Reynolds, T.D. 1997. Wildfire become a routine procedure. length-to-breadth ratio of elliptical forest case study: Butte City Fire, southeastern fire patterns. In: Donoghue, L.R.; Martin, Idaho, July 1, 1994. Gen. Tech. Rep. R.E., eds. Proceedings of the Eighth INT–GTR–351. Ogden, UT: USDA Forest This special issue of Fire Manage­ Conference on Fire and Forest Service, Intermountain Research Station. ment Today devoted to fire behav­ Meteorology; 1985 April 29–May 2; ior, and the two others that will Detroit, MI. SAF Publ. 85–04. Bethesda,

Volume 63 • No. 3 • Summer 2003 7 Byram, G.M. 1954. Atmospheric conditions Haines, D.A.; Sando, R.W. 1969. Climatic Schaefer, V.J. 1957. The relationship of jet related to blowup fires. Stn. Pap. No. 35. conditions preceding historically great streams to forest wildfires. Journal of Asheville, NC: USDA Forest Service, fires in the North Central Region. Res. Forestry. 55: 419–425. Southeastern Forest Experiment Station. Pap. NC–84. St. Paul, MN: USDA Forest Sneeuwjagt, R.J. 1974. Evaluation of the [Reprinted as: National Fire Equipment Service, North Central Forest grass fuel model of the National Fire System Publication NFES 2565 by the Experiment Station. Danger Rating System. M.S. thesis. National Wildfire Coordinating Group, Headley, R. 1936. Fire Control Notes offers Seattle, WA: University of Washington. Boise, ID.] its services. Fire Control Notes. 1(1): 3–4 Stocks, B.J. 1987. Fire potential in the Byram, G.M. 1960. A problem analysis and [reprint: Fire Management Today. 60(1): spruce budworm-damaged forests of proposed research program for the 6–7]. Ontario. Forestry Chronicle. 63: 8–14. Southern Forest Fire Laboratory. Macon, Helms, J.A. 1979. Positive effects of pre­ Stocks, B.J.; Flannigan, M.D. 1987. GA: USDA Forest Service, Southeastern scribed burning on wildfire intensities. Analysis of the behavior and associated Forest Experiment Station, Southern Fire Management Notes. 40(3): 10–13. weather for a 1986 northwestern Forest Fire Laboratory. Johnson, V.J. 1964. Chronology and analy­ Ontario wildfire: Red Lake No. 7. In: Chandler, C.C. 1976. Meteorological needs sis of the Hughes Fire, 1962. Res. Note Postprint Volume, Ninth Conference on of fire danger and fire behavior. In: NOR–8. Juneau, AK: USDA Forest Fire and Forest Meteorology; 1987 April Baker, D.H.; Fosberg, M.A., tech. coords. Service, Northern Forest Experiment 21–24; San Diego, CA. Boston, MA: Proceedings of the Fourth National Station. American Meteorological Society: Conference on Fire and Forest Lindenmuth, A.W., Jr.; Davis, J.R. 1973. 94–100. Meteorology; 1976 November 16–18; St. Predicting fire spread in Arizona’s oak Stocks, B.J.; Walker, J.D. 1973. Climatic Louis, MO. Gen. Tech. Rep. RM–32. Fort chaparral. Res. Pap. RM–101. Fort conditions before and during four signif­ Collins, CO: USDA Forest Service, Rocky Collins, CO: USDA Forest Service, Rocky icant forest fire situations in Ontario. Mountain Forest and Range Experiment Mountain Forest and Range Experiment Inf. Rep. O–X–187. Sault Ste. Marie, ON: Station: 38–41. Station. Canadian Forestry Service, Great Lakes Cheney, N.P.; Gould, J.S.; Catchpole, W.R. McCaw, L.; Maher, T.; Gillen, K. 1992. Forest Research Centre. 1998. Prediction of fire spread in grass­ Wildfires in the Fitzgerald River National Street, R.B.; Alexander, M.E. 1980. lands. International Journal of Wildland Park, Western Australia, December 1989. Synoptic weather associated with five Fire. 8: 1–13. Tech. Rep. No. 26. Perth, WA: major forest fires in Pukaskwa National Custer, G.; Thorsen, J. 1996. Stand-replace­ Department of Conservation and Land Park. Int. Rep. SSD–80–2. Toronto, ON: ment burn in the Ocala National Management. Environment Canada, Atmospheric Forest—A success. Fire Management McCaw, L.; Simpson, G.; Mair, G. 1992. Environment Service, Ontario Region. Notes. 56(2): 7–12. Extreme wildfire behaviour in 3-year-old Thomas, D.A. 1991. The Old Faithful Inn DeCoste, J.H.; Sackett, S.S. 1966. Baptism fuels in a Western Australian mixed euca­ fire run of September 7, 1988. In: by fire. Southern Lumberman. lyptus forest. Australian Forestry. 55: Andrews, P.L.; Potts, D.F., eds. 213(2656): 169–170. 107–117. Proceedings of the 11th Conference on Euler, D.H. 1946. The sand box as a fire- Nelson, R.M., Jr. 1993. Byram’s derivation Fire and Forest Meteorology; 1991 April control training tool. Fire Control Notes. of the energy criterion for forest and 16–19; Missoula, MT. SAF Publ. 91–04. 7(1): 37–39. wildland fires. International Journal of Bethesda, MD: Society of American Fogarty, L.G.; Jackson, A.F.; Lindsay, W.T. Wildland Fire. 3: 131–138. Foresters: 272–280. 1996. Fire behaviour, suppression and Norum, R.A. 1982. Predicting wildfire Thomas, D. 1994. A case for fire behavior lessons from the Berwick Forest Fire of behavior in black spruce forests of case studies. Wildfire. 3(3): 45, 47. 26 February 1995. FRI Bull. No. 197, For. Alaska. Res. Note PNW–401. Portland, Thomas, D.; Cook, W. 2002. Dude Fire Rural Fire Sci. Tech. Ser. Rep. No. 3. OR: USDA Forest Service, Pacific staff ride. Fire Management Today. 62(4): Rotorua and Wellington, NZ: New Northwest Forest and Range Experiment 4–5. Zealand Forest Research Institute and Station. Traylor, R.E. 1961. Correlation of weather National Rural Fire Authority. Pearce, H.G.; Alexander, M.E. 1994. Fire to fire spread in grass and brushland Forestry Canada Fire Danger Group. 1992. danger ratings associated with New fuel types on the Snake River Plains of Development and structure of the Zealand’s major pine plantation wildfires. southern Idaho. M.S. thesis. Missoula, Canadian Forest Fire Behavior Prediction In: Proceedings of the 12th Conference MT: Montana State University. System. Inf. Rep. ST–X–3. Ottawa, ON: on Fire and Forest Meteorology; 1993 Turner, J.A.; Lillywhite, J.W.; Pieslak, Z. Forestry Canada, Science and October 26–28; Jekyll Island, GA. SAF 1961. Forecasting for forest fire services. Sustainable Development Directorate. Publ. 94–02. Bethesda, MD: Society of Tech. Note No. 42. Geneva, Switzerland: Garvin, D.A. 2000. Learning in action: A American Foresters: 534–543. World Meteorological Organization. guide to putting the learning organiza­ Rothermel, R.C. 1991. Predicting behavior Van Wagner, C.E. 1971. Two solitudes in tion to work. Boston, MA: Harvard and size of crown fires in the Northern forest fire research. Inf. Rep. PS–X–29. Business School Press. Rocky Mountains. Res. Pap. INT–438. Chalk River, ON. Canadian Forestry Gisborne, H.T. 1929. The complicated con­ Ogden, UT: USDA Forest Service, Service, Petawawa Forest Experiment trols of fire behavior. Journal of Forestry. Intermountain Research Station. Station. 27: 311–312. Rothermel, R.C. 1993. Mann Gulch Fire: A Weick, K.E.; Sutcliffe, K.M. 2001. Gisborne, H.T. 1948. Fundamentals of fire race that couldn’t be won. Gen. Tech. Managing the unexpected: Assuring high behavior. Fire Control Notes. 9(1): 13–24. Rep. INT–299. Ogden, UT: USDA Forest performance in an age of complexity. Haines, D.A.; Main, W.A.; Simard, A.J. 1986. Service, Intermountain Research Station. San Francisco, CA: Jossey-Bass—A Wiley Fire-danger rating and observed wildfire Sackett, S.S.; DeCoste, J.H. 1967. A new Company. ■ behavior in the northeastern United mobile fire laboratory. Fire Control States. Res. Pap. NC–274. St. Paul, MN: Notes. 28(4): 7–9. USDA Forest Service, North Central Forest Experiment Station.

Fire Management Today 8 BLACKWATER FIRE ON THE SHOSHONE*

USDA Forest Service, Division of Fire Control

reliminary reports in hand as this issue goes to press show When the heavy wind started sweeping P that initial action on this light­ this way and that on Saturday, August 21, ning fire was alert, prompt and vig­ fifteen men lost their lives. orous—quite remarkably so, con­ sidering that the Shoshone National Forest is rated as a low- danger forest, and doesn’t even largely ignore it. But such fire always with a running accompani­ have lookout stations. The country accidents do happen and impress ment of a plan (sometimes uncon­ was high and steep—just below us all the more because of their scious) for the best way to safety timber line. In spots the lodgepole infrequency. This is the largest loss for the crew if something went pine and fir were dense and of life from a single National Forest wrong. limby—the familiar patches of fire since 1910. It is the irony of rather scrubby jungle found on the fate that it had to occur on a His statement is also a technical better sites at high elevations. National Forest which, so far as case history of the handling of men There were steep slopes covered can be determined from the in such crises. His record could be with dense but not jungly stands— records here, has had only one followed better with the aid of a just the setting for wind-driven other large fire during its whole map, but even without it much can crown fires of intense heat. history. The latest reports on size be gleaned from the story. Pictures of the area show bare of this fire put it at 1,100 acres ridge tops and open places here (445 ha). Heroic Conduct and there. Fuel on the ground As a record of unassuming heroic seems to have been quite light—as To the men who died in this disas­ conduct the statement needs no would be normal under such con­ ter, all fire control men everywhere comment. It was dictated straight­ ditions. One would guess that the pay tribute. To the bereaved fami­ away in the presence of D.P. fuel experts would rate the area at lies they extend the deepest sympa­ Godwin, with no rehashing or edit­ “Low rate of spread” and “Low thy. To the survivors, and particu­ ing except the correction of the resistance to control.” But when larly the exceptionally large num­ spelling of one name and the inser­ the “heavy” wind started sweeping ber of injured men, is extended tion of the name of Bert Sullivan this way and that on Saturday, appreciation and cordial concern in one place. Post has some bad August 21, fifteen men lost their from all those engaged in the high burns, and both hands and both lives. Six of these died from their adventure of protecting American sides of his face are heavily band­ burns after the blow-up. forests from devastation by fire. aged. He is out of danger, but will bear scars. Large Loss of Life District Ranger Post’s statement The danger from such accidents (excerpted in the sidebar on page Junior Forester Tyrrell of Ranger probably is statistically less than 10) is published as an authentic Post’s party died later from his the danger from automobile acci­ case record of the processes of burns. In speaking of him in his dents, which is so familiar we judgment in such situations where signed statement, Enrollee Alcario a man must think first and think Serros says: clearly about the safety of the men The Division of Fire Control in the USDA in his crew. His words will recall to Then we saw that we didn’t have Forest Service’s national office was the all experienced men many days of no chance to go back, so Ranger precursor of the agency’s Fire and Aviation harassed effort to get the line Post told Mr. Tyrrell to take care of Management Staff, Washington Office, ahead and the fire mopped up Washington, DC. us, and he took us up to the rim before something happened—but rock. The fire started from the * The article is reprinted from Fire Control Notes 1(5) [20 September 1937]: 305–306.

Volume 63 • No. 3 • Summer 2003 9 east, and then south, and then the west. It was the west fire that burned us. As the fire came closer to us we layed down on the rock ridge. Mr. Tyrrell layed on top of me. When the fire burned Mr. Tyrrell he ran and I ran, too, about 10 feet.

District Ranger Clayton, whose mes­ sage Ranger Post received, died with six of his men. A seventh got out, but died from his burns. ■

When cut off from the safety sought above timber line (center background), Ranger Post and 40 men who followed his instructions survived by taking what shelter they could, first on the left (north) side, then the right side of the bare spot on the ridge (center foreground). Junior Forester Tyrrell died later from burns inflicted by the flames and heat that swept over them.

The Blowup Begins

The following eyewitness account* describes the onset of the blowup on the Blackwater Fire on the after­ noon of August 21, 1937. The blowup resulted from the passage of a dry cold front.

Up until this time no wind was in I ran for some distance to the directly into the line of men on evidence. Almost like a shot out northwest and climbed a rocky line construction. In a few sec- of a gun, there was a heavy wind. point and saw below me a spot fire onds numerous fires appeared It swept through the area in as of considerable size burning to the below the line.... Almost at once near as I can determine in a northeast and around a ridge to it was clearly evident that further northeasterly direction, this car- the north of us. My impression was attempts at line construction in ried sparks over the constructed that this fire should be immediate- that area was out of the question. line and below us. I heard a fire ly taken care of and possibly aban- I sent out a call for all men to roar to the northwest and it don work on our line in order to abandon their work and proceed appeared to be a considerable dis- do it. to the ridge to the northeast. tance away. I called to Tyrrell and This was approximately three told him that something was I turned to summon help for this o’clock, P.M. going wrong and that I was going job when there was decided change to investigate. in the wind again and the spot fire U.J. Post, District Forest Ranger was swept into the southwest August 27, 1937 * Excerpt from “Statement by Ranger Urban J. Post,” Fire Control Notes 1(5) [20 September 1937]: 308–315. Photos taken of the blowup can be found in E. Kauffman, “Death in Blackwater Canyon,” American Forests 43(11) [November 1937]: 534–540.

Fire Management Today 10 THE FACTORS AND CIRCUMSTANCES THAT LED TO THE BLACKWATER FIRE TRAGEDY*

A.A. Brown

Original editor’s note: Included as a vital part of the full report on the Blackwater Fire was the report made by A.A. Brown after an exhaustive study of the fire behavior and the critical circumstances which converged to bring about the tragedy. Mr. Brown had been only recently transferred to Region 2 to head the fire control work of that Region, and he brought to the tasks involved in this disaster the sound knowledge and discern­ ment springing from long and successful fire experience.

Four Critical Factors hile no proof is available, The nature and circumstances the nature and circum­ of the blow-up on August 21 indicate W stances of the blow-up on that there were four critical factors involved. August 21 seem to indicate that an undiscovered spot fire, probably from the night before, to the north ragged edge to the burning area on Strong Wind of Clayton Gulch and over the the steep downhill side, with small The relative humidity at 1 p.m. on small, sharp ridge in Clayton Creek spots below the general front. As a August 21 was 6 per cent, with a (one-half mile [0.8 km] west and result the fire fighters found it temperature of 90 ºF (32 ºC) at the north of the point where the expedient to connect the fire line fire danger station at the Wapiti Clayton group was later found), below the hottest spots, leaving Civilian Conservation Corps (CCC) was the first critical factor in mak­ considerable unburned surface fuel camp, two and one-half miles (4 ing the trap in which the men were inside the line at the lowest point. km) away at 2,000 feet (610 m) caught and burned to death on lower elevation. At approximately that day. Apparently this spot fire The fourth critical element was the 3:30 p.m., with these critical cir­ at first spread up the slope imme­ nature of the forest fuel in this cumstances prevailing, a strong diately above to the northeast. This drainage. It consisted of a very gusty wind of apparently at least is clearly indicated by the note dense stagnated stand of Douglas 30 miles’ velocity per hour Ranger Clayton sent to Ranger fir with a varying mixture—5 to 15 (48 km/h) came up from the south­ Post at the time the spot fire per cent—of spruce and of alpine west. About 3:45 p.m. it swerved spread conspicuously just prior to fir. A dense overhead canopy exist­ and became a west wind. (These the blow-up. ed, with dead branches nearly to times are based on the circumstan­ the ground, with many small, tial evidence of other events of the The second critical circumstance brushy, dead or nearly dead sup­ fire.) was the fact that the timber above pressed trees as an understory, a the newly constructed line had not considerable volume of sound dead The duration of this strong velocity crowned out except for a small branches, logs and suppressed is uncertain because of the strong fringe along the south edge. The trees on the ground, and with vary­ convectional winds set up almost third critical circumstance was the ing amounts of moss throughout at once by the crowning. It is rea­ fact that “spotting” from the fire of the canopy and on all the dead sonable to suppose that the change the previous day had given a branches. In addition, slopes of in direction may have been largely 20 to 60 per cent prevailed. a convectional effect. At the start, timber began to crown above the When this article was originally published, These four factors set the stage for A.A. Brown was head of Fire Control for line and the whole fire there began the USDA Forest Service, Rocky Mountain what happened. to pick up in intensity and to Region. throw new spots below the line, as * The article is reprinted from Fire Control Notes 1(6) might be expected. Possibly this [December 6, 1937]: 384–387.

Volume 63 • No. 3 • Summer 2003 11 exerted a strong convectional pull The two crown fires then rapidly closed together on the spot fire below, which had with the cyclonic effect of such a circumstance, also begun to crown. sweeping clean the major portion of the head At any rate, the course of the drive of the Clayton Creek drainage. from the spot fire changed to the east and started directly up the drainage. The two crown fires then [322 m]) at the time the blow-up Trail Ridge and also called out the rapidly closed together with the occurred. men from the fire trail north of cyclonic effect of such a circum­ Trail Ridge. Pierce, who had been stance, which reached its climax at Clayton, who had been placed in near the highest point of the fire 4:20 p.m. As a result, the major charge as sector boss of the new trail before it dipped down into portion of the head of the Clayton construction, was following the Clayton Gulch, had already come Creek drainage from the spot fire fresh crew in and checking up on out to the first small park with the up to Double Mountain was swept conditions as he went. Apparently two CCC boys helping him because clean in a final crown fire confla­ he was checking particularly on of a flash of crowning just below gration which was completed by spot fires. The BPR crew were giv­ him, which apparently crossed the approximately 5 p.m. ing most of their attention to spot fire line but died down again at the fires at two points below the 30 little park. He attempted to get the In this conflagration 9 deaths chains (1,980 feet [603 m]) of line attention of the rest of the BPR occurred directly. Six additional they had constructed. They were crew, but, receiving no answer, men were so badly burned that about 20 chains (1,320 feet [402 decided they were withdrawing the death ensued, and 36 additional m]) in from Trail Ridge, except for other way. men suffered injuries from which Pierce, one of their members who they are recovering. had been left alone on hot line at a About 6 other CCC boys were also point about 10 chains (660 feet assembled at the park, and all came Firefighter Movements* [201 m]) in, where several logs out together at Wolcott’s alarm were on fire close to the fire call. The heavy crowning apparent­ Just before the crowning started, trench. Two men were left to help ly occurred shortly afterward the distribution of men on the him as Post’s crew came past and (about 4 p.m.). Post’s attention was newly constructed line in Clayton about 6 men were left with Saban attracted to the spot fire when it Creek had been as follows: Five and Clayton to work on spot fires. started crowning toward the north­ men of the National Park Service east up the slope on the north side crew on mop-up were operating as By the time this distribution was of Clayton Creek. His first thought far as the first small park, about 5 completed, about 3:45 p.m., Post, was to take his crew to it, but the chains (330 feet [101 m]) northeast Clayton, and Fifield, probably wind changed and the fire started of the ridge. Beyond them for 30 simultaneously, saw evidence of an up the gulch before he could take chains (1,980 feet [603 m]) were 6 uncontrolled spot fire. Fifield, any action to that end. Accordingly men of the BPR crew, who had according to his statement, was on he started moving his crew from been actively pushing the new line the rock point of Trail Ridge at the its path as best he could, as construction from this point on, time and thought at first that it described in his statement. and who had got as far as Clayton was the spot near the bottom of Gulch, plus a few men who had the first gulch which had previous­ Death Trap been dropped off from Post’s crew. ly been found and trenched, but Clayton’s movements are not so Beyond them were Ranger Post, discovered instead that it was in with Foremen Tyrrell and Saban clear in detail. It is evident from line with it, but over the small the note he dispatched to Post that and sub-foreman Hale, with about ridge just to the north. He at once 40 men who had taken up the new the spot fire had attracted his gave thought to Wolcott’s crew, attention, apparently from on the line at Clayton Gulch and had who were in this vicinity, but completed 16 chains (1,056 feet spur ridge just south of the gulch, found them coming out on where he was later trapped. Up to account of the crowning there. the time this fire started directly *See the maps on page 14 (from the article by Wolcott immediately went on up up the gulch it was a threat to the D.P. Godwin referenced in the sidebar on page 13).

Fire Management Today 12 With the direction of the fire’s path tions alone, without the spot fire, directly up the gulch, it probably acted as a would have created a dangerous situation, and would have no doubt furnace draft and became a death trap. forced abandonment of the newly constructed fire line, but without loss of life, since distances to safety line above which [it] must be Once off the ridge the full poten­ were not great. Exactly the same stopped, but probably did not tialities of the fire below would not strategy employed would likely appear to be dangerous to life. have been immediately as evident have succeeded without incident a Clayton saw it was the focal point if to him as it was to Post above. few hours earlier, or perhaps even the line were to be held, and that Presumably, as soon as he saw at that time of day if no sudden more men than the 7 with him what was about to happen, he change in wind velocity or direc­ would be required. turned back to get the men at the tion had occurred. spring. In doing so he had to go It does not seem likely that he back up the slope ahead of the fire Had the spot fire not been in line waited on the ridge above Clayton and probably got to his men just at with one already controlled, or had Gulch the 20 minutes or more that the time that every avenue of not been hidden by the sharp little seems to have elapsed from the escape was cut off. Had the fire ridge in the bottom of Clayton first active spreading of the spot been going across topography the Creek, it would have had earlier fire below until the general blow­ bare gulch might have served satis­ attention from the Park crew from up occurred. Probably he started factorily as a refuge. With the Trail Ridge, and again the situation down toward it, either with his direction of its path directly up the would have been changed. group or alone. If alone, he proba­ gulch, it probably acted as a fur­ bly left instructions for his group nace draft and became a death trap. More time on the part of either to await his investigation of poten­ Clayton or Post to fully scout out tialities below. Or, if he took his Reconstructed Tragedy the potentialities of the fire ahead group, he probably left one or In conclusion, the reconstructed of the crews might have prevented more men at the spring in the tragedy depended on each of the the tragedy. gulch to await reinforcements from four factors first discussed which Post. In either case, the natural contributed to the behavior of the Earlier arrival of the new crew, route of travel toward the spot fire fire, plus the distribution and even by as little as a half hour, would be down the gulch toward it. movement of the men at the time. would have resulted in completing The high wind and burning condi- the new line and would have con­ centrated the attention of all supervising officers and man-power Other Articles on the on all threats to holding it. This Blackwater Fire would have resulted in a different distribution of the crews and prob­ Godwin, D.P. 1937. The handling of the Blackwater Fire. Fire Control ably slight danger. Many other Notes. 1(6) [6 December 1937]: 373–383. premises may also be drawn, but the matter of timing of action of Brauneis, K. 2002. 1937 Blackwater Fire investigation: Boost for the fire vs. movement of men smokejumpers? Fire Management Today. 62(2): 24–26. gave the distinctive and fatal combination. ■

Volume 63 • No. 3 • Summer 2003 13 Aerial view of Blackwater Fire, looking southeast directly up the head of Clayton Creek. Dashed line indicates the lost control line. Center arrow shows the gulch where Ranger Clayton and seven men were trapped. Arrow toward the upper left shows the open point on the ridge where Ranger Post took refuge with his crew.

Fire Management Today 14 LESSONS FROM LARGER FIRES ON NATIONAL FORESTS, 1938*

Roy Headley

Editor’s note: Roy Headley took the “lessons learned” from field Failure to study the action on larger fires reports on “larger fires” (fires cov­ invites weaknesses that will let big ones ering more than 300 acres [121 grow into bigger ones. ha]). In places, he added com­ ments, indicated by brackets and the italicized notation “–Ed.” in one fire. Seventy-five thousand The first essential in such matters Among fire control men there is a acres (30,000 ha) were lost in the is to grasp the need for change, the real difference of opinion on the four largest fires. nature and importance of a prob­ extent to which study should be lem, the chance to introduce concentrated on the larger fires as Lessons Learned something better. With that fact in distinguished from all fires. Some Such an infinite variety of prob­ mind, the outline for 1938 reports believe that concentration of atten­ lems are involved in the manage­ on larger fires requested a record tion on large fires is dangerous and ment of large fire jobs that of lessons learned by the man or may result in neglect of the meth­ thoughtful men seldom fail to men who had most to do with each ods and principles which are of learn from each one something fire. Some of the most suggestive importance in keeping small fires which should be guarded against answers received are quoted in this from becoming larger. in the future, something which article. Perhaps these notes will should be done differently, some help reduce the number of times The answer is probably a matter of cherished belief which must be lessons have to be “relearned” by proper balance. Both classes modified or abandoned. For 35 different men—or by the same should be studied. Failure to years I have been working on or men. review the action taken on even observing suppression jobs, but I the smallest fires invites weakness­ still learn something from every Northern Region (R-1) es that will let more little ones fire I reach. Absaroka—Chico fire—429 acres grow into big ones. But larger fires (174 ha). The fire got over the have characteristics peculiar to Sometimes, alas, we “learn the line because of incomplete mop-up themselves. Management of the same lesson over and over”—or do and men being gathered in a larger fire jobs is a rather distinct we? For example, I have learned bunch to get water. The foreman branch of control—and a backward throughout many years that there in charge was a young, inexperi­ branch. Failure to study the action is some flaw in our management of enced administrative guard who, in on these larger fires invites weak­ larger fires which keeps us from his zeal to be helpful, left his crew nesses that will let big ones grow getting a reasonable output of held to help on another part of the line. into bigger ones. These bigger ones line from a crew of a given size. Had he had more experience, he hurt, too. Out of a total of 219,000 Plenty of other people have learned would have recognized that his acres (88,000 ha) burned in 1938, the same thing. But, untrained as crew lacked training and knowl­ 35,000 acres (14,000 ha) were lost we are in the science and art of edge of fire fighting, and he would management, we have not found not have left them. This might also ways to act satisfactorily on what be traced to the fact that the sector When this article was originally published, Roy Headley headed the Division of Fire we have learned. Our learning has was too long for efficient handling Control, USDA Forest Service, Washington, too often failed to lead to produc­ by the sector boss. The country DC. tive action. was so steep that the man in

* The article is an abridged version of sequential arti­ charge could not cover his whole cles published in Fire Control Notes 3(3) [Summer line often enough. 1939]: 6–17 and 3(4) [Fall 1939]: 30–45.

Volume 63 • No. 3 • Summer 2003 15 For 35 years I have been working on Splendid cooperation with the or observing suppression jobs, but I still learn United States Weather Bureau in something from every fire I reach. their special fire-weather forecasts during the entire fire season, and specifically during the progress of this fire, made possible a more Kaniksu—Goose Creek fire—544 conditions that existed at the time accurate planning program for fire acres (220 ha). An organization of this fire. However, a spot fire got personnel needs and strategy to be principle was violated in that away, and caused the second run; employed (particularly on backfire authority on the ground was divid­ therefore, the number of men or work). ed between two heads and coordi­ the organization of them must nation was not effected. Two men, have been at fault. Records show and this fire con­ either qualified to handle the firms: Our late fall fires are a prob­ entire fire, were on the job. Each Bulldozers were used to construct lem we have not met in southern assumed the other to be in charge. approximately three-quarters of a California. In the absence of early Each aggressively worked on oppo­ mile (1.2 km) of line on this fire. rains, in September and October site sides of the fire, expecting the The main criticism was on account we get Santa Anna conditions, other fellow to see that balance of some of the burning fuel having extremely low humidity and high was maintained. been covered with duff and dirt wind velocities up to 50 miles per which added hours to the time hour (80 km/h). These conditions The men planning the attack were required to do the mop-up job. sometimes last a few days—some­ confronted by a fire burning very times 3 weeks. Practically all seri­ fast and throwing spot fires ahead. The time of arrival, the rate of ous fires in southern California for Fire was in a heavy, dry, white-pine spread, and the size and value of the last 10 years have started under slash area which had been logged the rapid-spread area determined these conditions, indicating we during the summer and was the action to take. I believe a bull­ have pointed our efforts to the nor­ pushed by a 20-mile (32-km/h) dozer line under not too difficult mal season conditions and gained wind. It was figured that the fire topographic conditions can be con­ much ground, but now we must would travel in a northeasterly structed in less time than any point to these abnormal fall condi­ direction, which it did to a certain other machine we have on hand. tions and plan to meet them by: extent, but not as fast as was antic­ The cost of mop-up is, of course, a ipated. Consequently, more men problem, but corral is the fist job. 1. Study more intensively, the were figured for the northeast side Because of the value of bulldozers, behavior of these late fall fires. than actually needed. If our fore­ and the possibility of breakdowns, 2. Provide a 24-hour patrol and sight had been as good as our I recommend an extra machine on lookout service. This was a night backsight, more men would have a fire-line construction job. fire—as have been two or three been placed on the west side where others. there were barely enough men. California Region (R-5) 3. Make more intensive use of clo­ Angeles—San Antonio fire— sures regardless of private prop­ Kootenai—Rocky Creek fire—380 3,270 acres (1,323 ha). We were erty interests. acres (154 ha). The old story that slow in completing lines on some 4. Shorten elapsed time between an undetected spot fire blew up sectors, because of using crews too discovery and first action by and caused this to be an extra-peri­ large—20 men per experienced movement of equipment and od fire holds true here. crew leader. We should have (for men so as to concentrate in high this particular fire, and it is true in hazard areas and high occur­ In an analysis of the action on this most fall fires that spread with rence zones. fire by the supervisor, administra­ erratic lines and many spots) used 5. Intensify use of secondary look­ tive assistant, two rangers, and an small crews, 5 men to each trained out points. alternate and dispatcher who were fire fighter, and made sectors small 6. Have night suppression crew on on this fire, it was agreed that the enough to provide closer supervi­ duty, dressed and ready to go. 25 men who were searching for sion and more effective perform­ spot fires were sufficient under the ance.

Fire Management Today 16 In summary: Put additional heat The large area and rapid spread of this fire on every control effort we normal- was due entirely to the dense and continuous ly practice for regular summer stand of “cheat grass.” season.

Klamath—Red Cap fire—16,196 acres (6,554 ha). The suppres­ on October 1 I was forced to dis­ would have reduced the size of this sion action on this fire was charac­ miss them with the expressed hope particular fire. However, it was terized by insufficient manpower that they could find enough to do thought that we were deficient in and overhead in the first five burn­ through the winter to come back the number of tank trucks avail­ ing periods, then a sudden build­ mentally and physically fit to tack­ able at the beginning of the fire up of men in a belated attempt to le another season. and during the fire. A tank truck at conform to the Forester’s policy of Arrowhead ranger station would be immediate control. This sudden San Bernardino—Arrowhead very desirable and had there been build-up led to placing more men fire—12,362 acres (5,003 ha). one there it would have arrived on in the field than could be ade­ The Arrowhead fire originally start­ the fire about 7 minutes after the quately serviced—with consequent ed in a cabin on the crest of the fire was reported instead of 36 loss of efficiency. For example, on San Bernardino Mountain range. minutes as was the case. Even divisions 1 and 2, zone B, it took After very thorough investigation though there had been a tank 400 men (including camp workers, as to the cause of the fire, it is truck at Arrowhead ranger station, etc.) 4 days to build and backfire believed that it was due to faulty it is doubtful that it would have 404 chains (26,664 feet [8,127 m]) flue construction. The defect was made any difference in this fire. of line. Theoretically, 400 men believed to have been where the But as a presuppression lesson we should have built the line in 1 day, thimble fastened to the brick can see the importance of having which would have been in accor­ chimney. favorable distribution of tank dance with the Forester’s policy; trucks. but actually, they lost 50 percent Extreme fire weather conditions efficiency because of lack of food existed at the beginning of this fire All line constructed and lost was and bedding; and it is estimated and continued throughout. A high uncompleted line. All backfire that about half of the remaining wind was blowing at gale force, 45 work that was done was held effort was dissipated in too wide to 50 miles per hour (72–80 km/h), although it slopped over in places. line construction, so 100 men fully and the humidity was very low, Orchard torches were used. No serviced and adequately supervised about 7 percent. acreage was burned through back­ could have accomplished the same firing which would not have been job in the same time. The cabin in which the fire started lost by the fire anyway. had wooden, untreated shingles, The Weather Bureau fire-weather also others that burned near this Sequoia—Fish Hatchery fire—500 field station forecasts were used one. Due to the high winds the acres (200 ha). Investigation throughout the fire. The assurance burning shingles were probably the shows that regardless of any sup­ of this station that certain favor­ greater cause of the many spot pression action taken after the able conditions would continue fires. crews arrived on the fire, it could made it possible to plan and con­ not have been controlled while still struct a line into the head of the The rate of spread of this fire was small. This fire was simply moving Red Cap Canyon, thereby saving extremely great during the first too fast after it got up a headway. It about 4,000 acres (1,600 ha) that few hours. Approximately 8,000 is barely possible that had we been had appeared to be doomed. acres (3,200 ha) burned during the able to put about 50 men on this first 7 hours or over 1,000 acres fire at the time the first crews The marvelous “do or die, stick to per hour (400 ha/h). arrived, the fire could have been the bitter end” spirit of the short- held to a small acreage. However, term force was magnificent. It was Because of the high rate of spread there was no failure in first attack with a feeling of deepest regret it is doubtful that any means of action, as the first crew traveling that with the first big rain storm preparedness in presuppression 10 miles (16 km), arrived 17 min-

Volume 63 • No. 3 • Summer 2003 17 Records show and this fire confirms: the size it did as a direct result of Our Santa Ana fires are a problem crowning through the unburned we have not met in southern California. canopy on the day following con- trol. Past records will show that this is a common occurrence on fires in the Canyon Creek areas, utes after the origin of the fire. We than I had thought possible, and and a special effort had been made feel, therefore, that the only way as a consequence, the fire made an to mop up the entire fire before the this fire could have been prevented advance beyond the point where it burning period of the day following from becoming large would have was estimated it would be held. control. been for the district ranger to have taken such action earlier in the Shasta—Big Lake fire—745 acres In spite of the efforts of 65 men season as would have prevented (301 ha). The most important with backpack pumps and a power the fire from starting. This might thing I learned on this fire was pumper, the fire flared up from have been possible had he specifi­ that in slash type cover one should inside and crowned out in the cally designated a dumping ground never try to attack the head of a unburned canopy of live oak. for the use of the Kern Country fire if it reaches an area of more Juvenile Camp and had this than 3 acres (1.2 ha). To attack the If there is any lesson to be learned ground thoroughly fire-proofed at rear first and pinch it in on both from this experience, I believe it is the stated of the season. sides, putting the heaviest attack the necessity of having enough on the side toward which the wind men with water equipment to The fire was in very rough country is blowing. cover the entire fire before the where night work was difficult. beginning of the next burning peri­ District Ranger Stathem now Shasta—Mount Hedron fire— od. This is a rather difficult and believes that it would have been 8,300 acres (3,400 ha). This fire costly procedure, especially when better business to have reduced his originated in a grass sage–juniper control is not achieved before day­ night crews in size, concentrating fuel type in gently rolling country light. The flare up on this fire most of his manpower on the line at a time when the wind was blow­ began at 12 noon, which did not during daylight hours. ing approximately 20 miles per allow much time to cover the hour (32 km/h). The fire was entire fire. Also, it is not advisable The lower fire line was built paral­ attacked by 20 men and 2 tank to send men into the burn before it leling an oiled road and in some trucks within 11 minutes of its has had time to cool sufficiently to places within 800 feet (240 m) of origin, and at that time it was allow them to work safely. this road. The reason for this was about 3 acres (1.2 ha) in size and that it was believed at the time spreading rapidly. Cooperators Control of this fire the day follow­ that it was necessary to do this to were immediately called upon for ing was achieved by following the protect a high-voltage transmis­ help, but by the time they arrived above procedure to a large extent. sion line that was between the fire the fire was completely out of con­ However, weather conditions were line and the road. Looking back on trol and was heading north before more favorable until the late after­ this, it is easily seen, however, that a strong tail win. noon, which allowed more time to there was an opportunity to drop mop up. back to the road and hold the fire The action of the fire indicated and also protect the transmission that it probably would have been a In summarizing the above, perhaps line with a much smaller crew, 300-acre (120-ha) fire even though the most important point is to inasmuch as the line would have there had been a full camper of achieve control as early as possible been easier to hold on the road. CCC enrollees at Leaf which was the first night so that the fire will only 8 miles (12 km) away. It cool down before daylight, and Shasta—Salt Creek fire—1,690 seemed to be one of those fires allow more time for mop up. acres (684 ha). The most impor­ that had to make its run. tant lesson that I learned from this Trinity—Little Bear Wallow Fir— fire was that it burned more rapid­ Trinity—Glennison Fire—370 2,200 acres (890 ha). Civilian ly at night, especially downhill, acres (150 ha). This fire reached Conservation Corps (CCC)

Fire Management Today 18 enrollees are ineffective on a fire An important principle is that when not thoroughly trained in fire the front of a fast running fire is often suppression, and they must be untouchable. But a lull always comes. accustomed to hard hiking in rugged country and able to do The job is to be sure to grab it effective work after arriving at the for keeps during that first lull. fire. In the case of this fire, the ini­ tial attack was made by an experi­ enced guard and two green for several days, which clearly 1. Must keep roads cleared as fire enrollees. The boys, who were brought out the fact that pack breaks. transferred here with a New York stock must be relied upon at times. 2. Must construct additional cross company a day or so previously, roads. were soft, untrained, and not More use should be made of the 3. Must make definite plans for accustomed to hiking, and not only indirect method of control on fires suppression (outline prepared). slowed down the guard’s travel of this size and in this type of 4. Overhead, in general, needs time to the fire, but were useless topography. There is always the more “mop up” training. after arriving there. human aversion to burned area 5. CCC enrollees must be shown that is continually cropping up on “how” by fire foremen and straw North Pacific Region fires of this kind.* Men work too bosses. (R-6) close to the fire, thus sacrificing 6. On very large fires it is better to Siskiyou—Cedar Camp (Chetco) time and labor. More satisfactory use machinery for trench con­ Fire—34,627 acres (14,014 ha). results can be obtained by drop­ struction at the sacrifice of Something must be done to stop ping back and taking advantage of some acreage than to depend on incendiary fires from being started natural breaks in topography. hand work too close to the fires. to keep large fires burning. On the division in which I worked, it is a A stream, unless it is large and Investigation of the cause of break- positive fact that one fire was set, clear of brush, makes a poor fire over on August 26 showed that and I am relatively sure that others line. The variable winds up and while a good trench and clean were set in the same area. Take the down streams, the possibility of burn had been obtained on the local man’s advice as to fire behav­ rolling logs and falling snags, and extreme southeast corner of the ior and wind drafts if he is reliable. the rapid spread of a fire up the fire, and mop up had been carried Reliable local men are indispensa­ sloe once it crosses a stream, back of the line an adequate dis­ ble as scouts, especially in unfamil­ makes it exceedingly dangerous to tance for ordinary circumstances, iar country. use. the crowning of a small green pine tree on a low ridge about 300 feet Siskiyou—Nome Fire—5,800 Fremont—Bonanza Fire—9,155 (90 m) inside the line threw sparks acres (2,300 ha). Planes proved to acres (3,705 ha). The large area into the thick cheat grass and be an excellent means of supplying and rapid spread of this fire was juniper timber which quickly fire camps both from the stand­ due entirely to the dense and con­ crowned and spread under the point of speed and economy. Camps tinuous stand of “cheat grass” brisk breeze. Men working in the were serviced that otherwise would (Bromus tectorum) with very little vicinity of the break took action have had to be supplied by man timber overstory. Consequently, almost immediately, but were pack, as was done for several days the usual “forest” fire standards are unable to cope with the heat of the when planes were grounded, which not fully applicable. However, the crowning junipers and did not required 8 hours of hard packing following was learned about this have time to use the indirect by good men to get 40 pounds of particular country. method. supplies into camp. Rocky Mountain * Editor’s note: American Indians routinely used fire Planes will not eliminate the use of for natural resource management, a practice that some Region (R-2) horses. During the time that this European settlers adapted to their needs. But by the Roosevelt—Jenny Lind Fire—664 fire was burning the smoke became 1930s, the culture of fire control was so deeply ingrained that “aversion to burned area” seemed quin­ acres (269 ha). I believe the most so dense the planes were grounded tessentially “human.”

Volume 63 • No. 3 • Summer 2003 19 To summarize, it is apparent that entirely too narrow for safety in extreme effort should be made handling backfiring in windy weather. [Importance of training to satisfy the requirements of the in judgment again emphasized. Forester’s policy as to control in More common error is to make the early burning stages. fire lines wider than need be for backfiring. –Ed.] important point to be noted is the Southwestern Region Intermountain Region great difficulty of insuring ade­ (R-3) (R-4) quate initial action on a very Wasatch—Shepherd Creek Fire— Gila—Lookout Ridge Fire—575 aggressive fire in its initial stages. 960 acres (388 ha). Particularly acres (233 ha). Class 6 burning [Three miners, 100 feet (30 m) significant is that the fire, carried conditions prevailed and were not away, saw fire and attacked it 5 by a high wind, spread over 700 recognized. The fuels were very minutes after it started. District acres (280 ha) within 2 hours after dry and the wind was strong. The ranger with one man arrived at fire it started. It traveled so fast that lesson learned here is that a fire 35 minutes after it started. But the fire truck, plow unit, and men danger station would have fur­ larger crews did not reach fire in could not keep pace. Forty men nished a definite check on burning time to prevent 39- to 54-mile were on the fire within 20 minutes conditions, and this could then (63–87 km) wind from starting fire after it started, and, with our fire have been promptly followed by on its first run. Probably most truck which suppressed a mile (3.2 proper strengthening of the guard important lesson to be learned is km) of fire edge, and the plow unit force to meet class 6 or emergency that on a forest not provided with that made over 3,000 feet (900 m) conditions. [This forest is now sup­ regular detection and other facili­ of fire line, it was impossible to plied with a fully equipped station. ties common to western fire cope with the high rate of fire Danger stations and danger ratings forests, a fire starting in slash in a spread. [But the wind dropped and are designed to avoid just this sort gale of wind could be and was the fire died down, after which the of failure to recognize changes in stopped at 664 acres (269 ha). –Ed.] crews had a chance and did not fire danger. –Ed.] muff it. An important principle is Almost uniformly there was too that the front of a fast running fire Poor line location due primarily to much tendency to bury burning is often untouchable. But a lull trying to hold fire to a minimum material along the fireline in such always comes. The job is to be sure acreage with subsequent loss of a way that it only prolonged the to grab it for keeps during that excess acreage was an important burning and the mop-up period. first lull. –Ed.] factor. This is merely a matter of The correct use of dirt has been so training in suppression. [Yes, but much emphasized that there are training in judgment, which is the Southern Region (R-8) few if any of the foremen ignorant most difficult and most backward Conecuh—“Big Fire”—576 acres of the undesirability of partially field of training. –Ed.] (233 ha). In the attack on this burying heavy material; but there fire, as in most other fires on the seemed to be great difficulty in Gila—Iron Creek Fire—2,318 Conecuh, (wire-grass type of getting thus particular fault cor­ acres (938 ha). Here again class 6 ground cover), everything depend­ rected. burning conditions existed unrec­ ed upon the success or failure of ognized. Too much line was lost by the tank truck. From the time of A common fault which was per­ working too close to the fire. attack with this truck until the haps the most serious was that of [Understand that this fire and this water supply was exhausted (15 inclosing ragged fire perimeter weakness were used by the region minutes) 19 chains (1,056 feet inside a line without taking any for training by the case method. [321 m]) of line were built. A reor­ measures to clean burn. This –Ed.] Width of fire lines must be ganization has been made to pro­ needs to be given a great deal of governed by (a) fuel on ground; (b) vide for replenishment of the sup­ emphasis on all forests. steepness of slope; (c) wind veloci­ ply of water for the tank truck. ty. In the past, fire lines have been

Fire Management Today 20 A whirlwind picked up burning The most important lesson that I learned material from as far as 100 yards from this fire was that it burned more rapidly (90 m) inside the line and dropped it outside in the rough, resulting at night, especially downhill, immediately in a break-over of than I had thought possible. large proportions. On at least two occasions whirlwinds hit the fire line while suppression was being the construction of permanent the fire was sadly at fault. The day carried on with tank trucks, meandering 8-foot (2.4-m) fire­ was not an extremely bad one; resulting in loss of sort lengths of breaks on the higher ground with­ however, a fresh south wind was line and excessive use of water. in the type will prove very helpful blowing and any fire located within Variability of wind direction and in suppression of any future fires a quarter or a half mile (0.4–0.8 prevalence of whirlwinds are fac­ which may occur. km) south of the boundary certain­ tors not ordinarily considered in ly should have merited both the danger rating. Summarizing, I would say that the fire truck (tank truck and Panama most important thing this fire pump) and a full 16-man crew Evidence indicates there was time showed was the inability of the with the standard fire-fighting tool to have cut a line and backfired ranger to make efficient use of the box. If the junior foreman had had across the head. This would have manpower and equipment that was a full 16-man crew and fire truck speeded up control and promised available at the time of the fire. he would have unquestionably held better chance of holding to a The matters of new and better the fire with only an acre (0.4 ha) smaller acreage. equipment are secondary. [The or so of national forest land ranger may take comfort from the burned. Conecuh—Bradley Fire—792 fact that many others, including acres (321 ha). There is a definite big shots in fire control, suffer DeSoto—Saucier Fire—635 acres need for a mobile tractor-plow from the same inability—but this (257 ha). The weather conditions suppression unit, capable of rapid­ should not deter him from seeking made the hazard extreme. There ly plowing a line from which back­ to be mentally prepared for effi­ had been no rain for 3 days. The firing, could be done. This type of cient management of the next fire. relative humidity was 33 at noon. equipment appears to offer the –Ed.] Wind velocity varied from 8 to 16 only practicable means for stop­ miles per hour (13–26 km/h) dur­ ping the head of the occasional Kisatchie—Slash fire—407 acres ing the fire. The ground cover was bad fire short of natural breaks. (165 ha). Two hundred and ten broomsedge grass. The flames had acres (85 ha) of the fire had only a range of 40 to 60 feet (12–18 m), The relative humidity on the day recently been planted to slash pine. crowning at times and spotting of the fire (December 3) did not The slash pine of course was totally ahead. reach a point that could be called destroyed. The rough was com­ low. This shows that relative posed of wire grass, broomsedge, Instructions were to backfire from humidity, in itself, cannot be and a fairly heavy carpet of hard­ a fire lane on the west side. This relied upon as an indicator of fire wood leaves. The rough was of 5 firebreak had been constructed 2 danger. years’ accumulation, and can be years ago and had not been main­ considered heavy for longleaf type. tained since this time. As a result, Ocala—Pleasant Flat Fire—2,161 it was grown over and offered no acres (875 ha). The ground of Investigation revealed that the fire effective barrier from which to the timber type through which was handled very well by the look­ backfire. The extreme weather this fire burned is generally cov­ out men and the dispatcher. The conditions made it improbable that ered with water during periods of discovery, report, and getaway even if the firebreak had been normal weather, but dries out sev­ times are excellent. However, the recently maintained, the crew eral times a year so that it decision of the superintendent to would have been able to hold their becomes inflammable. Since it is send only 6 men with the junior backfire on this narrow line. not very accessible by tank trucks, foreman in a pickup to check on

Volume 63 • No. 3 • Summer 2003 21 A whirlwind picked up burning material and directed the crews to drop from as far as 100 yards inside the line and back several hundred feet from the fire edge to construct line. The dropped it outside in the rough, resulting technique used, which consisted of immediately in a break-over of large proportions. raking a line and backfiring from it, was considered satisfactory under the circumstances. This DeSoto—Hell Hole Creek Fire— North Central Region dropping back was essential 509 acres (206 ha). This fire (R-7) because whirlwind conditions emphasized how immediate and caused fires to spot 100 to 200 feet Gardner—Tower Fire—489 acres continued work on the flanks of a (30–60) ahead of the main fire. (198 ha). This was one of a series large fire can keep the acreage to a of 25 fires burning on the district minimum and prevent new heads Knowing the conditions as are on this date. The fire danger was from forming when the wind shifts. known now, fire management class 6 and the wind velocity was Crews worked the flanks continual­ could have been improved by back­ 20 miles per hour (32 km/h). The ly so that when the head was final­ firing from the truck trail and fire spread rapidly in highly ly stopped the entire fire was cor­ keeping the fire from crossing it. inflammable leaf and grass fuel. ralled within a very few minutes. However, the use of this technique would have required at least 25 Topography was rugged with steep The head of this fire presented a additional men as there were sepa­ slopes, which accelerated the rate wall of flames 20 feet high (6 m), rate heads traveling toward the of spread. Winds shifted frequently with a range of 30 to 40 feet (9–12 road. and small whirlwinds occurred m). It was burning in a 5-year along the fire front. As a result the rough. Several of the fire fighters The general belief that all fires in fire had a number of heads, with had narrow escapes while attempt­ Missouri hardwoods can be han­ one traveling along the top of each ing to stop the head. Two different dled by direct or parallel methods main ridge. The fire advanced in a pump men were forced to abandon is false. On certain days, conditions solid sheet of flame, leaping 6 feet their pumps in order to outrun the are such that the only practical (2 m) high in most places, and the fire. method is to drop back to natural men could not get near the fire or cultural barriers, such as the because of the intense heat. Spot This fire, as well as other large road in this case, and sacrifice fires, jumping in many cases over ones on the DeSoto, shows that in some burned acreage in order to 100 feet (30 m) ahead of the main extreme weather conditions a great insure control. fire, were common. There was 2 to deal depends on the dispatcher’s 3 years accumulation of leaves on judgment in placing crews, not A more extensive use of water the area. only for the head but for work on should be resorted to on such days. the flanks. The speed at which On this fire three men could not When Assistant Ranger Barry these fires travel offers little oppor­ control a 5-foot (1.5-m) spot fire arrived, he took charge of the fire tunity to correct tactical mistakes by use of rakes. A marine pump immediately. He made a quick on the fire line. would have saved the first line size-up of the situation, deter­ built on this fire. ■ mined that direct attack was inef­ fective [It had already failed. –Ed.],

Fire Management Today 22 LESSONS FROM LARGER FIRES ON NATIONAL FORESTS, 1939*

Roy Headley

Editor’s note: Roy Headley took the “lessons learned” from field The Chestnut Ridge fire showed that weather reports on “larger fires” (fires cov­ conditions may rapidly change during early spring, ering more than 300 acres [121 ha]) in the Eastern, Southern, and and that the field organization North-Central Regions. He indicat- must be kept on its toes. ed the source following each report. His own summaries and comments are indicated by brack­ change very rapidly during the is reported as 2 to 19 miles ets and the italicized notation early spring, and that the field (3.2–31 km/h); in another, 25 to “–Ed.” organization must be kept on its 38 miles (40-61 km/h). Five days toes. –Unsigned forest report since last rain of 0.28 inch (0.71 Eastern Region (R-9) cm). Number of chains per man George Washington—Chestnut A complete failure of district hour, 2.3 (152 feet [46.3 m]). Much Ridge Fire—872 acres (353 ha). ranger to recognize degree of fire confusion and disorganization on This fire on March 9, under very danger and plan accordingly. the line. Maximum number of men serious burning conditions, was Corrective action has been taken. engaged, 65. –Ed.] The initial tac­ probably caused by children living The most valuable lesson in this tics used by the fire boss were cor­ in the vicinity. The fire caught the case involves the continuous daily rect for this fire, and had the heavy preparedness organization of the use of fire-weather stations to equipment functioned properly, district napping. Towers were not determine degree of danger follow­ the fire would have been held to manned. Weather stations and dis­ ing rain. Very rapid changes involv­ less than 100 acres (40 ha). trict fire truck not in operation, ing only a few hours from no dan­ and the district ranger not in ger to high danger are to be De Soto, Leaf River District—Leaf touch with his dispatcher. Failure expected during March, April, and No. 35 Fire—416 acres (168 ha). to recognize emergency conditions May. –R.M. Evans, regional forester [Fire was 75 acres (30 ha) when and to build organization on the discovered 2 hours and 17 minutes basis of emergency conditions Southern Region (R-8) after guessed time of origin. It results in fires of this class. De Soto, Leaf River District— spread 25 acres (10 ha) during the Plantation No. 24 Fire—1,682 53 minutes required for report, Once organized the suppression acres (681 ha). [When discovered get-away, and the travel. –Ed.] action was satisfactory. However, 5 minutes after known time of ori­ because of lack of preparedness it gin, this incendiary grass fire was Low visibility accounts for the size required 3-1/2 hours to get the 3 acres (1.2 ha). Report and get­ of fire when discovered. There suppression organization to the away took 5 minutes. Three miles were five sets and the purpose of fire. The result was a large fire, and (4.8 km) of auto travel, 10 minutes. the fire was to provide sheep the lessons learned by all con­ When reached by the first crew of range. The exposed ridge site and cerned are that regardless of previ­ 17 men, 20 minutes after origin, high winds made suppression diffi­ ous weather, conditions may the fire was 20 acres (9 ha). But it cult. A fire usually occurs in this burned 1,682 acres (681 ha), location about the same time each When this article was originally published, including 870 acres (352 ha) of year. Next year it is planned to Roy Headley headed the Division of Fire plantation, before it was corralled patrol these areas closely in an Control, USDA Forest Service, Washington, effort to prevent the fire or catch DC. 10 hours and 25 minutes after origin. In one place, wind velocity the incendiary. –V.B. * The article is an abridged version of an article pub­ McNaughton, fire assistant lished in Fire Control Notes 4(1) [Winter 1940]: 15–17.

Volume 63 • No. 3 • Summer 2003 23 The lesson from the Ludlow fire It does not hold, however, that this is to promptly investigate if there is procedure can be followed on all smokes, the number and places of any doubt in the mind of a dispatcher or occurrence making this prohibi­ towerman as to the location of a smoke. tive.

It is planned to adopt a policy of De Soto, Biloxi Ranger District— imperative that we get a crew on closer investigation of doubtful University No. 28 Fire—1,238 the head of the fire as soon as pos­ smokes on days of poor visibility acres (501 ha). [Another grass, sible. when no crossing-out is possible, stockman, incendiary fire but with even at the risk of increasing the only one set. This time the lookout For this reason we dispatch a small false-alarm cost. Better a false man snapped in with discovery crew on a “hot shot” pick-up and alarm than a 600-acre (240-ha) within 4 minutes after guessed follow this unit with a 20-man fire! time of origin, when the probable crew. The small crew, upon reach­ area was one-tenth acre (0.04 ha). ing the fire should go to the head Every fire training meeting or con­ But report took 42 minutes. In the and take whatever action is neces­ ference stresses the need of exer­ 65 minutes required for report, sary to check it. As I see it, there cising good judgment in all phases get-away, and 9 miles (14 km) auto are only two possible conditions of fire control. It is a little unfortu­ travel, the fire spread to 350 acres present; either (1) they can hold nate that this qualification must (140 ha). Wind velocity reported as the head by direct attack, or, (2) attend an individual’s every action 13 to 18 miles per hour (21–29 they can’t. In the latter case, their in fire control. Very often, certain km/h) when fire was first reached only alternative is to backfire action which has merited commen­ and during biggest run. –Ed.] against the head and depend on the dation because of its practical follow-up crew to handle the application in hundreds of previous De Soto, Biloxi Ranger District— flanks. cases, falls down in a particular Camp Branch No. 16 Fire—419 case, and the cry “poor judgment” acres (170 ha). On this fire, if the Bienville—Ludlow Fire No. 20— is raised and not altogether unjus­ one-half ton pick-up which was 1,056 acres (427 ha). [The report tifiably, as with this fire. equipped with a Panama pump and shows that the fire started one-half 55 gallons (208 L) of water had not mile (0.8 km) outside the bound­ The lesson to be learned from the bogged down, a successful attack ary. Discovered at 10 acres (4 ha), 1 Ludlow fire is definitely this, in my could undoubtedly have been made hour and 14 minutes after guessed opinion: “Where the slightest on the head. –V.B. McNaughton, time of origin, and reported to dis­ doubt exists in the mind of a dis­ fire assistant. [Illustrating why patcher. But report time is 45 patcher or towerman as to the Region 8 is interested in four-wheel hours and 6 minutes. Hence, with location of a smoke, because of drive trucks. –Ed.] 5 minutes get-away time and 1 poor visibility or any other reason, hour and 5 minutes for 4 miles prompt investigative action should The fire was not scouted and no (6.4 km) travel on foot, a total of take place. Although it was known one reached the head until it had 46 hours and 16 minutes elapsed that debris burning was being done burned into the forks of the creek from discovery to arrival. The fire in the affected area, too great a about 1 mile (1.6 km) south of was 500 acres (200 ha) when chance was taken on these particu­ where the initial attack was made. reached. It did not enter the forest lar hazardous days to let a smoke This fire illustrates what happens until the second day after it start­ of this nature, never definitely when the initial action on fires in ed. –Ed.] located until the third day, go this country is wrong. The climatic uninvestigated. Out of this lesson conditions and fuel types favor an It is known at this time that a crew should come an unqualified policy extremely rapid rate of spread on should have been dispatched to the of action on smokes which literally ■ practically all of our fires, and it is smoke shortly following discovery. cry out: “I need investigating!”

Fire Management Today 24 LESSONS OF THE MCVEY FIRE, BLACK HILLS NATIONAL FOREST*

A.A. Brown Editor’s note: Early issues of Fire Control Notes often prefaced arti­ Making a pretty complete kill cles with substantive remarks by on 22,000 acres of highly productive country, the journal’s editors. A.A. Brown’s the McVey fire was easily article on the McVey Fire is a good example, with a preface long the worst fire of the 1939 year. enough to stand on its own (see the sidebar). km) of held line. Because of the The following comments represent blow-up on July 11, only 10 miles The heading under which this an attempt to sort out experiences (16.1 km) of line were still held by must be written, and its implied of the McVey fire from which les­ 8 p.m. of that day. The additional purpose, impose an automatic cen­ sons may be drawn that are of 36 miles (58 km) were built in one sorship on its contents. A lesson in interest to officers concerned with work period, from 8 p.m. Tuesday fire control to be of interregional fire-control strategy on big fires. night to 9:15 a.m. Wednesday. The interest should introduce some­ For the most part they do not rep­ rate at which the fire burned out thing new, or should at least give resent anything new in principles was remarkable for a timber fire. new emphasis to principles or to of fire control. Once the entire perimeter was con­ aspects of their application that trolled, the whole area went cold have not yet been fully learned. The McVey fire itself was a confla­ almost overnight. This seems to Whether or not any experience gration in flashy fuels, in rolling have been due, in a large part, to qualifies on either of these two topography, with variable winds. the absence of heavy dry material, counts depends on the class of This combination is designed to to the burning out done on all fire individual. test the resources of any fire boss. lines, and to the extreme heat pro­ Rates of forward spread on wide duced by the slash on the ground. McVey Fire fronts up to 120 chains (7,920 feet A list of the lessons learned by the [2,414 m]) per hour occurred. The lessons to be learned from this Civilian Conservation Corps (CCC) Direction of spread varied, and rate fire for the local organization have boys on the McVey fire would have of increase in terms of acreage the usual and familiar character of a very different content from the burned went as high as 2,900 acres lessons in fire strategy, in organi­ list that would be most appropriate (1,200 ha) per hour during the run zation, in speed of attack, in meth­ for their foremen. Similarly, some of Tuesday, July 11. The cover type ods of attack, and in fire preven­ of the lessons learned by members of the whole area was ponderosa tion and preparedness. All of these of the supervisors’ staffs of the pine, interspersed with small have been discussed at some Barney and Black Hills Forests meadows. Over two-thirds of the length, and most of them on which would sound trite to a southern area burned consisted of thinned it was agreed that something defi­ California fire fighter, although stands of ponderosa pine, in which nite and constructive could be other features of the job might the thinning slash formed a con­ done locally have been duly listed have considerable thought-provok­ tinuous layer of fuel of varying in the conclusions of the Board of ing challenge to outsiders. density and stages of decay, dating Fire Review. For the most part, from 1933. they will not be repeated here, but an attempt will be made to go back When this article was originally published, Although this fire reached a total a little further in a consideration of A.A. Brown was head of Fire Control for the USDA Forest Service, Rocky Mountain area of 22,000 acres (8,900 ha), it the significance of facts brought Region. was controlled as a second-period out by fires such as this one. fire with a total of 46 miles (74 * The article is reprinted from Fire Control Notes 4(2) [Spring 1940]: 63–67.

Volume 63 • No. 3 • Summer 2003 25 Inadequate Theory kinds of fires. A 1-acre (0.4-ha) fire Most of our theory of fire-control One of the first things that seems may be only a smoldering spot or it tactics is based on a two-dimension important to recognize is the inad­ may already be as dynamic as a idea, and the control of a fire is equacy of existing fire-fighting small tornado. usually pictured as the solution of methods to meet and overcome all a problem in plane geometry. It

Editors’ Preface to A.A. Brown Article

Making a pretty complete kill on work-period policy. But about 11 over 3.3 miles (5.3 km) of front 22,000 acres (8,900 ha) of highly p.m. the wind freshened a little as of 6 p.m. on the first day. The productive country, where every and the fire marched out author’s figure of 10 miles (16.1 cubic foot of timber that can be through the unworked gap of km) to build as of 6 p.m. evident- grown is needed to sustain the “less than half a mile [0.8 km]” ly considers expected spread after dependent communities, the and started a spread which was that time, but does not include McVey fire was easily the worst not corralled until 34 hours the 6.9 miles (11.1 km) of line or fire of the 1939 year. To find and later. edge as of 6 p.m. which was in follow through the clues leading the same place at 9 a.m. the next to increasing mastery of such The surge of the fire at about 11 morning. fires offers the sharpest possible p.m. the first day was the result challenge to fire-control stu­ of recognizable causes or it was Output of held line up to the dents. Such losses are intolera­ not. Can we discover why this completion of the final control ble, but it will take more than and similar fires behave the way line was 0.1 chain (7 feet [2.1 our usual insight and methods of they did? Whether, like a rattler, m]) per man-hour_which again study to find out why they occur. they actually gave a warning falls within our semistandard of When we find the answers, it will which we might learn to recog­ 0.06 to 0.16 of a chain (4 to 11 take more than our usual forms nize in future before it is too feet [1.2–3.3 m]) per man-hour. of training to get the essential late? No figure is available on lost principles of control so embed- line. ded in the minds of enough men Fire Perimeter that these principles will be Control Spot Fire Problem applied in future crises. The data submitted by Staff Spotting was naturally very bad. Technician Skinner shows that at Can’t we develop better ideas for Breaking the Rules 6 p.m. on the first day, 5-1/2 dealing with spot fires? As a sim- At 11 p.m., 9 hours after the hours after discovery and 4 ple example, how can we get arrival of the first 68 men on the hours and 40 minutes after men to watch and comb the fire, those in charge recognized arrival of the first cooperator places where spot fires may start for the first time that they “were crew, the fire had a perimeter of instead of watching the fire from up against a fire that wouldn’t 10.2 miles (16.4 km) (including a which the sparks come? Sounds act according to rule.” The fire 25-percent addition to a machine easy, doesn’t it? had spread less than 250 acres count of map miles). Of this (100 ha) in the previous 5 hours total, 4.6 miles (7.4 km) were in Before the first crew arrived, two and had reached a total of the same location as when the post cutters working nearby saw approximately 1,600 acres (650 fire was finally corralled_mostly and tried to stop this fire. They ha). Wind from 7 to 10 p.m. had on or close to roads, fields, and tailed and had difficulty in escap­ been “nearly imperceptible.” prairies; and 2.3 miles were held ing. They were suspected of Completion of the control line in until 9 a.m. the next morning. starting the fire, but it was later the early morning hours The spread from 6 p.m. of the proved to be a hang-over light­ “appeared certain” long before first day to 9 a.m. of the second ning fire. the time required by the first day came, therefore, from not

Fire Management Today 26 seems to me that this type of Nine hours after arrival on the fire, thinking has carried over too far those in charge recognized for the first time into ideas of application on the job so that the experienced fire fore­ that they “were up against a fire man himself comes nearer to a full that wouldn’t act according to rule.” realization of the potentialities of a given fire in the third dimension down at night. Yet fires do so so its run, prevailing winds and past than does the so-called fire-control commonly that both the forester’s experience with such fires in this expert. and the regional policy are based locality have revealed that the odds on overnight action which is are four to one that the greatest This of course, sums up to our judged to be adequate on a big fire conflagration threat will be to inability to cope with crown fires. if sufficient manpower and facili­ country to the east and northeast It is not a new subject but it has ties are mobilized to control a of the fire. This was well known, been rather studiously ignored in more or less static fire perimeter yet the timing of control effort was discussions of fire fighting. before 10 a.m. of the next day. On such that the most northerly and Certainly, this one factor looms the McVey fire, forest officers westerly extensions of the fire were large in the history of the McVey allowed themselves a large safety the last to be controlled. As a result fire. margin, but as it turned out, not the blow-up here flanked all the enough, since they did not foresee hard-won control line to the east. Missed Opportunity the critical necessity of a safe Assuming that nothing could be perimeter that could be held This was a question of judgment at done during the height of the against all odds by 11 p.m. of the the time and not as obvious a fail­ spread of this fire through the first night. ure as it may sound in review. The crowns, there were lulls during worst fuels and the most aggressive which it might have been con­ Regardless of how feasible such an head of the fire were given first trolled if certain things could have accomplishment might have been attention, and the hardest part of been done quickly enough. These on this particular fire, it, of course, the fight seemed already won when “ifs” comprise debatable points presents a very real question in the the fire made its new drive north­ from which some lessons may be case of future fires. Will it ever be ward on its west flank. This exten­ drawn. During the first afternoon possible to predict the behavior of sion set the stage for the final run and evening, from 6 p.m. to 11 an exceptional fire far enough in eastward on a front of conflagra­ p.m., the fire dropped out of the advance to insure the exceptional tion proportions. The west flank of crowns and stayed relatively quiet. intensity of attack required? Ways the fire had already burned out to a There were 10 miles (16.1 km) of in which Region 2 officers believe wide expanse of noninflammable fire line to build, which it was such foresight might be improved grassland which gave a strategic expected would be complete by 2 are: Through improved fire-danger advantage. Had the progression of a.m. As it turned out, this would ratings, improved fire-weather safe-control line been entirely from have been completed, probably by forecasting, and through better the grassland eastward along the midnight, but the blow-up at 11 recognition of potentialities of north flank and from the complet­ p.m. with less than one-half mile local fuel hazards. ed south line northwestward along (0.8 km) of line still to build the east flank, the investment in resulted in losing nearly all that Unsound Strategy fire-control line would have been had been accomplished. A plan of The next lesson of critical import better protected against loss and attack designed to complete the job and perhaps the one of most con­ the same blow-up with the same before 11 p.m. might have saved structive value of all is that of the amount of uncontrolled perimeter the day. strategy of control in an aggressive farther east at 11 p.m. would have fire such as this. Although the been far less disastrous, in terms of It has long been realized that it is main head or heads of the fire took both a narrower new head and less unsafe to stake everything on the one direction, then another, during lost line. assumption that a fire will slow

Volume 63 • No. 3 • Summer 2003 27 Forces were reorganized and a fresh ing the main fire. But when the attack was made—a sound strategy, two met they seemed to “bounce” right back through the unburned but its execution failed from circumstances crowns toward the fire line without which may also carry some lessons. any noticeable drop in intensity. This did not always occur but hap­ This lesson in importance of sound tion of a reasonably safe control pened so often as to make backfir­ strategy would not have become line alone. Both on this occasion ing ineffective at close range and apparent at all had the fire and at other times during the fire very risky at longer distances. remained quiet an hour longer or the rate of production of a reason­ Apparently the nature and force of had the fresh crews scheduled for ably acceptable control line, the convection drafts is the decid­ this sector been the first to arrive including burning out, did not rep­ ing factor. It is a challenge to back­ and start work instead of the last. resent the rate at which the fire firing theory and strategy. Similarly, more manpower or was being controlled. increased efficiency in line con­ Challenges to Theory struction, sufficient to complete all This raises a question of tactics and Practice control line before 11 p.m., might where fire line is being built rapid­ Out of these experiences there may have resulted in success with the ly. The job of holding the line built be resolved several points of chal­ timing used. calls for far more than is implied lenge to existing fire-fighting theo­ by either the term patrol or mop- ry and practices. These points, Failed Execution up. It represents on a fast-running rather than the lessons discussed, After the fire had made its run fire a desperate defense action, will be listed in the form of ques­ northward another lull occurred since the advance crew cannot wait tions. between the hours of 2 and 7 a.m. to see the fire finally die down of the morning following. Forces behind it, but must pass on to the 1. How can forest fire-fighting the­ were reorganized and a fresh attack follow-up crews the responsibility ory and practice take better was made during this period which of making good the advantage account of fire as a force to be carried considerable promise of gained. When 2 to 3 miles (3–5 dealt with which varies as a success in spite of a greatly km) of such line is handed to a sec­ problem not only with the area increased perimeter. In this case tor boss he must drastically change involved but also with the vary­ little fault can be found with the the approved method of organizing ing rate at which heat units are strategy planned, but its execution the job, which was based on mop­ being released and with the con­ failed from circumstances which ping up a dying fire edge. vection forces being generated? may also carry some lessons. Apparently this was not well 2. How can the requirements of enough done. The conclusion then successful control action for The critical sector of the new line is that organization of crews must each individual fire be predicted was being pushed rapidly from be handled on a very different basis more successfully? both ends by experienced fire fore­ than the conventional mop-up or 3. What principles of fire-control men using the one-lick method patrol action when fire line is put strategy can be defined and set and burning their line clean as in rapidly by an advance crew or by up for general application? they went. Follow-up crews were mechanical means. 4. What changes in crew organiza­ organized to guard the hot line and tion should be adopted to insure to control spot fires. The latter A further experience on the McVey holding long stretches of fire- failed to accomplish this. No one fire that was baffling to sector control line built rapidly along knows exactly why, except that this bosses was the tendency of back­ the perimeter of an active fire? job turned out to be far more fires to spread inward toward the 5. How can the use of backfiring be exacting and required far more oncoming crown fire in the surface reduced to a dependable practice supervision and more ingenuity litter just as intended, and to go in combating crown fires? ■ and action than did the construc­ into the crowns just before reach-

Fire Management Today 28 AN ANALYSIS OF THE HONEY FIRE*

C.F. Olsen

Original editor’s note: Although the Honey Fire occurred almost 2 Fires of the extreme intensity and years ago the following analysis is rate of spread of the Honey Fire in the published because of its value to longleaf/slash pine type are the exception other fire control personnel. The data were gathered by research rather than the rule. personnel who were present throughout the period of the fire spread and resistance to control, (24 km) north of Alexandria, head­ and who were free from any sup­ was obtained.** quarters of the forest supervisor. pression duties. A “Board of Fire The central tower of the ranger Review” held by the regional This fire record will be discussed district, the Catahoula Tower, is forester in 1938 brought out addi­ and analyzed in detail with three located 1-1/2 miles (2.4 km) east of tional information and criticisms objectives: Bentley. The highways, railroad, which were furnished to the firebreaks, and other physical Southern Forest Experiment 1. To indicate the rate of spread improvements on the fire area are Station for its study. and behavior of fires burning shown on the accompanying maps. under extremely critical condi­ Employees on a south-bound tions. The area is typical of open cut-over freight train in north-central 2. To describe the action that was longleaf pine land in the Upper Louisiana carelessly disposed of a taken to suppress the fire. Coastal Plain. The topography piece of burning waste from a hot 3. To use the experience gained ranges from flat to gently rolling, box on a crisp January morning in from this fire as a guide in plan­ with occasional depressions of wet 1938. The bit of flame landed in ning the action to be taken on and boggy land. Several small dead grass at the edge of the tracks other fires burning under simi­ creek bottoms occur on the area, inside the boundaries of the lar circumstances. but because they are in general Kisatchie National Forest; 30 min­ very narrow, they contribute little utes later this small flame had Location or nothing as natural barriers to grown into a forest fire with a fast-spreading fires. The principal perimeter of almost 4 miles (6.4 The Honey Fire, briefly described by Headley (1939), was known fuel in the area is grass, km) and had spread almost 2 miles broomsedge (Andropogon sp.) (3.2 km) with the wind from the locally also as the Dyson Creek Fire. It occurred in the center of being predominant. Other compo­ point of origin. A crew of four men, nents of the fuel are various herba­ assigned by the Southern Forest the Catahoula District of the Kisatchie National Forest. The fire ceous plants, pine needles and Experiment Station to studies in cones, and hardwood leaves. fire behavior, was on the scene started along the east side of a rail­ road right-of-way 1-1/2 miles (2.4 within 3 minutes after its start and Approximately 150 acres (60 ha) an unusually complete record of km) north of Bentley, a small set­ tlement approximately 15 miles covered by this fire were burned this fire, including its rate of over on February 21, 1935. The remainder of the area, or about ** In order to obtain better data, the forest supervisor had previously agreed that the members of this crew 950 acres (384 ha), had been were relieved of any obligation to assist in fire suppres­ unburned for at least 6 years. When this article was originally published, sion. Usually they were in effect dispatched with the C.F. Olsen was a forester for the USDA regular suppression crew; in this instance they hap­ Inasmuch as all of the area had Forest Service, Southern Forest pened on the fire at about the time it was reported. been unburned for 3 years or Experiment Station, Asheville, NC. They were criticized for not trying to control it; but with two fences and a railroad between them and the longer and the site is better than fire, there is no doubt that their truck was unusable on * The article is reprinted from Fire Control Notes 5(4) this fire. It was very definitely too big for them to hold average, a uniform and extremely [Fall 1941]: 161–178. with hand tools alone. heavy stand of grass covered the

Volume 63 • No. 3 • Summer 2003 29 entire area and contributed Weather Conditions This instrument consisted of a markedly to the intensity of the The weather conditions that pre­ Byram-type fan anemometer rest­ fire. Also scattered residual seed vailed in the region immediately ing on a universal joint and trees and some reproduction were before the fire started and on the 2 mounted on a tripod. The instru­ present in the western half of the previous days are shown in table 1. ment was placed a sufficient dis­ area and the eastern half supported tance from the fire to be unaffected moderately well stocked stands of Since this fire was mapped for rate by the drafts and currents created longleaf pine saplings, light to of spread in connection with a during rapid combustion. The heavy stands of blackjack oak, and study of fire behavior and the anemometer was set with the spin­ open grass. Following the 1935 recording of current weather data dle 3-1/2 feet (1.1 m) above the fire, 20-foot-wide (6-m-wide) dri­ was an integral part of that study, ground, the standard height for vable firebreaks were constructed weather records were made, begin­ measuring wind velocity in all over portions of the area. An 858­ ning shortly after the start of the studies of rate of spread. acre (347-ha) plantation of slash fire. The factors measured were pine was established by the Forest relative humidity (by a hand-oper­ A record of the average and maxi­ Service in the center of the burned ated psychrometer), air tempera­ mum wind velocity, relative area during the 1936–37 planting ture, wind movement, wind direc­ humidity, air temperature, and fuel season. tion, and sky condition. General and soil moisture content (based notes also were made. Wind move­ on dry weight) is presented in All of the burned area lies within ment was taken at 1-minute inter­ table 2. the national-forest protection vals with a portable anemometer boundary, but only the east half is developed by the California Forest national-forest land, the remainder and Range Experiment Station. being privately owned.

Table 1—Weather conditions preceding the Honey Fire (observations made at Catahoula Tower).

Temperature Rainfall Wind Date Max. Min. Amount Time Sky condition Direction Rate Visibility Jan. 23 73 °F 62 °F 0.93 Noon– Cloudy South Light Poor inches 6 p.m. Jan. 24 66 °F 48 °F 0 — Cloudy and SW–NW Moderate Poor threatening (3 miles) a Jan. 25 42 °F — 0 — Clear NW Moderate Good

a. 10 a.m.

Table 2—Record of weather, fuel moisture, and soil moisture conditions during Honey Fire.

Moisture Wind velocity content (mph) Relative Air (% dry weight) humidity temp. Period Avg. Max. Time (%) (°F) Time Fuel Soil 10:07–10:37 a.m. 6.7 9.9 10:06 a.m. 33 42 10 a.m. 15.8 34.1 11:28–11:59 a.m. 11.3 16.6 11:20 a.m. 27 50 11:55 a.m. 12.1 34.1 12:20–12:33 p.m. 12.7 14.3 12:33 p.m. 26 46 1:45 p.m. 11.6 — — — — 2:17 p.m. 26 46 — — —

Fire Management Today 30 The Fire The combination of high, shifting winds Time of Start. The Honey Fire and low fuel-moisture content prevailing at the started at 9:50 a.m. on January 25, time of the fire created critical burning conditions. 1938, in the manner described. The fire-behavior crew, having been of the fire advanced at a rate rang­ during the course of this fire; all of traveling south along U.S. 167 ing from 25 to 35 chains (1,650 to the heads labeled from B through about 1 mile (1.6 km) behind the 2,310 feet [503–705 m]) for each 5­ H developed either by wind shifts train, arrived at the fire at 9:53 a.m. At the moment, the fire had minute period. This head, labeled A along the north flank of the main advanced more than 100 feet (30 on the accompanying map (fig. 1) head A and its subsequent heads or m). showing the progress of those parts from spotting across roads or of the fire that could be reached for burned firebreaks. Rate of Spread. Three members of mapping with the limited fire- There were great differences in the the crew started mapping the fire behavior crew available, stopped of rates of spread of the flames at dif­ at 9:55 a.m., while the fourth its own accord when it reached an ferent points on the fire. In the member collected fuel and soil abandoned Civilian Conservation main, however, the fire spread for­ samples and set up instruments to Corps (CCC) camp site on which a ward at the rate of 5 to 6 chains obtain weather data. The main heavy cover of carpet grass was (330 to 396 feet [101–121 m]) per head of the fire and the north flank present. minute. The greatest rate of spread were mapped at that time and at Because of distinct shifts in wind measured in a forward direction each 5-minute interval thereafter. direction from northwest to south­ was 8 chains (528 feet [101 m]) in After the first 5 minutes, during west, however, a new head B devel­ 1 minute. In the easternmost part which it had moved almost 6 oped along the north flank (fig. 1). of the area, where the fire was chains (396 feet [121 m]) forward A total of eight heads were mapped finally brought under control and from the point of origin, the head

Figure 1—Map of the Honey Fire, showing rate of spread and fire behavior.

Volume 63 • No. 3 • Summer 2003 31 where there were dense stands of 20 to 25 feet (6–8 m), with the acres (160 ha) were in the slash- blackjack oak in which the carry­ width of the burning line 15 feet pine plantations mentioned earlier. ing fuel was considerably less than (5 m) or more. in the open longleaf pine areas, the Available Suppression Crews and forward rate of spread dropped to There were numerous cases of Equipment. On the morning of as low as 1 chain (66 feet [20 m]) spotting for a considerable distance January 25, nine crews of fire per minute. ahead of the fire; in one instance, fighters were available to the fire when the wind velocity was 13 dispatcher for fire duty. The crews The perimeter and area increases miles per hour (21 km/h), fire were made up of CCC and Work that accompany the forward rates spotted over 200 feet (60 m) in Projects Administration men who of spread on fast-moving fires are advance of the head. An unusual worked either at the Stuart also very high; the figures for the case of spotting occurred when a Nursery or on planting and road main head A are given in table 3. dead snag, located 95 feet (30 m) maintenance jobs within easy driv­ from the nearest edge of fire, ignit­ ing distance of the central tower. Fire Behavior. Besides collecting ed at a height of 12 feet (4 m) The crew organization is shown in data on the rate of spread of the above the ground. Hardwood table 4. fire, the mapping crew recorded leaves, especially those from black­ observations of various items of jack oaks, were responsible for all The standard fire tools, with which fire behavior that influence fire- spot fires noted. all except crew No. 1 were suppression action. Among these equipped, consisted of flaps (swat­ were flame height, width of the The spread of the fire was stopped ters), hand-operated back-pack burning line, incidence and dis­ and the fire corralled at 2:43 p.m.; pumps, fire rakes, water buckets, tance of spotting, and difficulties the fire was controlled and mop-up railroad “fusee” torches, axes, etc., experienced by the fire fighters. completed at 6:45 p.m. all of which were kept in a wooden box on the trucks transporting the The flames at the head frequently The final total area burned in this men. The pumper truck was reached out in long tongues fire was 1,092 acres (442 ha), of equipped with dual wheels, a 350­ extending 100 feet (30 m) or more which 493 acres (200 ha) were on gallon (1,325-L), water tank, and a in advance of the actual burning of national-forest land and the pressure pump unit driven from the fuel; on the flanks, a slight remainder, or 599 acres (242 ha), the fan-belt of the engine. shift in wind direction would on privately owned land within the increase the flame height from an national-forest boundary. Of the Discovery. The fire was discovered average of 3 to 4 feet (1–1.2 m) to national-forest land burned, 396 by the lookout on the Catahoula

Table 3—Rate of spread on main head (A) of Honey Fire.

Forward progress Perimeter Area Time (chains [ft]) (chains [ft]) (acres) elapsed after start Total Increase Total Increase Total Increase 5 min 5.7 (376) 5.7 (376) 13.6 (898) 13.6 (898) 0.9 0.9 10 min 32.8 (2,165) 27.1 (1,789) 70.4 (4,646) 56.8 (3,749) 10.2 9.3 15 min 57.4 (3,788) 24.6 (1,624) 121.3 (8,006) 50.9 (3,359) 27.1 16.9 20 min 93.4 (6,164) 36.0 (2,376) 195.9 (12,929) 74.6 (4,924) 60.5 33.4 25 min 126.0 (8,316) 32.6 (2,152) 263.4 (17,384) 67.5 (4,455) 109.2 48.7 30 min 148.2 (9,781) 22.2 (1,465) 311.9 (20,585) 48.5 (3,201) 167.4 58.2 35 min 175.9 (11,609) 27.7 (1,828) 368.6 (24,328) 56.7 (3,742) 250.8 83.4

Fire Management Today 32 Table 4—Crews and equipment available for suppression of Honey Fire.

Supervisory Crew No. Firefighters personnel Total Equipment 1 1 1 2 Pumper truck, 350-gallon capacity. 21 21 13 Standard fire tools. 3 5 1 6 Do. 42 41 25 Do. 5 20 2 22 Do. 62 11 22 Do. 71 72 19 Do. 81 81 19 Do. 91 11 12 Do. Misc. superv. pers. — 8 8 — Total 129 19 148 —

Tower, located 2 miles (3.2 km) to fire. It left Catahoula Tower, where the train at the road crossing. It is the east, at 9:52 a.m. (2 minutes it was standing by for emergency estimated that the fire had a after the start) and reported imme­ use, at 9:53 a.m. and went directly perimeter of 40 chains (2,640 feet diately to the fire dispatcher. A to the origin of the fire near U.S. [805 m]) upon their arrival. They cross-shot was obtained from the 167. While enroute, the driver started to extinguish fire along the Colfax Tower lookout, 4 miles (6.4 determined definitely that the fire north flank near the head for the km) to the west at 9:53 a.m. The was on the east side of the railroad purpose of checking the fire by fire dispatcher, therefore, had a track. The train momentarily cutting it into a cultivated field reasonably definite location of the delayed him by blocking the road located about three-quarters mile fire within 3 minutes of its start. At crossing leading into the fire. He (1.2 km) southeast of the origin. that moment, however, it was then made an attempt to reach the These tactics failed when the head impossible to determine with fire, but the pitcher-plant passed to the north of the field; the absolute certainty on which side of (Sarracenia spp.) land which it was forward progress of the south the railroad track the fire was then burning was so wet and boggy flank, however, was checked when burning. There was still a possibili­ that this was impossible. He it reached the field. The crew ty that the fire was burning in the returned, therefore, to Catahoula leader then split his crew; he left 7 150 feet (50 m) of grassland Tower and received instructions to men to suppress the fire, starting between U.S. 167 and the west side wet down the fuel and be prepared from the tail and working toward of the railroad track or even west to extinguish spots along the east the head, and led the remaining 5 of the highway. side of Tower Road, starting near men (crew 2B) on foot across the tower, while crews 3 and 4, country to the west firebreak. Crew Dispatch and Initial Attack. who had meanwhile reported at The map showing the initial and Catahoula Tower for fire duty, The main head of the fire, mean­ subsequent crew locations (fig. 2) burned a backfire along the west while, had reached Tower Road, indicates the points at which the side of Tower Road. where it was stopped by the aban­ fire was attacked before control doned CCC camp and the fireline was attained. Meanwhile, crew 2, the leader of burned by crews 3 and 4 and the which was fire boss, left Catahoula pumper truck; crews 3 and 4 then Crew 1, consisting of the pumper Tower for the fire at 9:55 a.m., went to the west firebreak to burn truck with a driver and hoseman, going to the point of origin. Thus more backfire similar to that which was the first crew dispatched to the crew was momentarily delayed by they had just completed on Tower

Volume 63 • No. 3 • Summer 2003 33 Figure 2—Map showing points at which the Honey Fire was attacked.

Road. There the fire boss assem­ and was held at that point when Tower Road before adequate back­ bled crews 3 and 4 and his own 5 the head hit the backfire. fires were completed. men (a total of 37 men) on the west firebreak and attacked the Later Attack and Tactics. In extinguishing the fire along the north flank near the head and Following the initial attack north flanks, the crews, with one worked toward the tail. Progress by described, during which two differ­ exception, attacked the fire from a this large crew in extinguishing ent heads on fast-moving fires were point near the head and worked the fire along the north flank was stopped by indirect attack or back­ toward the tail. The only fire fight­ encouraging; a boundary length of firing but on which the fire on the ing on the north flank from the tail 33 chains (2,178 feet [634 m]) was north flanks was not controllable, toward the head was done by the put out and being held successful­ the crews resorted to further back­ original crew (2A) of 7 men, who ly. firing along the west firebreak, were left near the tail when crew 2 Tower Road, State Highway 19, and was split. As a consequence, a con­ Up to this time, the wind had the east firebreak, and to patrolling siderable distance along the north blown steadily out of the north­ the east side of these roads to pre­ flank (for the most part, 1 mile [1.6 west, with little or no evidence of vent the formation of new heads by km] or more) between the tail and shifting markedly in direction. At spotting when an oncoming head the west firebreak was left to burn 10:44 a.m., the wind distinctly reached a backfire. freely, and with each shift in wind veered from northwest to south­ direction to the southwest new west, resulting in a big sweep in Efforts to extinguish the north heads would develop. Examples of the flank and the formation of a flanks were made only when it was mapped heads that resulted from new head. The suppressed crew reasonably certain that the head this situation are C, D, and F; sev­ was forced to yield ground. The could be held at the backfire. In all eral unmapped heads developed new head (B), which wiped out all the heads mapped, the backfire previously in the area west of the the line that had been held up to effectively stopped the progress of west firebreak, causing heads C, D, that time, reached the west fire­ the head of the fire. Heads G and H and F by spotting across this fire­ break in 9 minutes (at 10:53 a.m.) resulted from spotting across break.

Fire Management Today 34 Fire along the south flank was of was solid enough so that it would table 6 are only for stringing fire in minor concern during the run of not bog down. Throughout the a straight line without regard to the fire, small roving patrols were final attack, effective and rapid the width of the backfire burned. stationed at strategic points along suppression on the north flank was State Highway 19 to backfire wide accomplished by working from the Ratio of Line Actually strips and to watch for spotting tail toward the head and mopping Extinguished to Total Needed for across the road when heads up the edges of the fire simultane­ Control. As pointed out above in reached the road. The large culti­ ously. the discussions of the initial attack vated field located three-quarters of and the output of held line, consid­ a mile (1.2 km) southeast of the Output of Held Line. During the erable line was lost because of the origin did much to lessen the fire course of the fire, measurements behavior of the fire at certain danger along the south flank. The were taken of the line extinguished points. No accurate record of the shift in wind direction from north­ by various crews. The amount of total line extinguished but later west to southwest also reduced the line held per man-hour varied lost is available. It has been conser­ fire danger along the south flank, greatly at different points on the vatively estimated by Kisatchie since it resulted in a relatively fire, depending upon the behavior National Forest personnel, howev­ slow-moving flank burning into of the fire and the conditions er, that of the 864 chains (57,024 the wind with only occasional under which it was being fought. feet [17,831 m]) of line actually minor sweeps when the wind In table 5 a record is given of the built to corral the fire, 240 chains changed back to the northwest. output of held line per man-hour (15,840 feet [4,828 m], or 27.9 per­ for different sized crews with cent) were lost during the suppres­ Final Attack. In the final attack on remarks concerning the conditions sion action and did not contribute the fire, during which it was under which they worked. toward the control of the fire. The brought under control, wider fire­ difference, or 624 chains (41,184 breaks were burned along Tower Record of Stringing Backfire. feet [12,553 m]), therefore, would Road and the east firebreak. The Data on the rate at which backfires have been sufficient to attain con­ suppression crews also attacked the or burned firebreaks were strung trol of the fire. north flank of the fires from the by different methods were obtained tail toward the head by reinforcing at several points on the fire. These The ratio of line actually extin­ the small crew left originally at the data do not include the manpower guished, but later lost, to the total tail to work east. The pumper required to keep the backfire under needed for control is 240 chains truck did very effective work along control by patrolling and mopping (15,840 feet [4,828 m]) to 624 the north flank where the ground up spot fires. The data given in chains (41,184 feet [16,667 m]);

Table 5—Rate of held line per man-hour at various points on Honey Fire.

Number Held line per Location of suppression men in man-hour action crew (chains [ft]) Remarks North flank—head A 19 4.7 (309) No shift in wind. North flank—head A 6 –11.6 (–765) Line lost—shifting wind. North flank—head D 14 5.8 (385) Some crew members idle or resting. North flank—head G 14 9.1 (601) In heavy cover of oak leaves. North flank—head H 18 9.15 (604) In heavy oak brush—medium fuel. North flank—heads F, G Pumper truck 126.6 (8,356) Average difficulties with trees and soft ground. North flank—head A 10 5.03 (332) Average conditions—men placed too much reliance on water.

Volume 63 • No. 3 • Summer 2003 35 thus, 38.5 percent more line was their hands. The fire fighters expe­ Critique built than actually needed. The rienced considerable difficulty in Recognition of Danger. In evalu­ final perimeter of the fire when walking against and in using their ating the suppression action taken controlled was 934 chains (61,644 equipment in the strong wind cur­ on this fire, it must be realized that feet [18,789 m]). The discrepancy rents created near the head of the no satisfactory methods and tech­ between the final perimeter and fire. Thus, the efficiency of the fire nique were then available to the the length of line actually needed fighters stationed at these points to fire dispatcher to rate the fire dan­ for control, or 310 chains (20,460 extinguish break-overs and spot ger existing at the time of the fire. feet [6,236 m]), is accounted for by fires was greatly reduced. Further, His experience in judging fire dan­ the fact that practically none of the the roar of the fire and wind at ger during several preceding fire south flank required suppression, these points made it impossible for seasons, however, made it clear to because the fire went out of its the crews to hear verbal orders of him that the weather conditions own accord when it reached the their foremen. then prevailing would cause a fast cultivated field, the abandoned rate of spread and that speedy and CCC camp, Highway 19, and the The heat on the north flanks when adequate dispatch of suppression backfire along Tower Road. the wind shifted was oppressive, crews was essential. Consequently, and the danger of a crew getting he had prepared and organized all Difficulties Encountered by Fire trapped by the high, oncoming crews for speedy dispatch. Even Fighters. The suppression crews flames was great. The hose man on after dispatching all available fire­ were under tremendous handicaps the pumper truck was particularly fighters, he was quick to recognize and personal discomfort, caused by handicapped by the heat because the extreme conditions and to the heavy, choking smoke and the he had to get very close to the fire inform the supervisor’s office that dense cloud of ashes, soot, and to place the water effectively. the fire was out of control by sparks, when patrolling an onrush­ reporting, “I cannot hold it.” ing head as it hit a backfire. At Because of the relatively flat ter­ such times, it was impossible for rain and the dense smoke, the fire A fire-danger meter, recently devel­ them to face toward the head; not boss was unable to get the clear oped by the Southern Forest only was the visibility extremely and complete picture of the Experiment Station (Bickford and bad, but also the dense ashes and progress of the fire that he needed Bruce 1939), should prove exceed­ sparks, carried swiftly by the draft for the most effective use of his ingly useful to a fire dispatcher in of the fire, compelled the fire fight­ men in controlling the fire. recognizing fire danger, particular­ ers to turn their backs to the fire ly under conditions similar to and cover their smarting eyes with

Table 6—Rate of stringing fire for backfires on Honey Fire.

Length burned per man-hour Backfire location (chains [ft]) Medium used Remarks West firebreak—near 52.7 (3,480) Gasoline torch Inexperienced men. heads C, D, F Tower Road—head D 26.9 (1,776) Bunches of grass Backfire only 20 ft wide, not enough to keep head C from crossing Tower Road. Tower Road—head F 51.8 (3,420) Rakes — Highway 19—head E 26.2 (1,726) Bunches of grass In oak brush and leaves. East firebreak—head G 20.9 (1,378) Bunches of grass In heavy blackjack oak. Tower Road—head D 45.1 (2,978) Rakes —

Fire Management Today 36 those prevailing at the time of the The fire fighters experienced considerable difficulty Honey Fire. The present fire-dan­ in walking against and in using their equipment ger meter, for these conditions, would have shown the danger as in the strong wind currents created class 5, or extreme. With a prompt near the head of the fire. recognition of the fire danger and with adequate plans for the fire the tail of the fire; but since this given clearly and definitely by the action to be taken in the event of a fire started on the east side of the fire dispatcher. The chain of com­ fire, it is expected that fires occur­ track the preferred action, had the munications to the individual ring during times of great danger real danger been fully recognized, crews previous to their initial dis­ can be checked early and con­ would have been to place the initial patch to the fire was, for the most trolled while still of small size. crews along the west firebreak to part, very satisfactory. Some delay string backfire and send subse­ was experienced in reaching the Fire-Discovery Time. Fire-discov­ quent crews to extinguish the fire crews that were working in the ery time was excellent and the tow­ along the north flank. Stuart Nursery quickly, because ermen are to be commended for the telephone in the nursery office their alertness. With good visibility It can be argued that, under the was unmanned for several minutes; and with a clear view of the origin circumstances, the fire dispatcher when word finally reached the of the fire available from Catahoula should have momentarily delayed nursery, the three crews were Tower, conditions were very favor­ initial dispatch of crews until he promptly dispatched to the fire. able for quick discovery. Further­ had received verification from the more, when the train made an Catahoula towerman on this seem­ Supervision. All the crews were unusual stop, it was viewed with ingly trivial point. However, to supervised by men who had had suspicion by the Catahoula look­ have made such verification at that considerable experience fighting out. time would have been contrary to grass fires in the cut-over longleaf the forest supervisor’s instructions; pine type. Fires of the extreme A cross shot was quickly obtained under other circumstances, even intensity and rate of spread of the from Colfax Tower, located 4 miles the slightest delay in dispatch Honey Fire, however, are the (6.4 km) west of the origin. Be­ would have been costly insofar as exception rather than the rule. cause of the obtuse angle of this size of fire was concerned. Consequently, it was natural to cross shot, which the fire dispatch­ expect that some mistakes in judg­ er received within 3 minutes of the Preparedness. Adequate prepara­ ment and action on the part of the start of the fire, it was impossible tions and crew organization had supervisory personnel would be to get a precise location. Moreover, been made for fighting fires on bad made. The writer points out what he was under specific orders from fire days. A total of 148 men, divid­ he considers as mistakes only to the ranger and forest supervisor to ed into 9 properly supervised, guide the actions of supervisory dispatch the stand-by crew imme­ trained, and equipped crews, were personnel in the future under simi­ diately upon obtaining a cross ready to respond promptly to a fire lar circumstances. shot. call from the dispatcher. These crews were distributed at strategic Every member of the supervisory Under ordinary conditions the points on the ranger district and personnel, including the fire boss, location of the fire, as indicated by had telephone connections with used a flap, a back-pack pump, or the reported conditions, would be the fire dispatcher’s office. All feasi­ some other fire-fighting tool. It is highly satisfactory for a prompt ble measures of preparedness had commendable that they were so attack. In this particular case, how­ been taken. earnest and eager to get the fire ever, a precise location was essen­ extinguished that they helped in tial, since the subsequent fire Dispatch of Crews. The dispatch the physical work, but it is much action depended upon this point. to the fire of all the crews available more important and necessary for Had the fire started on the highway on the ranger district was effected those in charge of fire crews to west of the railroad track, the logi­ promptly and with a minimum of expend their energies and use their cal action would have been for the confusion. Their assignments to superior training in analyzing ever- fire boss to lead the first crews to specific points on the fire were changing situations on a fire, in

Volume 63 • No. 3 • Summer 2003 37 The crew leaders should use their heads and eyes line. Large crews working as a unit instead of their hands. Had this been done, are generally inefficient either because they stumble over one they would have quickly realized the futility another or because the work is of the tactics chosen. unbalanced, the first men bearing the brunt of the attack and the directing their men to work effi­ ning crew locations and actions. In stragglers expending their energies ciently, in discovering and remedy­ similar situations, the fire boss chiefly by running to keep up with ing weaknesses in their work, in should always size up the fire them and doing relatively little anticipating and planning actions, either from a highpoint such as a productive work. The morale of a and in urging the men toward ridge or tree top, or by cruising the crew weakens when the work load their best efforts. The crew leaders area by car, sending scouts out for is not evenly divided among all its should use their heads and eyes information, or by referring to aer­ members. instead of their hands. Had this ial photographs. It is strongly rec­ been done, they would have quick­ ommended that on a large fire the Equipment. On the whole, the ly realized the futility of suppress­ fire boss have at his disposal two or fire-fighting equipment was in ing the north flanks from the head three men to reconnoiter and to excellent condition. In only one toward the tail of the fire. Actually, serve as messengers to carry his instance was failure of equipment the physical efforts of the supervi­ orders to the leaders of the individ­ noted, namely, the railroad fusees sory personnel in suppressing the ual suppression crews. intended for stringing backfire, fire were of minor consequence, which would not ignite, undoubt­ considering the fire as a whole. Morale of Fire Fighters. The edly because they had become morale and determination of all damp from atmospheric moisture. The fire boss should make it his men were excellent, and in many The crew attempting to extend the job to keep up with every change cases remarkable. Virtually all of backfire along the west firebreak in the situation, know the location them used their flaps and back­ one-half mile (0.8 km) north of of all his crews, and continually pack pumps effectively, showing Highway 19 was delayed while try­ plan the action to bring the fire that the training they had received ing to make the fusees ignite. The under control at the earliest was very much worth while. result was that heads C, D, and moment. His decisions should be During the hot flank attacks, how­ later F crossed the graded fire­ direct, definite, and well-planned. ever, the flapmen relied heavily break, and eventually led to heads On the Honey Fire, the fire boss, upon the pumpmen spraying water G and H. The need for having all instead of placing himself at all to knock down the flames. The equipment in perfect order was times at a central point to gather men should be trained to rely less strongly exemplified by this one information regarding the situa­ upon water in fighting the flanks small but important failure. tion and to direct and dispatch by having the crew leaders tem­ crews, was off on the fire line with porarily stop suppression and rest The supply of tools for all fire suppression crews for considerable the crews when the wind shifts on fighters was automatic in that each periods of time. As a result, the a flank, resulting in a very hot fire crew had an adequate amount of desired movement of some crews to fight. More line on the flanks standard fire-fighting equipment in to critical areas was delayed. will be extinguished and held by the truck on which they traveled to resting a crew while the fire is the fire. No delays were noted on Anyone on the Catahoula Tower burning intensely and then effi­ this account. could have obtained an excellent ciently directing them when the grasp of much of the situation. heat and flames have diminished. The urgent need for accessory When the fire passed close to the equipment on fires of this type was tower the smoke was heavy and Crew leaders should strive to keep brought out by the handicaps and visibility was bad; later, however, their crews working in units of five difficulties encountered by the fire the view from the tower would or six men. A crew of this size is fighters. Emphatically, the crews have given the fire boss a compre­ very flexible and mobile and, when assigned to backfiring and hensive picture of the fire and trained for perfect coordination patrolling backfires should be sup­ helped him tremendously in plan­ and teamwork, it can hold a long plied with smokeproof goggles so

Fire Management Today 38 that they can work efficiently in More line on the flanks will be extinguished and the smoke and flying ashes. The held by resting a crew while the fire is burning value of such goggles was indicated by the fact that one fire fighter, intensely and then efficiently directing them when normally working as a welder in the heat and flames have diminished. the shop, wore his dark welding glasses while patrolling, and later Technique of Attack. The logical toward the head. Crew 6 of 22 men commented that he experienced no point of initial attack on the Honey and, if needed crew 7 of 19 men, great discomfort from ashes, soot, Fire, as already discussed, depend­ should then have been dispatched or sparks when the head reached ed on a very accurate location of its to patrol highway 19 and the south the backfire. Respirators should origin. As soon as the fire boss saw flank. The 2 remaining crews, com­ also be investigated to determine that the fire was definitely on the prising 31 men, if called at all by whether or not the patrol crews east side of the railroad track with the fire dispatcher and fire boss, could perform more effectively a large area of dense grass before it should have been held at the with such equipment. Since the for its run, he should have directed Cataholua Tower as stand-by crews. hose man on a pumper truck is his crew (No. 2) together with crew These positions are shown on the subject to intense heat and smoke 3 to start backfiring immediately suggested locations and tactics over a prolonged period, he also and liberally along the west side of map (fig. 3). should be provided with special the west firebreak. The pumper equipment to enable him to do his truck should also have been avail­ Initial backfiring should have been job better. Asbestos hoods and suits able at this point to support the started promptly along the west have already been developed for fire fighters. Crews 4 and 5, com­ firebreak instead of along Tower such use and might upon trial prising 47 men, should then have Road. The woods road, located one- prove ideal for this specific pur­ been dispatched to the tail to extin­ half mile (0.8 km) to the west, pose. guish the north flank from the tail could not have been used for the

Figure 3—Map showing suggested crew locations and tactics.

Volume 63 • No. 3 • Summer 2003 39 It is very important that backfires be sufficiently the Tower Road and the east fire­ long to stop the onrushing head even if a shifting break would necessarily mean deliberately burning part of the wind has changed its direction, and sufficiently plantation. The acreage consumed wide to prevent spotting across the backfire. in backfire, however, is negligible; each mile of backfire 100 feet (30 initial backfiring because (1) insuf­ At all places where an adequate m) wide requires only 12 acres (5 ficient time would have been avail­ firebreak had been burned, the ha). This would have been a trivial able to string an adequate backfire onrushing head was checked and loss for the great protection it along the woods road before the the few spots that started in the offered. main head reached it, and (2) there unburned grass across the fire­ could be legal complications and break were quickly extinguished by The following technique for string­ violation of policy had a backfire the patrol crew. The checking of ing backfire has been effectively been strung on privately owned heads B, D, and F are good exam­ and safely used and is recommend­ land not bordering Government- ples of control with backfiring. ed for use whenever backfiring owned land, even though national- When the backfiring is delayed or must be resorted to in order to forest land was seriously threat­ the patrol crew inadequate, as for obtain control: Organize the crew ened. Prompt backfiring along the example on heads C, G, and H, into fire stringers and patrolmen. west firebreak would have over­ break-overs occur almost invari­ The latter should take their posi­ come both of these difficulties. ably, greatly delaying control of the tions across the line from which fire. the backfire is being burned. Their At least 20 minutes would have only job is to keep alert for possible been available to the crews for Backfiring. Backfire can be safely spotting along the entire backfire stringing backfire along the west strung at a fast rate even with the line and to extinguish spots quick­ firebreak, since the main head did crudest, equipment, as shown in ly, as they occur. The stringers, not reach this point until 10:20 table 6. The greatest precautions three or four selected men in each a.m. This would have been ample must be taken, however, to keep crew, should be given special train­ time for the three crews available the backfire always under control ing in the methods of stringing to have strung at least one-half and to avoid the misfortune of let­ fire, using bunches of grass, a fire mile (0.8 km) of backfire 100 or ting it get away. At the same time, rake, a torch, or other available more feet (30+ m) wide. It was a backfire, to be effective, must be equipment. The first man should impossible to burn such a protec­ of sufficient width and length to string his line of fire parallel to and tion strip around the plantation hold the main head being fought as approximately 10 feet (3 m) from earlier in the season because the well as any additional heads that the line from which the backfire is land involved was privately owned. may subsequently develop before being made. The width will, of The 20-foot-wide (6-m-wide) grad­ the flanks are controlled. course, depend upon numerous ed firebreak surrounding the plan­ factors, among which are the type tation was scraped clean of all veg­ On the Honey Fire, the backfiring and density of the fuel, the wind etation and was an exce1lent line crews strung fire too timidly, par­ velocity, and the width and condi­ from which to backfire safely. ticularly insofar as the length of tion of the line (road, etc.) from the backfire was concerned. It was which it is being made. Had the initial attack of stringing most fortunate that the backfires backfire along the west firebreak were so successful, since they were The greatest precautions should be failed so that the fire spread into seldom more than 50 feet (15 m) taken to put the initial backfire the plantation, the next attack wide. Furthermore, had the crews line in safely. Waiting until the first should have been to string back­ not been so reluctant to string man has safely burned approxi­ field along the Tower Road, using long backfires, control of the fire mately 100 feet (30 m) of his line the stand-by crews available at could undoubtedly have been of fire, the second man should Catahou1a Tower. gained much earlier. This reluc­ start his line of fire parallel to but tance can in part be accounted for 20 feet (6 m) from the first line. by the fact that backfiring along The third stringer, in turn, should

Fire Management Today 40 wait until 100 feet (30 m) of the his office after attending a court The futility of attempting to con­ second line has been, safely burned trial in Alexandria, on his own ini­ trol the flanks by suppressing a and then string his line parallel to tiative phoned ahead to his dis­ rapidly spreading fire from a point but 30 or 40 feet (9–12 m) from patcher to call in all work crews for near the head toward the tail of the the second line. If four stringers the emergency and to send two of fire is brought out. Such tactics are used with intervals between them to stand by at a CCC side lead to a great loss in what would lines of 10, 20, 30, and 50 feet (3, camp located about 10 miles (16 otherwise be held line and make it 6, 9, and 15 m), a backfire 110 feet km) north of Bentley. The ranger is possible for new uncontrollable (34 m) wide can be burned with to be commended because he took heads to form with each relatively great rapidity. There should be an definite action when he saw the slight change in wind direction. interval of at least 100 feet (30 m) need. between stringers at all times. The heads of rapidly spreading fires Summary cannot be stopped by direct attack The crew foreman must be very A detailed analysis of the Honey with the equipment now available; alert when backfires are being Fire is presented in order (1) to one must resort to an indirect burned so that they do not get out show the rapid rate of spread and attack involving the use of backfir­ of control. If two crews are avail­ the behavior of a fire burning ing. Fighting fire with fire can be able, the backfire should be started under critical weather conditions very dangerous, however, and the at a point where the main head is in the southern pine type of a greatest care must be exercised in expected to hit and the crews coastal plain, (2) to describe the its use if an adequate backfire is to string backfire in opposite direc­ suppression action taken, and (3) be attained and if break-overs are tions along the line from which it to offer constructive criticism and to be prevented. A method of back­ is being burned. The crews should suggestions as a guide in planning firing, in which fire is simultane­ continue to string fire for at least suppression action for future fires ously strung by three or four men seven hundred or even 1,000 feet burning under similar conditions. separated by definite distance and (210–300 m) beyond the points width intervals, is outlined. It is where the danger is critical; a crew The combination of high, shifting very important that backfires be will not be criticized for stringing winds and low fuel-moisture con­ sufficiently long to stop the too much backfire if it has done so tent prevailing at the time of the onrushing head even in case its safely. fire created critical burning condi­ direction of burning has been tions. The rate of spread was changed by a shift in the wind, and Stand-by Crews and extremely high, the maximum for­ sufficiently wide to prevent spot­ Reinforcements. As previously ward increase measured being 8 ting across the backfire, which may brought out, all of the firefighters chains (528 feet [160 m]) in 1 start new heads. available on the ranger district minute or at a rate of a mile in 10 were dispatched to this fire before minutes (97 km/h). Acknowledgment it was finally brought under con­ Grateful acknowledgment is given trol. Two other fires occurred on In order to control such a fire it is to A.H. Antonie, R. Brooks, and the Catahoula Ranger District dur­ necessary to have an adequately C.A. Bickford for invaluable assis­ ing the Honey Fire; they were equipped suppression force avail­ tance not only in mapping and extinguished by crews dispatched able at a moment’s notice. The dif­ recording data but also in con­ from the Honey Fire without ficulties experienced by fire fight­ structive criticism and review of undue loss of time or acreage, indi­ ers at various parts of the Honey the manuscript. cating that the fire organization Fire are stated and suggestions are was prepared to cope with serious made for the use of accessory References fire conditions. equipment in overcoming such Bickford, C.A.; Bruce, David. 1939. A tenta­ handicaps. There is a distinct need tive fire-danger meter for the longleaf­ The local force was strengthened also for efficient supervision of slash pine type. Occ. Pap. No. 87. by the fine cooperation and judg­ Asheville, NC: USDA Forest Service, each crew, as well as able leader­ Southern Forest Experiment Station. ment of the ranger on the adjoin­ ship, including well-planned tac­ Headley, R. 1939. Lessons from larger ing district, who, upon passing the tics, by the fire boss. fires in 1938. Fire Control Notes. 3(4): scene of the fire while enroute to 40–41. ■

Volume 63 • No. 3 • Summer 2003 41 THE BOWER CAVE FIRE*

Leon R. Thomas

he Bower Cave Fire of August 13, 1947, on the Tuolumne Extended attack went according to schedule, and T District of the Stanislaus the night work was so efficient that by midnight National Forest is being reviewed most of the lines had been built and burned out. to show how a fast-moving fire, which was burning in steep terrain and in heavy cover, was readily and tered ponderosa pine and black oak First Report quickly controlled, after the first with a heavy ground cover of man­ The first report was received in the attack had failed, by effective use of zanita and scrub oak. Elevation district ranger’s office at 9:25 a.m. modern equipment, and by the ranged from 2,350 feet (720 m) at August 13, 1947. The ranger, fire local people and trained Forest Bower Cave to 3,400 feet (1,040 m) control assistant, and the district Service personnel working effec­ at the higher reaches of the fire clerk-dispatcher were in the office tively as a team. perimeter. The sets were on a when the report was phoned in by grassy, pine–oak flat with the the caretaker at Bower Cave. He Dangerous Fire Month steeper slopes and heavier cover reported that there were several August is a most dangerous fire just to the north. small fires burning on the upper month in this area. The weather is side of the McCauley road between normally hot, fuel moisture is low, For many years there has been an Bower Cave and the miners’ cabins and rapid spread of fires can always incendiary problem on this part of one-half mile (0.8 km) up the road be expected. At the time of this fire the ranger district. Sets are always (fig. 1). The area is blind to all normal weather conditions pre­ in high hazard types; and several lookouts and not until 9:40 a.m. vailed throughout the district and severe fires have been the result. did the lookout on Pilot Peak no important changes were fore­ This fire appeared to be another of report smoke coming up over the cast. It is estimated that there was that type. ridge that blanked out the area for an upslope wind of 8 miles per him. hour (13 km/h) blowing at the fire when the first crews arrived. It increased in velocity during the day but never blew hard. There is a normal downdraft at night with increase in humidity.

Bower Cave, a former resort, is located near the old Coulterville– Yosemite road where the road crosses the North Fork of the Merced River (fig. 1). The ridge between Bower Cave and McCauley ranch and Scott Ridge to the north of the river have moderate to steep slopes and are covered with scat-

When this article was originally published, Leon Thomas was a fire control officer for the USDA Forest Service, Sequoia National Forest, CA.

* The article is reprinted from Fire Control Notes 10(2) Figure 1—Map of Bower Cave Fire, Stanislaus National Forest, CA. [Spring 1949]: 26–31.

Fire Management Today 42 The Kinsley Station crew, being The early and efficient dispatching of the nearest organized fire suppres­ personnel and equipment by the district and sion crew, consisting that day of two men and a light pickup tanker, the central dispatcher was an important factor. was immediately dispatched. They arrived at the fire at 9:48 a.m. The at the toe of the steep slope. The crews, Pacific Gas & Electric Co. fire control assistant and one fire­ Bower Cave caretaker had also construction crew, and the local man arrived at 9:56 a.m. arrived and was assisting the fore­ ranchers. Two 15-main district man. road construction crews and a 50­ The ranger station crew of two man district blister rust crew were men and a tanker was dispatched Initial Attack Failure also ordered. A fire camp was to be as was the third organized fire The fire was burning very hot and set up at Bower Cave. It was calcu­ crew on the district, the McDiar­ was spotting badly up the steep lated that this number of men mid Station crew of four men and slope when the ranger arrived at could corral and hold the fire on a tanker. 10:05 a.m. He fell in with the other the ridge that afternoon with an men in attempting to cut off the estimated area of about 200 acres The district ranger helped the clerk head and control the spots. The (80 ha). After 10 a.m. the men notify the central dispatcher and fires which had now burned into began to arrive rapidly, as indi­ a few local people and was at the one were just too hot to handle cated by the number on the fire fire by 10:05 a.m. with a radio- and the light tanker was ineffec­ in table 1. equipped pickup. tive. The forest fire control officer flew The Kinsley crew found five sepa­ By 10:15 a.m. it was fully realized the fire at 11:15 a.m. in a conven­ rate fires burning on the upper that the initial attack had failed. tional aircraft and reported to the side of the road within a distance The heavier tankers from the ground by radio that the flanking of about 100 yards (90 m). Two ranger station and the McDiarmid action was making good progress Pacific Gas & Electric Co. power Station did not arrive until too late and that it had a very good chance line construction employees, who to be of value on initial attack. for success by the early afternoon. had seen the smoke while working on a nearby line, were already put­ The ranger and the fire control Spot Fire ting a line around the fire nearest assistant made plans for and imme­ At 12 noon the ranger and a local Bower Cave. These two men cor­ diately started a flanking action to rancher scouted the ridge in front ralled this fire, the smallest of the keep the fire narrow and possibly of the fire and kept in communica­ five, at about one-tenth of an acre pinch it out on the ridge above, tion with an SX radio. At 12:16 (0.04 ha) in size. The Kinsley fore­ should sufficient help arrive soon p.m. the lookout on Pilot Peak man left his one man on the sec­ enough. There was a very good reported a smoke in the bottom of ond fire and attempted to handle possibility of control on top of the the Merced River about one-half the other three alone with the aid ridge above Bower Cave since once mile (0.8 km) to the northeast of of a light tanker. These three fires the fire reached the top it would the original fire (fig. 1). In a few were the largest of the five and have to burn along the ridge or minutes the ranger could see the were rapidly burning together. downhill for a considerable dis­ smoke from his position on the They were burning in the grass and tance. The cover was also lighter ridge. It appeared to be burning on pine needles and were working along the ridge and on the north both sides of the canyon and toward the steep slope above. slope. spreading toward Scott Ridge and the McCauley ranch as well as back Little effective control work had The dispatcher was notified of the toward the original fire. been accomplished on the upper situation by radio and a request four fires when the fire control was sent in to the supervisor’s The forest fire control officer again assistant arrived at 9:56 a.m. He office to have the fire and the area scouted the fire from the air at 2 and the one man with him joined scouted from the air. Orders were p.m. The ranger in the meantime the Kinsley foreman in attempting sent in to get all the help possible had gone around and scouted the to cut off the head of the main fire from the local sawmills, woods new fire from the Scott Ridge side.

Volume 63 • No. 3 • Summer 2003 43 Table 1—Personnel on the Bower Cave Fire, by time of arrival.

Personnel on the fire Time of arrival Local labor Forest Service Cumulative total 9:30 a.m. 2 0 2 9:48 a.m. 1 2 5 9:56 a.m. 0 2 7 10:05 a.m. 0 2 9 11:00 a.m. 81 40 130 12:00 noon 21 0 151

Note: The fire was first reported at 9:25 a.m.

Through radio discussion with the arrived early but were of little use The fire control officer and the for­ fire control officer in the plane and on the original fire. est supervisor came into the camp with the ground scouting informa­ at about 4 p.m. With the aid of tion, it was determined that the Extended Attack scouting information and aerial fires would burn together before All effort was now turned toward photos the final control routes they could be controlled. It was handling the two fires as one along were determined. The fire was then decided that both fires should the following plan: The hand line divided into four divisions and the be handled as one. that had been constructed along division overhead personnel were the west side of the fire above briefed on the construction and the The cause of the spot fire was not Bower Cave was to be dropped in backfiring plan. By 6 p.m. all determined. It may have been to the river to the north and held. crews, tractors, and other equip­ another set. No attempt was made The front of the fire on Scott Ridge ment were on the line and pre­ to send men to it as it was spread­ was to be headed and a line pared for a night operation. ing rapidly when first discovered dropped to the river along each and an initial action crew would flank. The line on the west side was The plan went according to sched­ have been ineffective. to tie to the hand line at the river. ule and the night work was so effi­ Each of the lines from Scott Ridge cient that by midnight most of the It was realized now that control was a bulldozer show until they lines had been built and burned lines would embrace an area of a reached the steep river slope. A out. Many of the dangerous snags thousand acres (400 ha) or more line was to be built from McCaul­ were felled by power-saw crews and that a good deal of the line on ey’s over the ridge to the north and before the backfires were started. Scott Ridge and the McCauley area then to the river and tie to the east This was an important factor in was a bulldozer show. Additional line from Scott Ridge. The road reducing the possibility of spot tractors were ordered. Two D–7 from the McCauley ranch to Bower fires as well as cutting down mop- caterpillars were walked to the Cave was to be backfired. Four of up and patrol work later. The fire McCauley ranch from a nearby the bulldozers were walked to was declared to be under control Forest Service road construction Scott Ridge where two were to by 9 a.m. job. One D–7 caterpillar was work on each of the lines from trucked in from another Forest Scott Ridge to the river. Two trac­ Mop-Up Service road job on the district and tors were to operate from A look at the available Forest a TD–14, the Forest Service fire McCauley’s. Service manpower in the early standby tractor, was trucked in evening indicated that there was from the supervisor’s headquarters. The camp was now in full swing not enough for the mop-up job the Two bulldozers were already on the and all incoming men were organ­ next day. Needs were calculated fire, a D–6 from a nearby gold ized into crews with sufficient and an order was placed for one dredge and an AC tractor from a Forest Service overhead for good division team from another forest nearby sawmill. These last two management. and 150 off-forest laborers. The

Fire Management Today 44 division team was flown in from The use of aerial scouting and aerial photos the Sequoia Forest and the 150 for plotting the fire and the control line laborers were picked up in Stockton in the San Joaquin Valley. aided greatly in early control. All were at camp in time for the next day’s shift. The Sequoia team available. Exceptionally good coop­ portable SX sets. A mobile unit and did an excellent job on a division eration was received from the local then an SX set were used by the unit and returned to their home people—labor from the sawmills, fire boss. A mobile set was used in forest after one shift. woods crews with power-felling the fire camp. equipment, electric power line Mop-up proceeded rapidly during construction employees, and expe­ A telephone connection was made the early morning and the next day rienced local ranchers. There were to a nearby line and run to the fire with tractors widening lines, with 210 men on the line during the camp. This took a load off the radio tankers working along the bulldoz­ night shift and 272 on the line dur­ net. er lines and the roads, and with ing the next day. Men were power saws felling the remaining released rapidly after the end of the The use of aerial scouting and aeri­ snags. Especially important on second day’s shift. al photos for plotting the fire and mop-up was a 4-by-4 blister rust the control line aided greatly in spray rig. This four-wheel-drive The effectiveness of the work dur­ early control. This combined with unit with its long light hoses ing the first night was an outstand­ limited ground scouting proved reached many places that were ing factor in the early control of very effective. inaccessible to the conventional the fire. Control could not have tankers. been effected by 9 a.m. the follow­ The camp was well located near the ing morning, however, even with fire and was rapidly put in full Two Pacific marine portable the manpower available had it not operation by experienced person­ pumpers and hose were taken into been for the efficient work of the nel. Lunches, lights, water, and the river on the east side of the fire tractor operators in the heavy other equipment were always ready by pack horses and were used very manzanita cover. Lights on the six to go before departure time sched­ effectively on mop-up on the river tractors enabled them to work all uled for crews. Adequate trans­ slopes. The fire boss was equipped night. Wide effective lines were the portation was available and ably with a jeep and a portable radio result. Total perimeter of the fire coordinated under the camp boss. during the mop-up period. He was was 598 chains (39,468 feet [12,030 able to cover all of the fire lines in m]) handled as shown in table 2. This fire, burning in steep heavily the jeep except the steep river covered terrain, was readily con­ slopes. The fire was declared to be Excellent radio communication trolled before the second burning officially out on August 23. during the entire fire made admin­ period at 1,223 acres (495 ha) istration fairly easy. The radio net because of the effective use of mod­ Factors for Success centered around the Pilot Peak ern fire fighting equipment, the There were many factors working lookout who used a T set for excellent cooperation of local peo­ together that contributed to the receiving and relaying messages. ple, and the efficient work of ■ control of the fire prior to the Division bosses were equipped with Forest Service personnel. burning period of the second day. The most outstanding ones are list­ Table 2—Fireline on the Bower Cave Fire, by type. ed in the following paragraphs. Line constructed Line backfired The early and efficient dispatching of personnel and equipment by the Type of line Chains Meters Chains Meters district and the central dispatcher Hand 160 3,200 50 1,000 was an important factor. Men and materials were ready to go. Tractor 200 4,000 200 4,000 Sufficient experienced Forest Road 238 4,760 238 4,760 Service and local men were readily

Volume 63 • No. 3 • Summer 2003 45 THE POSSIBLE RELATION OF AIR TURBULENCE TO ERRATIC FIRE BEHAVIOR IN THE SOUTHEAST*

George M. Byram and Ralph M. Nelson

Editor’s note: Early issues of Fire Control Notes often prefaced arti­ If the degree of local stability in the atmosphere cles with substantive remarks by proves to be a key factor in unexpected the journal’s editors. A.A. Brown, and dangerous fire behavior, skill in control an early leader in Forest Service fire research and management, of large fires can be greatly advanced. prefaced the article by Byram and Nelson with the remarks in the the Coastal Plain of South Carolina fuel moistures which, combined sidebar. during a few days of the 1950 with high winds, resulted in a large spring fire season. number of fires and a large acreage Fire control men have long sus­ burned. Apparently there were two pected that there are unidentified Southeastern States experienced an rather definite types of severe fires. factors that contribute to the unusually severe season during The first, driven by high winds, was strange behavior and spread of that period. A prolonged drought, characterized by high rates of some fires. H.T. Gisborne, G.L. interrupted only by occasional spread, especially while crowning. Hayes, and A.A. Brown, among oth­ rains, began in November and per­ From the standpoint of the safety ers, have believed that atmospheric sisted in some areas until May. of suppression crews and their instability might in part explain This brought about abnormally low equipment, this type of fire has not some of the western blow-up fires. Brown (1950) and Crosby (1949) speaking more generally have stat­ Prefatory Remarks by A.A. Brown ed that when sufficient heat is gen­ erated by a fire in an unstable “Blowups” and other forms of unaccountable fire behavior, that charac­ atmosphere, erratic fire behavior terize many of our more disastrous fires every year, were a special topic can be expected. In a report from that was given much emphasis at the fire meeting held in Ogden, UT, Australia (Foley 1947), turbulence in January 1950. New research on this problem was urged and all is stated to be an important factor research men were urged too to make available every bit of new infor­ in the degree of severity of bush mation that might be useful to the fire strategist even though this fires and that it is of value in com­ might mean some reporting ahead of final evaluation. piling forecasts of fire weather. There is now reason to believe that I am very happy to present this report by George M. Byram and Ralph turbulence may also be associated M. Nelson, two members of the research group, as new information with certain severe fires in the that I personally believe is highly significant though further confirma­ South. Some evidence on this tion, evaluation, and means of prediction are needed and will require point was obtained from fires that much further investigation. If the degree of local stability in the atmos­ burned in an unusual manner in phere proves to be a key factor in the unexpected and often dangerous behavior of many of our large fires, and it can be identified in advance, one more of the unknowns will be eliminated and skill in control of When this article was originally published, large fires can be greatly advanced. George Byram and Ralph Nelson worked for the USDA Forest Service, Division of This preliminary report should be a challenge to all experienced fire Fire Research, Southeastern Forest Experiment Station, Asheville, NC. fighters and research men alike. Have we been ignoring one of the con­ trolling factors in big fire behavior? * The article is reprinted from Fire Control Notes 12(3) [Summer 1951]: 1–8.

Fire Management Today 46 been considered dangerous for In one instance, the plow crew observed flames experienced firefighters in the directly overhead while the main fire Southeast. The second type differed from the first in that its peculiar was still some distance away. whirling nature and unpredictable behavior made even a flank attack trouble for the plow crews. indicated a high degree of atmos­ dangerous. It is with the second Because the early spread of the fire pheric instability also on that day. type that this report is concerned. was nearly at right angles to the road, no short cut to the head was There were three large whirls in The Buckle Island and possible. Therefore, a flanking this fire and at least two small Farewell Fires* attack on both sides was made. As ones. The paths of the larger whirls Following is a description of two the plow crews progressed, the fire were approximately parallel and whirling fires that occurred on the on the left flank had a tendency to were separated by less severely Francis Marion National Forest in cross in front of the crew, and on burned strips 75 to 100 feet (23–30 South Carolina. So far as can be the right flank, behind the crew. m) in width. Needles on tree ascertained, they had characteris­ One large counterclockwise whirl crowns in the strips were not con­ tics of behavior common to fires or two such whirls, one on each sumed, and in a number of places that burned elsewhere in the flank, could account for this the tops of crowns of trees 25 feet Coastal Plain of that State during strange behavior. It was later found (8 m) high remained alive. How­ the five or six most severe days of that there was at least one small ever, in the paths of the whirls the the spring season. whirl on the right flank, although crown foliage was generally com­ from the plane observer Mitchum pletely consumed, on some trees to The Buckle Island No. 144 fire saw only one large whirl. a height of 80 feet (24 m). Needles burned on March 26 in a densely are not completely consumed stocked stand of loblolly pine 10 to The Buckle Island fire differed unless they are well within the 35 feet (3–11 m) in height. The day somewhat from other whirling flames, so it is estimated that the was sunny with little wind during fires in that it apparently main­ flames may have ranged from 50 to the morning hours. The relative tained a fairly constant direction of 150 feet (15–45 m) in height. humidity was medium (about 35 spread. The wind velocity was also percent) and the records of the greater and steadier than on other After becoming established, the Weather Bureau airport station, fires. Even so, the spread was errat­ whirls moved rapidly ahead of the located approximately 40 miles (64 ic. At times the fire would quiet main fire with sufficient updraft to km) from the burned area, indicat­ down and then suddenly burn with carry burning embers aloft. These ed a layer of highly unstable air fierce intensity. These bursts may embers are reported to have started about 400 or 500 feet (120–150 m) have been caused by the almost fires a considerable distance ahead deep at 10 a.m. The layer probably simultaneous ignition of several of the main fire. Airplane observer had become even more unstable acres by the whirl. In one instance, Mitchum believes that two of the and somewhat deeper at the time the plow crew observed flames three whirls burned at the same the fire started in the early after­ directly overhead while the main time. He also observed that flames noon. However, the increased tur­ fire was still some distance away. came out of the center of the tops bulence may have been partly off­ of the cone-shaped whirls. The set by an increase in wind velocity The most severe fire on the Francis flames did not spring directly which took place at about noon. Marion National Forest during the upward but had the same rotary spring season from the standpoint motion, spiralling upward, as the Apparently no large whirl devel­ of erratic behavior and its whirling smoke in the outer parts of the oped until the fire reached a size of nature was Farewell No. 172. It whirls. 40 or more acres (16+ ha). One burned on April 11, with a light then enclosed the head and created but variable southwest wind in a Some Characteristics stand of loblolly reproduction 10 to of Erratic Fires 35 feet (3–11 m) in height which * Acknowledgment is made to John T. Koen, formerly Evidence from the Buckle Island ranger on the Francis Marion National Forest, to John contained a scattering of mature and Farewell fires, and from others T. Hills, Jr., and Aiken Mitchum of the National Forest trees. Weather Bureau records staff for eyewitness accounts of the fires reported upon.

Volume 63 • No. 3 • Summer 2003 47 Fires with erratic behavior • In flat country it is doubtful that are most likely to occur on sunny days large whirls could develop if the air were absolutely calm, regard- when there is strong surface heating. less of turbulence. Some wind movement would be necessary to that burned in South Carolina dur­ size when born. After becoming move them over fresh fuel. A ing the spring season of 1950, indi­ established, the whirls apparently high wind, on the other hand, cates that erratic fires in the can move rapidly away from the would reduce turbulence and Southeast may have some common main fire and take the direction might also tend to break up the characteristics. Also, there appear of the light wind prevailing at whirls. This does not mean that to be certain conditions of weather, the time. They can consume fires burning in a high wind will fuel, and type—not fully identi­ strips of reproduction 500 to 800 be less intense than fires burning fied—which are conducive to such feet (150–240 m) wide. The in a light wind. A large majority fires. Although some of the follow­ velocity with which these whirls of severe fires probably burn on ing conclusions regarding fire travel is one of their most dan­ days of high wind velocity, and behavior and possible causes are gerous characteristics because rate of spread will increase with speculative, they appear reasonable their speed may be equal to, or increasing wind velocity. in view of what is known of certain nearly equal to, the velocity of • The effects of turbulence in areas physical laws. Confirmation or dis­ prevailing winds. of rough or rolling topography proval will require further observa­ • Field men state that most of the would be considerably more tion and analysis. worst fires occur on days with a complex than in flat country. southwest wind. This indicates a Turbulence near the ground sur­ • Fires with erratic behavior are characteristic pressure system face would probably never be as most likely to occur on sunny which may also account for some great as in flat country but this days when there is strong surface of the turbulence. Observers in would possibly be more than off­ heating. There may be little or airplanes have noticed that the set by complex topographic no wind during the early part of air was always bumpy on days effects. For example, large whirls the day, and even while the fire is when whirling fires occurred. could travel up slope rapidly even burning, usually in the after­ Also, when the flying became in an absolute calm. noon, the general wind is light or smooth in the late afternoon • An important factor in the occur­ moderate. The most favorable whirls did not occur. rence of the whirling type of fire wind for this type may be some­ • It cannot be assumed that whirls may be the fairly recent change where between 8 and 16 miles will always rotate counterclock­ in stand type in much of the per hour (13–26 km/h) as meas­ wise like large-scale vortex Southeast. During the past 15 ured 20 feet (6 m) above tree storms such as hurricanes in the years extensive stands of dense tops. northern hemisphere. The coun­ pine reproduction have become • Erratic fires have a tendency to terclockwise rotation of the hur­ established on areas formerly develop one or more violent ricane is caused by the rotation kept clear of pine by repeated whirls after reaching a certain of the earth. This should have fires. This may be one reason critical size. The size is probably but little effect on small-scale why there have not been more of not the same for all fires and whirls like dust devils or these fires in former years. On may be somewhere between 20 whirling fires. For this reason the other hand, they may have and 75 acres (8–30 ha). This, the chances are probably about occurred more often than is sup­ however, is merely conjecture. equal that the whirls will be in posed. It is difficult for ground • From the air, the diameter of either direction. crews to recognize large whirls larger whirls appeared to remain • The depth of the turbulent layer because of smoke and a limited approximately constant and to may be a dominating factor in field of view. They can be seen cover an area of about 10 acres determining the maximum size best from the air. (4 ha). An increase in size after of the whirls, although other • Perhaps too much emphasis they had formed was not variables such as quantity of fuel should not be placed on just the observed. Possibly they appear should also have some effect. whirling characteristics of fires suddenly and may be nearly full- burning in turbulent air.

Fire Management Today 48 Turbulence could have a pro­ nounced effect on the draft of a fire long before it reaches the whirling stage. It was noticed that on days when there was high turbulence, even small fires burned with strong drafts. The opposite of this has long been familiar to fire fighters. Fires usually undergo a pronounced change in behavior in late after­ noon and evening when the atmosphere becomes more sta­ ble. This change has often been attributed to the increase in rela­ tive humidity which accompanies the drop in temperature of the lower air layers. It is possible that an increase in air stability Figure 1—Ten a.m. lapse rates for March 26, April 11, 17, and 24, 1950. The straight dashed lines represent the dry adiabatic lapse rates, i.e., a decrease of 0.53 °F per hun­ may have as great, or greater dred feet (0.96 °C/100 m) in height. influence on behavior than a combination of increased fuel dred feet (0.96 °C/100 m) in There is usually some turbulence moisture and decreased fuel tem­ height. At this rate of decrease the on clear sunny days, but the aver­ perature. atmosphere is neutrally stable. The age value of the turbulence factor* greater the drop in temperature is not known for the Coastal Plain The Influence of with height, the greater the air in early spring. Its value may be Atmospheric Instability instability. Conversely, the less the somewhere between 20 and 50. In on Fires drop in temperature with height, contrast, the turbulence factors for Unusual fire behavior, not previ­ the less the instability. For exam­ March 26, April 11, 17, and 24, ously experienced, was reported for ple, on calm, clear nights, the air were respectively 110, 160, 390, the Francis Marion National Forest temperature often does not and 135. It is possible that there on March 26, April 11, 17, and 24, decrease with height but even were other days during the spring by suppression crews. This behav­ increases. The atmosphere is then season that had equally high tur­ ior, characterized by one or more highly stable and the upward bulence factors, but whirls or whirlwinds and by sudden fierce movement of smoke and heated erratic fire behavior were not upward bursts of flames, could not gas may stop completely after observed. If highly turbulent days be accounted for by any exception­ reaching a certain height. From did occur, it may be that fires were al conditions of fuel or wind. This the graph it will be seen that the controlled while small or before led the writers to suspect that unbroken lines, representing lapse they reached the breaking point, or some unusual atmospheric condi­ rates for the 4 days, inclined that they did not burn under the tions existed at the time of the sharply to the left of the dashed fuel and stand conditions most fires. Accordingly, 10 a.m. lapse- lines for a distance equivalent to a favorable to turbulence. Further rate records for the 4 days men­ height of 300 to 500 feet (90–150 analysis should clarify this point. tioned were obtained from the m). This means that layers of high­ Weather Bureau station at the ly unstable air existed at these * A turbulence factor T will be defined by the equation Charleston airport. These are depths. These conditions of air tur­ Le graphed in figure 1. bulence, coinciding with certain T = 100 – 1 fuel and stand conditions, and size {L a } The straight dashed lines in the of fire or rate of energy output, are graph represent the dry adiabatic believed to explain the strange fire where Le is the existing lapse rate and La is the dry adiabatic lapse rate. When Le = La, the air is neutrally lapse rates, that is, a decrease in air behavior on the days mentioned. stable and T = 0. Whenever T is greater than 0 there is temperature of 0.53 °F per hun- always some turbulence.

Volume 63 • No. 3 • Summer 2003 49 An important factor in the occurrence of the southeastern fires are not known, whirling type of fire may be the fairly recent but they are strong enough to carry burning embers for consider­ change in stand type in much of the Southeast. able distances ahead of the main fire. As has been pointed out previously, Atmospheric Instability very severe fires can occur on days and Dust Devils Williams (1948) gives the range in when the atmosphere is relatively size for dust devils as varying from There appears to be similarity in stable. On March 27 the turbulence 20 to 200 feet (6–60 m) in diame­ some of the conditions which favor factor was only 16, but this was a ter and from 10 to 4,000 feet the development of whirls on some severe fire day. As a result of high (3–1,200 m) in height. It thus erratic fires and dust devils. These wind velocity—30 to 40 miles per appears that the largest dust devils are strong surface heating on clear hour (48–64 km/h) with gusts occupy an area only about one- days, and winds of not more than reaching almost 60 miles per hour tenth as great as the area of the moderate velocity. Ives (1947) gives (97 km/h) at the Charleston air- larger whirls on the South the following account of the condi­ port—there were intense, fast- Carolina fires. The main difference tions favorable for their formation: spreading fires which did great between dust devils and whirls on damage. There was nothing erratic fires is that the former must obtain In geographically favorable or baffling in their behavior, how­ all of their energy from the poten­ areas dust devils occur most ever, that could not be explained in tial energy of the atmosphere, frequently in clear weather, terms of wind and fuel conditions. whereas the latter obtain their when the surface has been energy from burning fuel as well as heated for some hours, and When a fire burns in a stable the atmosphere. In the same arti­ there is little surface wind. atmosphere, the hot gases must cle Williams states: Under these conditions the not only expend energy as they lift surface air is very hot with their masses through the stable air, These occurrence times were respect to that a few hundred but they also expend part of their from 1 to 5 hours before the feet aloft. Typically favorable energy in dragging a part of the times of maximum tempera­ conditions, measured during a surrounding air upwards. The sta­ tures. The reason for this fact “Great-Basin-High regime” ble air acts like a ceiling so that on is that the wind speeds nor­ are: surface temperature, 160 a calm clear evening the smoke ris­ mally increase as the times of °F (71 °C); one foot (0.3 m) ing above a fire will reach a certain maximum temperatures are above surface, 142 °F (61 °C); height and then level off. The con­ approached and certain criti­ five feet (1.5 m) above surface, ditions are entirely different when cal speeds are reached beyond 116 °F (47 °C); 500 feet (150 a fire burns in an unstable atmos­ which the dust whirls cannot m) above surface, 100 °F (38 phere. The gases do not expend exist. These critical speeds °C); 2,000 feet (610 m) above energy as they rise but in their have not yet been determined, surface, 92 °F (33 °C). ascent they may even acquire ener­ but vary with lapse rate, gy from the atmosphere. Their topography, and probably Such a pronounced drop in tem­ path upward creates a chimney other factors. perature means, of course, an into which the surrounding unsta­ extremely unstable atmosphere ble air is drawn. The potential Brown (1950) states that dust dev­ near the ground. Ives further states energy of the unstable air is then ils are an ominous sign to fire that the upward velocity of the air converted into kinetic energy as it fighters. Whirls of a similar nature in a dust devil may exceed 35 miles enters the chimney created by the on fires may account for many per hour (56 km/h) and that meas­ fire. When the total rate of energy blow-ups. ured horizontal winds within the release (rate of energy output of whirl can accelerate from near zero fire plus rate of energy change in to speeds of from 50 to more than Lapse Rate is Related the unstable atmosphere) is great 90 miles per hour (80–140+ km/h) to Fire Control enough, then whirls should devel­ and then return to their former If the conclusions reached regard­ op. velocity within 30 to 100 seconds. ing the effect of air turbulence on Velocities within the whirls on the fire behavior are substantiated by

Fire Management Today 50 additional work (see the sidebar), a Although erratic fires in the Southeast new aspect of fire control in the may not be common, their potential danger Southeast will have been recog­ nized. Although erratic fires in this to suppression crews will justify taking extra section may not be common, their precautions during especially hazardous periods. potential danger to suppression crews and damage to timber short, they could be warned to ability of this green fuel for com­ stands, particularly in the younger expect crowning and the sudden bustion was increased by an unsta­ age classes, will justify the taking formation of large whirls, unusual ble atmosphere plus a high rate of of extra precautions during espe­ backfire behavior, exceptional rates energy release in the ground fuels. cially hazardous periods. Radio­ of spread considering existing wind sonde observations, where avail­ velocities, gustiness and quick References able, will be helpful but the extent changes in wind direction, and the Brown, A.A. 1950. Warning signs for fire of the adjacent area to which these likelihood of danger even in mak­ fighters. Fire Control Notes. 11(3): apply will have to be determined. ing flank attacks. 28–30. Forecasts of high impending tur­ Crosby, J.S.1949. Vertical wind currents and fire behavior. Fire Control Notes. bulence a day or two in advance It should be emphasized again that 10(2): 12–15. would be most useful, although a the changing fuel and stand types Foley, J.C. 1947. A study of meteorological forewarning of even a few hours conditions associated with bush and occurring in the Southeast may be grass fires and fire protection strategy in might mean considerable for the a necessary condition for the large Australia. Commonwealth of Australia, safety of men and equipment. whirling fires which burned in Bureau of Meteorology, Bulletin. 38: South Carolina last year. These 51–52. Suppression crews during such Ives, R.L. 1947. Behavior of dust devils. fires burned in dense stands of Bulletin of the American Meteorological periods could be alerted to make reproduction (predominantly Society. 28(4): 168–174. the fastest possible attack so as to loblolly pine) in which the compact Williams, N.R. 1948. Development of dust restrict any fire to the smallest whirls and similar small-scale vortices. crowns constituted the main Bulletin of the American Meteorological possible acreage and before it source of fuel. In turn, the avail­ Society. 29(3): 106–117. ■ reached the breaking point. In

Additional Evidence

Since this report was written, data have been obtained from the Weather Bureau which give the upper air temperatures at the Charleston airport for all days in the period from March 20 to April 30. Although a com­ plete analysis has not yet been made, these data indicate that there were only eight days in this period with a highly unstable atmosphere. Four of these days were March 26, April 11, 17, and 24 when severe whirling fires occurred.

On the other four unstable days, no whirling fires were reported. On April 27, the atmosphere was very unsta­ ble at 10:00 a.m., but 0.41 inches (1.04 cm) of rain fell later in the day before 5:00 p.m. Similar turbulent conditions existed on April 6 and 27, but 0.33 inch (0.84 cm) of rain fell on April 5 and 0.60 inches (1.52 cm) on April 27 and 28. The chances were very slight for fires starting and building up to a high rate of energy output on these days, especially in dense stands of young loblolly pine. On April 19 there was a highly unsta­ ble layer of surface air but it was only 150 feet (46 m) deep. In addition, the next layer above was deep and sta­ ble. It is doubtful if large whirling fires could develop on such a day. However, the shallow layer should have had a marked effect on the behavior of small fires.

Volume 63 • No. 3 • Summer 2003 51 THE PINYON–JUNIPER FUEL TYPE CAN REALLY BURN*

Dwight A. Hester

n the Rocky Mountain Region, we are rapidly losing any illu­ We soon learned that natural barriers, I sions that our fuel types are of the “asbestos” variety. Aspen used such as ridges, cliffs, and roads, to be considered fairly fireproof were of no value in heading off this type of fire. until certain crown fires, gathering speed in adjacent conifer stands, western Colorado. Land managers deep underground for some 20 rolled through without loss of were not concerned since there years, chose to explode. This explo­ momentum. The moist, high-alti­ had been normal snowfall during sion, according to an eyewitness, tude spruce type has been even the winter, and the early spring occurred at 3:10 p.m., and the fire more deceptive on disastrous occa­ had been cold, if dry. The spruce seemed to be in the crowns at sions. type well above the pinyon–juniper once. By 5 p.m. the fire had trav­ still held considerable snow. By eled about a mile (1.6 km) “on the But at the lower elevation, in the June 10 the weather had turned back of a strong wind” and showed southwestern part of the region, is warm, and strong winds came up no signs of abating (fig. 2). the familiar pinyon–juniper type, with regularity during the after­ and this never gave any trouble. noons. Relative humidity was down Most of it is outside the national Extreme Fire Behavior to 7 percent. forest boundaries, and it is usually We soon learned that natural barri­ grazed so heavily that all fuel is ers, such as ridges, cliffs, and It was during this period that a gone except the trees themselves roads, were of no value in heading coal mine, abandoned and burning (fig. 1). The records show that our off this type of fire. The country neighbor to the south, the Mesa Verde National Park, had a big fire in such a type in the drought-rid­ den thirties, but that seemed to be a “one-in-a-million” occurrence. Unusual Conditions Then, in 1950, we suddenly found out that under extreme conditions the fuel-sparse pinyon–juniper type will not only burn, but will literally explode. Since this type is wide­ spread through the Southwest, perhaps other fire control person­ nel could profit by our experience.

During the early part of June 1950, the weather was fair and dry in

When this article was originally published, Dwight Hester was a district ranger for the USDA Forest Servicae, Grand Mesa National Forest, CO.

* The article is reprinted from Fire Control Notes 13(1) Figure 1—Typical pinyon–juniper type, showing scattered stand, sparse vegetation, and [Winter 1952]: 26–29. intermingled areas of bare ground. Photo: USDA Forest Service.

Fire Management Today 52 Figure 2—The fire as One cannot count on the oak seen from a point 15 miles away, 2 hours brush above the pinyon–juniper after origin. Photo: type to serve as a buffer. On our USDA Forest Service. fire, the oak brush, although only about one-half leafed out, burned readily and crowned out in most places. As was found in Maine in 1947, hardwoods are not immune to crowning. Fire Effects was too broken and rocky for bull­ Suppression Tactics Although the bulk of the trees dozers to be used effectively. The Judging from the behavior of our remain standing after the fire, the shaggy bark of the juniper made fire, I believe that the head of such heat is quite intense and leaves the fire brands to Satan’s liking. a fire should not be attacked until ground well cooked (fig. 3). Re­ Flaming strips of this bark, often 2 the crowning stops, unless there growth of any kind is bound to be feet (0.6 m) or more in length, are means available for creating slow and erosion will be a problem. were hurled ahead to wrap them­ extremely wide barriers. Once the On the fire described, the wind selves around other trees which fire is out of the crowns, men can started drifting the soil before the caught fire with a roar and gave off work relatively close to the fire and fire was out and continued throug­ ropelike strips of bark to repeat the can work in most of the burn with­ hout the summer. Only two rain­ process. in 2 hours. storms of relatively light intensity occurred during the summer, but Distance between trees and width I believe the best bet is to fell a small gullies were in evidence by of natural barriers seemed to have swath of burning trees at least 100 fall. little influence on this type of yards (90 m) wide, working from spread. In one instance, a cleared, the edge toward the interior of the While our pinyon–juniper type can 40-foot (12-m) fire lane was burn. One power saw per 4-man hardly be classified as a high fire crossed in its entire length by the crew seems to be the answer for risk, it is not fireproof. When con­ fire without detectable hesitation. this work. In this short-tree type 2 ditions are right, it can be quite Backfiring was not practicable men can operate the saw with a explosive, resulting in fires that are since the only fuel was standing reasonable degree of safety, and the difficult to control. A burn in this trees which had to be crowned out other 2 haul away the felled debris. type will be slow to heal and can to burn, and a crowning juniper in Mop-up usually has to be done result in a long-term watershed a high wind is not to be fooled with little or no water since much problem. ■ with. of this type is without “living” water of any kind. Not only can the fire explode dur­ ing the afternoon, it can continue this blowup well into the night. On Figure 3—The intensi­ our fire, the expected evening wind ty of the fire denudes the soil to a point shift did not take place until about where watershed dam­ 8 p.m. This occurred as a 90­ age of long duration degree change of direction (a will result. Ditch and down-mountain draft) with no gully in the foreground were cut before the appreciable change in wind veloci­ area burned. Photo: ty, and the fire really rolled down­ USDA Forest Service. hill. The rapid rate of spread con­ tinued until 11 p.m., at which time the wind velocity fell from an esti­ mated 20 to 30 miles per hour (32–48 km/h) to a gentle breeze.

Volume 63 • No. 3 • Summer 2003 53 A FIREWHIRL OF TORNADIC VIOLENCE*

Howard E. Graham

hirlwinds occasionally have been reported occurring The winds in this tube were so extreme W within various types of that a Douglas-fir, which at breast height fires. Accounts sufficiently detailed was about 40 inches, was quickly twisted to give the reader a definite idea of what the reporter had actually seen and broken off are rare. Since the fire-whirlwind is a phenomenon of considerable the flash of a powder keg when the reformed a moment later, repeat­ importance to fire fighters, I will whirl passed by. Within the tube, ing this procedure at least 3 times attempt to describe one which was gases and debris were moving during a 10-minute interval. observed by Robert S. Stevens, upward at a high velocity. The Forester, Oregon State Board of whirling column remained nearly The general fire was on a 50-per­ Forestry, and myself at 2 p.m., stationary during its activity, mov­ cent south–southwest slope. The August 23, 1951, on the Vincent ing a little more than 50 yards (45 trees were widely spaced with fuels Creek Fire in southwest Oregon. m). Had that not been the case, consisting of low brush, weeds, Figure 1 portrays the spectacular extremely rapid fire spread might snags, and down logs typical of an wind conditions. have resulted. The whirlwind rap­ old burn in this region. The fire idly disappeared and as rapidly front was moving steadily along Tornado-Like Tube From our vantage point about 200 yards (180 m) away it was evident that violent whirling surface winds existed over a diameter of some 100 to 200 feet (30–60 m). In the middle of this circulation was a dark tornado-like tube which extended upward, the top being obscured by drift smoke above approximately 1,000 feet (300 m). The winds in this tube were so extreme that a green Douglas-fir tree, which at breast height was about 40 inches (100 cm) in diam­ eter, was quickly twisted and bro­ ken off about 20 feet (6 m) above the ground.

In the area of the whirlwind, the fire flames leaped several times higher than those surrounding. A large tree top burst into flame like

When this article was originally published, Howard Graham was a meteorologist for the Fire Weather Service, U.S. Weather Bureau, Portland, OR.

* The article is reprinted from Fire Control Notes 13(2) [Spring 1952]: 22–24. Figure 1—Diagram of a fire-whirlwind observed on the Vincent Creek Fire.

Fire Management Today 54 the contours and extended up the Meteorologists know that these whirlwinds entire slope, about one-half mile are present only where the atmosphere is in (0.8 km) from top to bottom. Flames along the front were about a particular condition of unstable equilibrium. 5 feet (1.5 m) in height. Shorter flames persisted to a distance of tends to rise, and conversely, the edge of the main fire progressed, about 50 feet (15 m) behind. cooler air aloft tends to sink. The the fuels in the area of the whirl result is intensive vertical currents were consumed and the volume A slight spur ridge projected from throughout the unstable layer. and heat of the ascending gases the slope so that updrafts were became apparently insufficient to moving from both the south and In this case, we have the heat from support the whirlwind. As the fire the southwest into the area of the the fire which caused the unstable moved on to new fuel and new whirlwind. The fire-whirlwind conditions. However, this is an topographic features no further developed a few feet behind the fire entirely normal situation over a disturbance was noted. front 150 yards (140 m) from the large fire. Since these violent fire- summit of the main ridge and on whirlwinds are infrequent, there From this analysis it would seem the spur ridge. must be some condition other than likely that there are certain ideal heating to cause their formation. combinations of conditions under Atmospheric Instability Perhaps the answer lies in the which this type of fire-whirlwind of The meteorological condition of interplay of wind currents and extreme violence might occur. The the atmosphere was one of condi­ topography. necessary factors seem to be for tional instability. Overturning of the fire to be on the lee slope shel­ the air in the lower layers could Interplay of tered from the prevailing ridge top readily occur if the surface were Wind/Topography winds, a moderate or stronger heated sufficiently. No cumulus In the case under discussion, con­ wind at the ridge top, and strong clouds were to be seen. Winds at sider the position of the whirl near converging surface updraft cur­ ridge top were north–northeast the top of a sunlit south–southwest rents along the burning or sun- from 10 to 15 miles per hour slope where it was fed by upslope heated slopes. (16–24 km/h). Above the ridge top drafts from the south and south­ level, winds were from north– west in the surface layers. Above More Accounts northwest to north–northeast and the level of the ridge the rising Needed ranged from 8 to 16 miles per hour currents from the fire were played It would be desirable to have the (13–26 km/h) up to 10,000 feet upon by the prevailing gentle to necessary combination of condi­ (3,000 m). The relative humidity moderate north–northeasterly tions more positively identified so was 46 percent and the tempera­ wind. that fire fighters could learn to ture 67 ºF (19 ºC), neither of which anticipate at least this one type of was unusual. Perhaps herein lies the answer. blow-up fire behavior. Additional There were two opposing air cur­ detailed accounts of similar phe­ There has been much written on rents with a column of rapidly ris­ nomena would contribute to the various types of whirlwinds and ing gases between. This is an ideal understanding of their causes and their causes. Much is yet to be condition for the formation of their effects on fire behavior. These learned. Meteorologists know that mechanically induced eddy cur­ accounts should attempt to these whirlwinds are present only rents. An eddy current, once start­ describe the topography, surface where the atmosphere is in a par­ ed, might be sustained by the ener­ wind, ridge top wind, fire intensity, ticular condition of unstable equi­ gy of the rising hot gases. This the­ cloud types, smoke column charac­ librium—where the temperature ory is substantiated by the repeated teristics, and the intensity of the decreases so rapidly with height reappearance of the fire-whirlwind fire-whirlwind. ■ that the warmer air below, being in the same spot. As the leading lighter than the cooler air above it,

Volume 63 • No. 3 • Summer 2003 55 RATE OF SPREAD ON A WASHINGTON FERN FIRE*

William G. Morris

ate of spread, fuel moisture, and climatic conditions were In behavior, this fire typified many R measured on the 4,000-acre early spring fires on cutover areas (1,600-ha) Livingston Mountain in the foothill zone of the Douglas-fir region. fire east of Vancouver, WA, on April 11 and 12, 1951. In behavior, this fire typified many early spring fires previous fires and occurred in large was 16 percent, rates of spread on cutover areas in the foothill blocks, small patches, or as single beyond the shelter of trees varied zone of the Douglas-fir region. trees. Scattered Douglas-fir seed­ from 660 feet per hour to 7,900 Some of the measurements of rate lings and saplings were also pres­ feet per hour (200–2,400 m/h) of spread will be given as an exam­ ent in some areas of bracken and (table 1). On April 12, when the ple of this type of fire. brush. wind speed was 7 to 8 miles per hour (11–13 km/h) and the relative Burning Conditions The weather was unusually favor­ humidity was 26 percent, rates of The fire burned across a gentle to able for fires. The preceding 10 spread beyond the shelter of trees moderate south-facing slope that days had been clear and unseason­ varied from 80 feet per hour to extended some 3 miles (5 km) from ably warm with dry winds. Before 1,600 feet per hour (24–490 m/h) level farmland to the top of a broad this period, rainfall had been nor­ (table 2). hill, 1,500 feet (460 m) higher. At mal. On April 11 the relative the steepest points, the slope was humidity fell to 14 percent, and On April 11 the fire spread so rap­ about 40 percent. wind speed measured with a idly that control through flank portable anemometer 6 feet (1.8 attack would have required con­ As in much of the foothill country, m) above the ground was consis­ struction of control line at a rate of ground cover was mostly western tently 18 to 20 miles per hour more than 1/2 mile per hour (0.8 bracken fern, intermingled with (29–32 km/h). Moisture content of km/h) along both flanks. During annual weeds, trailing blackberry, dead bracken fuel was 9 percent or short periods when flames rose in and salal, an evergreen shrub 6 to less even though the soil was still the crowns of scattered conifers 15 inches (15–38 cm) high. In this moist 1/4 inch (6.4 mm) below the and embers flew far ahead, more locality, bracken grows to a height surface. than 1-1/2 miles of control line of 2-1/2 feet (76 cm) and forms a would have to be constructed per dense ground cover. In April, how­ Rapid Rates of Spread hour (2.4 km/h). In flash fuel of ever, the dead crowns are bent to Lineal spread and contributing fac­ this kind, a frontal attack with a the ground and form a loose, flashy tors were measured on heads and slower rate of line construction fuel about 1 foot (30 cm) deep. flanks of the sprawling, irregular could be effectively used to check Hazel brush, about 8 feet (2.4 m) fire. Measurements were taken the head of the fire. tall and not yet in leaf, formed a mostly during short periods when sparse overstory. Here and there wind speed, direction of spread, Typical Rates of pole-size Douglas-fir had survived topography, and fuels were fairly Spread uniform. Spread was usually meas­ Although effects of wind speed and When this article was originally published, ured for a distance of 50 feet (15 relative humidity cannot be sepa­ William Morris was a forester for the m) or less although occasionally rated in these observations, rate of USDA Forest Service, Pacific Northwest for 0.1 to 0.5 mile (0.16–0.8 km). spread apparently increased 4 to 5 Forest and Range Experiment Station, Portland, OR. On April 11, when the wind speed times when wind speed increased was 18 to 20 miles per hour (29–32 2-1/2 times and was accompanied * The article is reprinted from Fire Control Notes 15(1) km/h) and the relative humidity by an appreciable drop in relative [Winter 1954]: 32–34.

Fire Management Today 56 humidity. Measurements on many Northwest. Spread is of course slash of these forests, rate of spread other fires show that a lineal grater on steep slopes. is greater than shown in table 2 spread of 400 to 900 feet per hour owing to large numbers of wind­ (120–270 m/h), as recorded in In the litter of the normally dense borne embers that set new fires far table 2, is typical of fires in brush, coniferous forests, rate of spread is in advance of the surface fire. ■ bracken, and weed cover on nearly a small fraction of that shown in level ground in the Pacific table 2. In the unburned logging

Table 1—Rates of spread and contributing factors measured on a Washington fern fire April 11, 1951.

Rate of lineal spread Direction of spread Slope a (ft/h) Position on fire with reference to slope (percent) Remarks 7,900 Head — 0 Crowns of scattered conifers afire. Spotted ahead. Measured on 0.1-mi spread. 3,300 do — 0 No spotting. 2,400 Flank Down 15 Spread only 350 ft/hour while flames drew inward just previous to this measurement. Drafts then became turbulent and flames surged outward. 2,400 Head — 0 — 2,400 do Down 25 Some spotting. Measured on 0.3-mi spread. 2,100 do — 0 Some spotting. Measured on 0.5-mi spread. 1,320 Flank Down 15 — 1,320 Head — 0 Flames 6–8 ft high and intermittently drawing outward or inward. 1,320 do Down 10 Measured on 0.5-mi spread. 1,000 do — 0 — 800 Flank — 0 — 660 do — 0 — 400 Head — 0 In dense stand of Douglas-fir 50 ft tall. Wind at treetops about 20 mph, and at 6 feet above ground, 4 mph. Fire spread in twig litter 2–3 in deep and produced a blazing border 4 ft wide in which flames were vertical and less than 2 ft high. 100 do — 0 Same stand and wind as above. Flames drew inward toward fire and formed a border only 1 ft wide.

Note: Wind, east 18–20 mph; relative humidity, 16%; temperature, 71 ºF to 75 ºF. a South aspect.

Volume 63 • No. 3 • Summer 2003 57 A lineal spread of 400 to 900 feet per hour is typical of fires in brush, bracken, and weed cover on nearly level ground in the Pacific Northwest.

Table 2—Rates of spread and contributing factors measured on a Washington fern fire April 12, 1951.

Rate of lineal spread Direction of spread Slope a (ft/h) Position on fire with reference to slope (percent) Remarks 1,580 Head Up 5 Spread at right angles to overhead wind. The flames, 1–3 ft in height, drew outward although just previous to this measurement they drew inward. 990 do — 0 — 730 do — 0 — 630 do Down 10 (E) — 530 do — 0 Flames 1–3 ft high in strip 2 ft wide drew inward. 500 do Down 10 (E) — 360 do — 0 — 340 Flank — 0 — 280 Head Down 10 (E) — 200 do Up 5 Spread at right angles to overhead wind. 150 do Up 5 Spread at right angles to overhead wind. Flames drew inward. 130 Rear — 0 Bracken flattened to 3-in layer. 120 Head Up 5 Spread at right angles to overhead and wind. Flames drew inward. 110 Flank Up 5 — 80 Head Up 5 Do.

Note: Wind, west 7–8 mph; relative humidity, 26%; temperature, 77 ºF. a South aspect, except for three observations (indicated by “E”) on east aspects.

Fire Management Today 58 FIRE-WHIRLWIND FORMATION AS FAVORED BY TOPOGRAPHY AND UPPER WINDS*

Howard E. Graham

fire started at a logging oper­ ation during the afternoon of Local violent winds, frequently whirlwinds, have A October l, 1952. Toward many times caused unusually rapid fire spread evening the size had slowly due both to direct fanning and to spotting. increased to 20 acres (8 ha). About 9:30 p. m. the fire suddenly became a raging inferno as a nearly cold 2-acre (0.8-ha) slash whirlwind frequently has a central whirling winds formed within the fire on the west slope of the tube made visible by whirling fire and abruptly multiplied its Washington Cascades near the smoke and debris. Extreme varia­ speed to such strength that chunks Columbia Gorge. Although the tions in height, diameter, and of wood and bark up to 8 inches lookout on a ridge a short distance intensity are common. Witnesses (20 cm) in diameter were thrown eastward reported east winds from have described fire whirlwind about like straws. Logger fire fight­ 50 to 60 miles per hour (80–96 diameters from a few feet to sever­ ers fled for their lives. Within min­ km/h), these winds had not been al hundred feet and heights from a utes the fire raced though un­ hitting the fire area. Suddenly an few feet to about 4,000 feet (1,200 burned areas for half a mile (0.8 intense whirlwind formed in adja­ m). Intensity varies from that of a km), increasing to 240 acres (100 cent green timber and passed over dust devil to a whirlwind that ha). The whirling winds remained the dormant slash fire. The fire pitches logs about and snaps off over the fire for about an hour, leaped to life with an eruption of large trees (Graham 1952). hurling burning embers for con­ sparks and flame and ran for over Velocities in the vortex are siderable distances and preventing a mile (1.6 km) finally joining a extremely high, and, as in other the loggers from pressing their second fire (whirlwind 9). forms of whirlwinds, the greatest attack (table 1, whirlwind 4). speed occurs near the center. A Fire whirlwinds have received lit­ strong vertical current at the cen­ Firewhirls and Large tle attention from meteorologists, ter is capable of raising burning Fires probably because such winds are debris to great heights. Fire whirlwind is a phenomenon usually observed only by foresters that has been known to be associ­ and fire fighters who are too busy Favorable Topographic ated with large fires (whirlwinds 1, fighting fires to make detailed Features 2, and 3) (Hissong 1926). It has weather observations. With greater The 28 fire whirlwinds that form become more common in recent attention being given to the study the basis for this discussion were years in the Northwest as a result of blowup fires (whirlwinds 1 and all observed in mountainous ter­ of the increase in number of nec­ 4), it is fitting that this particular rain. Their individual characteris­ essary slash burning operations. type of violent fire behavior be tics are indicated in table 1. Of the explored from both the empirical 28 whirlwinds, 20 occurred on lee In another fire that occurred and theoretical standpoints. slopes, 1 under calm conditions, 2 November 8, 1952, a Crown with wind at right angles to the Zellerback fire patrolman was Winds are of great concern to fire slope, and 4 on windward slopes. making a routine 8:30 a.m. visit to fighters. Fire spread is a function Of the several additional whirl­ of wind speed, although not a sim­ winds described to the author and ple function. Local violent winds, When this article was originally published, not included in table 1, all Howard Graham was a fire-weather fore­ frequently whirlwinds, have many occurred on lee slopes. caster with the U.S. Weather Bureau, times caused unusually rapid fire Portland, OR. spread due both to direct fanning The mechanical action of airflow * The article is reprinted from Fire Control Notes 18(1) and to spotting. A typical fire over a mountain is a factor in fire [Winter 1957]: 20–24.

Volume 63 • No. 3 • Summer 2003 59 Table 1—Descriptive details for 28 fire whirlwinds in the Pacific Northwest.

Whirls Relation Ridgetop wind Size of debris wind to Whirlwind Date, hour a Number Duration Diameter Height picked up topography Direction Velocity 1 7/20/51, 1600 1 2 h 1,200 ft 2,500 ft logs, lee slope N — 30 in by 30 ft 2 8/23/51, 1400 3 10 min 200 ft 1,000 ft large tree lee slope N–NE 10–15 mph broken off 3 9/21/51, 1300 1 1 h 200 ft 4,000 ft logs, lee slope N–NE — 15 in by 15 ft 4 10/1/52, 2140 Several Several min 50 ft 200 ft 6- to 8-in lee slope E 30 mph each for 1 h chunks 5 10/15/52, 1500 Several each 30 sec 40 ft 150 ft 3 by 4 lee slope NE 10 mph by 16 in 6 10/24/52, 1030 Several each 30 sec 40 ft 300 ft log, lee slope E 10 mph 12 in by 16 ft 7 10/25/52, 1100 Several each 2 min 20 ft 150 ft bark, ridgetop W — 2 by 4 by 12 in 8 11/4/52, 1400 1 few min Unknown Unknown Unknown parallel to SW 10 mph contours 9 11/8/52, 0830 1 Unknown Unknown Unknown small debris lee slope E 50–60 mph 10 11/8/52, 1500 1 1 h 5 ft 100 ft small sticks lee slope E 10 mph 11 9/ ?/53, 1100 1 10 min 50 ft 100 ft small limbs and snapped tree top, 8- to 10-in diameter flat clam — 12 9/16/53, 1345 1 2 min 75 ft 125 ft bark and wood, lee slope S–SW 4 mph 4–5 lb 13 9/21/53, 1530 1 3/4 h 25 ft 100 ft 6- to 8-in diameter lee slope SW 15 mph 14 9/29/53, 1000 Several each 30–60 sec 10 ft 50 ft up to 8 sq in windward SW 10 mph 15 9/29/53, 1130 1 1-1/2 h 400 ft 400 ft up to 3 by 3 in lee W 2 mph 16 10/3/53, 1600 Several each 2 min 10 ft 30 ft up to 15 lb windward? E 30 mph for 8 h 17 10/4/53, 1150 Several 3 min 40 ft 300 ft 2 by 6 by 18 in and lee slope NE 5 mph a cedar post 18 10/6/53, 1530 1 1-1/2 h 50 ft 200 ft Unknown calm calm — 19 10/6/53, 1530 Several 1/2 h 20 ft 60 ft bark, 4 by 6 in lee slope E 8 mph 20 10/7/53, 1500 Several each 30 sec 30 ft 100 ft 1-1/2 in diameter windward? SW 10 mph 21 10/8/53, 1600 Several each 4–20 sec 10 ft 200 ft under 11b windward W 10 mph for 4 h 22 10/9/53, 1640 1 8 min 140 ft 170 ft 2 by 6 by 24 in lee slope NE 5 mph and larger 23 10/9/53, 2000 Several 1–2 min each 75 ft 100 ft 1 by 2 by 3 in lee slope SW 10–15 mph for 2 h 24 10/10/53, 1730 5 1 h 50 ft 125 ft large sparks lee slope W 5 mph 25 10/13/53, 1830 Several 1 min each 30 ft 200 ft small twigs lee slope W 8 mph 26 10/15/53, 1500 Several each 1–2 min 50 ft 150 ft branches and bark lee slope W 5 mph for 1 h 27 10/29/53, 1830 1 10 min 100 ft 80 ft bark and limb, lee slope N 3 mph 4 by 5 in 28 10/29/53, 1500 Several each 10 sec to 30 ft 60 ft large material lee slope W 5 mph 4 min for 2 h including a 20-lb piece of sheet metal a Pacific standard time.

Fire Management Today 60 whirlwind formation. Aerodynamic Fire whirlwinds have become more common theory tells us that favorable condi­ in recent years in the Northwest as a result tions for the starting of a whirl occur where abrupt edges of of the increase in number of necessary mountainous terrain create shear slash burning operations. in the air stream. As has been found true with dust devils, shear­ ly unstable because of intense heat­ According to a U.S. Weather ing motion is undoubtedly a major ing near the ground. Bureau study of strong winds over factor in whirl formation. Although mountain barriers, the pressure mountainous terrain provides The distribution of upper air wind reduction over a mountain crest many topographic situations favor­ velocities also was checked from was proportional to the square of able to fire whirlwind occurrence, pilot balloon data at the nearest the wind speed. Where the air was the fact that it is not an essential weather station. The results saturated, the pressure deficiency condition is indicated by several showed that 75 percent of the was nearly doubled. The greatest examples which occurred on flat­ whirlwinds reviewed occurred, pressure deficiency occurred along land in Eastern United States with winds of less than 17 miles a mountain barrier with a ridge (Byram and Nelson 1951). per hour (27 km/h) below the profile corresponding to the upper 5,000-foot (1,500-m) level. surface of an airfoil where the Meteorological This is to be expected since the maximum drop would be near the Aspects majority of whirlwinds were on topmost part of the airfoil camber. Dust devils are normal in flat areas controlled burns. The remaining Theoretically a topographic barrier when the wind speed is low and the 25 percent showed rapid wind should best approximate an airfoil lapse rate is steep, i.e., relatively speed increase with height. The when the lee slope is less than 33 rapid temperature decrease with wind speed profiles are of variable percent and relatively smooth. height. Fire whirlwinds also appear shape and show no typical occur­ This corresponds very closely to to depend upon steep lapse rates in rence of the “jet point” discussed the upper limits of the change in the layer near the ground. Roy R. by Byram (1954) with relation to direction of airflow over the upper Silen, Pacific Northwest Forest and blowup fires. surface of airfoils on slow speed Range Experiment Station forester, airplanes. moved a fire whirlwind downhill by Mountain Barriers and rolling debris against the fuel in Their Effects on Conclusion the hot spot over which it had Airflow Because of the direct relationship formed. As the hottest portion of between fire whirlwind occurrence the fire was carried down the slope, The upper end of a fire whirlwind and combustion heat, the meteor­ the fire whirlwind followed. Fire when on a lee slope near a ridgetop ologist can predict likely areas of whirlwind occurrence seems to be seems to extend into a region of occurrence only if he is familiar directly related to the local thermal low pressure that occurs in the with both the attendant meteoro­ instability set up by the fire and vicinity of a ridgetop whenever logical and topographic conditions not otherwise relieved. windflow is at right angles to the ridge. This follows the Bernoulli and the occurrence of heavy fuel concentrations. The forester with The degree of upper air stability as principle which states that changes intimate knowledge of areas under indicated by the lapse rate between in pressure are inversely propor­ his management will usually be 850 and 500 millibars, i.e., pres­ tional to changes in fluid velocity. more able to predict combustion sure surfaces near 5,000 feet and Pilots are taught this principle as heat over a given area. 18,000 feet (1,500–5,500 m), at the explanation altimeter errors nearby weather stations has little experienced over mountains. Fire whirlwinds seem to develop or no effect on fire whirlwind more readily on lee slopes close to occurrence. Data on the lapse rate The theory of pressure reduction ridgetops. It is suggested that this at lower levels is unavailable. along a ridge oriented at right is favored by pressure deficiencies Obviously the lapse rate in the angles to the direction of airflow is resulting from flow over an lower level over the fire is extreme­ well supported by evidence.

Volume 63 • No. 3 • Summer 2003 61 The most favorable condition fighters should consider the danger for fire whirlwind occurrence is of fire whirlwind formation. over a hot fire near the top of a steep lee slope References with strong winds over the ridgetop. Byram, G.M. 1954. Atmospheric conditions related to blowup fires. Sta. Pap. 35. Asheville, NC: USDA Forest Service, abruptly terminating airfoil. The wind occurrence is over a hot fire Southeastern Forest Experiment Station. wind velocity above the ridgetop near the top of a steep lee slope Byram, G.M.; Nelson, R.M. 1951. The possi­ ble relation of air turbulence to erratic thus becomes an important factor with strong winds over the fire behavior in the Southeast. Fire in determining the likelihood and ridgetop. Fire whirlwinds are fre­ Control Notes. 12(3): 1–8. magnitude of a whirlwind. quently characterized by destruc­ Graham, H.E. 1952. A fire-whirlwind of tornadic violence. Fire Control Notes. tive violence. Therefore when any 13(2): 22–24. We may conclude that the most fire—large or small, quiet or run­ Hissong, J.E. 1926. Whirlwinds at oil-tank favorable condition for fire whirl­ ning—is on a lee slope, the fire fire, San Luis Obispo, Calif. Monthly Weather Review. 54: 161–163. ■

Fire Whirlwinds Studied in the Lab*

Because of their importance as a hazard to firefighters and as a cause of rapid and erratic fire spread, fire whirlwinds are one of the fire behavior phenomena being studied at the Southern Forest Fire Laboratory in Macon, Ga. These whirlwinds can be produced readily on a small scale and studied by modeling techniques.

*See Byram, G.M.; Martin, R.E. 1962. Fire whirlwinds in the laboratory. Fire Control Notes. 23(1) [Winter 1962]: 13–17.

Fire Management Today 62 RELATIONSHIP OF WEATHER FACTORS TO RATE OF SPREAD OF THE ROBIE CREEK FIRE*

R.T. Small

Original editor’s note: The Robie Creek Fire in the Boise National The greatest blowup potential Forest, ID, September 5–9, 1955, is is when the wind reaches maximum speed described, and concurrent weather within the first 1,000 feet above the fire conditions are analyzed. The fire exhibits four different types of and then decreases in speed with elevation. behavior during the 5 days. On four of the days, the behavior fol­ an analysis of the relationship of that day was 101°F (38 ºC) and the lows patterns previously recog­ those conditions to the fire behav­ relative humidity was 6 percent nized as being usually associated ior. resulting in a very high fire danger with the prevailing weather condi­ (Burning Index of 72 on the Forest tions. The exceptions occur on the There are several reasons why this Service Model 8 Meter). third day, which is meteorological­ fire adapts itself to an analysis of ly similar to the second day but this type: (1) The fire occurred only The fire apparently started on the exhibits a different fire behavior. 10 to 15 airline miles (16–24 km) east side of the Boise Ridge and at Some implications that this study northeast of the Boise Weather a point on a minor slope exposed has for forecasting and research Bureau Airport Station (WBAS) to the southeast. The point of igni­ are pointed out. where regular surface and upper tion was in well-cured grass in a air observations are made. (2) The light stand of chokeberry brush. Many observations have been made fire area was bracketed by two fire- Fuel in the general area consisted regarding the cause of forest and weather stations, Shafer Butte mostly of dry grass, several kinds range fire spread and a number of Lookout, six miles north of Robie of brush, and second growth pon­ well-qualified men have made Creek at an elevation of 7,590 feet derosa pine. The fire started at an investigations and contributed (2,313 m), and Idaho City Ranger elevation of about 5,000 feet (1,500 valuable reports and technical Station some 12 miles (19 km) m), but eventually spread over an papers on this complex subject. northeast of the fire, at an eleva­ elevation range from 4,000 to There is general agreement that tion of 3,950 feet (1,204 m), in the 5,500 feet (1,200–1,700 m). weather is the most important main Mores Creek drainage. (3) Although winds were light and variable in fire spread, and that the The fire went through four differ­ variable, the other factors were conditions which lead to “blowups” ent types of behavior-day: a very conducive to fire spread. are very complex and difficult to blowup, a long run, a potentially Within 2 hours of the time that predict. critical but quiet day, and a quiet fire began there were 15 to 20 peo­ day. ple from the nearby Karney Lakes This paper consists of a report of Resort, four smokejumpers, and a the weather conditions which Description of the Fire crew of 20 trained fire fighters at existed during the Robie Creek The Robie Creek Fire in the Boise the scene, but the rate of spread Fire in the Boise National Forest, National Forest started in the early was so great that the fire fighters Idaho, September 5–9, 1955, and afternoon of Labor Day, September had to retreat from the fire area. 5, 1955. It was a hot, dry day; the 45th day since there was measura­ The fire started on Monday, When this article was originally published, September 5, and was brought R. T. Small was with the Weather Bureau, ble precipitation in that area and U.S. Department of Commerce, Boise, ID. the 21st consecutive day with the under control on Friday, Septem­ maximum temperature above nor­ ber 9. Of the 5 days, major runs or * The article is reprinted from Fire Control Notes 18(4) “blowups” occurred on 3 days: [Fall 1957]: 143–150. It is an abridged version of an mal. The maximum temperature at article under the same title in the January 1951 issue of Monday, Tuesday, and Thursday. Monthly Weather Review. nearby Idaho City Ranger Station

Volume 63 • No. 3 • Summer 2003 63 Forecasters on large fires should consider Plotting the maximum surface carefully the wind speed profiles and surface temperatures at Shafer Butte, temperature distribution as well as temperature Idaho City, and Boise WBAS on the soundings show that superadiabat­ lapse rates and other observational material. ic lapse rates existed on Monday and Tuesday near the surface, but On Wednesday there were minor humidity at different heights. The the layer near the surface was flareups, but no sustained run decrease in temperature with alti­ more stable on Wednesday and occurred. There was very little tude is called the lapse rate. When Thursday. spread on Friday as established this value becomes 5-1/2 °F per lines were widened and mopup 1,000 feet (10 °C/1,000 m) the The surface conditions as shown in commenced (fig. 1). lapse rate is known as the dry adia­ table 1 reveal that the weather was batic lapse rate. With lapse rates hot and dry all 5 days, but that During the 5 days the fire spread considerably less than dry adiabat­ there was a definite cooling on over 8,310 acres (3,363 ha) of pri­ ic, the atmosphere is more stable. Thursday and Friday. vate and national-forest land. At Where the lapse rate approaches or the peak of the attack over 700 is greater than the dry adiabatic The wind speed profiles for the men were employed and total sup­ rate the air becomes unstable and 0800 mountain standard time pression costs were in excess of upward motion is greatly in­ (MST) and 1400 MST Boise winds $100,000. creased. aloft observations are shown in fig­ ure 3. The wind speeds over 7,000 Weather Conditions On the assumption that stability feet (2,134 m) above mean sea level increased gradually during In the attempt to determine which would be an important factor, a the first 4 days of the fire and then weather parameters had the most comparison was made of the twice- slacked off again at the end of the influence on the fire behavior dur­ daily Boise radiosonde observations week. ing the 5-day period, comparisons (fig. 2). The lapse rate was very were made of the various weather nearly dry adiabatic on Monday, data. The upper air measurements Tuesday, and Wednesday and only Fire Behavior give the values of temperature and more stable on Thursday and The fire behavior on Monday was Friday. very similar to that of Tuesday and most of the weather data were strikingly similar on those 2 days, except for minor changes in the winds aloft patterns.

Monday and Tuesday both had some of the characteristics associ­ ated with a blowup pattern; i.e., steep lapse rates, high tempera­ tures, low humidity, dry fuel, and relatively light winds aloft. On both Monday and Tuesday the major spread occurred in the mid­ dle and late afternoon and was accompanied by a nearly vertical smoke column which was topped by a well-developed cumulus cloud. Both Monday night and Tuesday night the smoke filled the surrounding valleys and remained Figure 1—Total area of the Robie Creek Fire showing location where fire started on Monday, September 5, 1955, and its spread on succeeding days. Grid interval equals 1 low until upslope motion mile (1.6 km).

Fire Management Today 64 On Thursday cooler air was obvi­ ously moving into the fire area with moderate westerly winds across the Boise Ridge and down onto the fire. In the early morning the fire was moving rapidly up the slopes exposed to the west, and throughout the morning and after­ noon the fire continued to spread in an easterly direction. Maximum temperatures were down about 20 F° (11 ºC) from Tuesday and mini­ mum relative humidity was up 10 to 20 percent. Although the fire covered nearly as great an area on Figure 2—Radiosonde temperature observations at Weather Bureau Airport Station this day as on Tuesday the behavior Boise, Idaho, during period of Robie Creek Fire. Daily maximum temperatures for Shafer Butte Lookout, Idaho City Ranger Station, and Boise Airport are plotted at their relative was different. The wind was rela­ elevations. tively consistent in both speed and direction and the fire moved from west to east, up slope and down. commenced at 1000 MST on Shafer Butte and 8 °F (4.5 ºC) at The forest officials described it as Tuesday and 1100 MST Wednesday. Idaho City. Winds aloft were weak­ more of a steady “push” than a er at low elevations and stronger at blowup. The smoke column leaned On Wednesday the fire spread over high elevations as shown by the to the east and although small only about 500 additional acres wind speed profiles. On Wednesday cumulus tops appeared frequently (200 ha) compared to over 3,000 there was no towering cloud- they disappeared almost as quickly acres (1,200 ha) on Tuesday. capped smoke column, only small as they formed. However, the temperature lapse areas of billowing smoke during rate was almost as steep as on the the afternoon. In contrast to the On Friday winds were light and previous 2 days and the minimum previous nights, the fire continued variable, temperatures were about relative humidity at Idaho City and to spread during the night, espe­ the same as on Thursday, and the Shafer Butte was the same as on cially near the ridgetops, and there relative humidity was higher by 5 Tuesday. There were minor was very little smoke hanging in to 10 percent. In the afternoon a changes in maximum temperature the valleys Thursday morning. few minor dust whirls were visible with a drop of 6 °F (3.3 ºC) at

Table 1—The maximum temperature and 1600 mountain standard time relative humidity for the 5 days of the Robie Creek Fire, Boise National Forest, ID, September 5–9, 1955.

Boise Weather Bureau Airport Idaho City Ranger Station Station Shafer Butte Max. Rel. Max. Rel. Max. Rel. Day temp. humidity temp. humidity temp. humidity Monday 97 °F 24% 101 °F 6% 84 °F 12% Tuesday 98 °F 23% 100 °F 12% 83 °F 14% Wednesday 97 °F 17% 92 °F 12% 77 °F 14% Thursday 81 °F 27% 80 °F 19% 62 °F 34% Friday 80 °F 30% 81 °F 25% 62 °F 40%

Volume 63 • No. 3 • Summer 2003 65 If it were possible to dispatch a mobile radiosonde light to medium fuels. Byram has observational unit to large fires the information classified the wind speed profiles into four main types, each with gained would be very valuable to the forecaster two or more subtypes (fig. 3). in predicting fire behavior. In comparing the wind speed pro­ in the ashes and smoke stumps, The rapid spread on Monday and files of the 1400 MST Boise winds but at no time was there a serious Tuesday corresponded to situation aloft reports for the 5 days of this flareup or threat to the firelines. 5, and the conditions on Thursday discussion we find that the profile By this time the suppression attack seemed to fit situation 4. On for Monday resembles Byram’s was organized and lines were well Monday and Tuesday there Type 1-a except for wind speed. The established and manned. appeared to be a “chimney effect” wind blowing up slope tended to reaching to an estimated 25,000 to offset this low velocity. Discussion 30,000 feet (7,600–9,100 m) which The wind speed profile at 1400 Arnold and Buck (1954) have listed induced a strong draft at the base of the column. MST on Tuesday for Boise closely five atmospheric situations under resembles Byram’s Type 3-a with which fire blowups may occur: Byram (1954) states that for the the jet point just above the fire greatest blowup potential the wind zone. This type has strong winds at 1. Fire burning under a weak high levels, but with a layer of inversion. should reach a maximum within the first 1,000 feet (300 m) above decreasing speed just above the jet 2. Fire burning in hot air beneath point. Byram says of this particular a cool air mass. the fire and then decrease in speed with elevation for the next several profile “for a fire near 7,000 feet 3. Combustible gases from a fire (2,100 m) it resembles the danger­ accumulating near the ground. thousand feet. He refers to this point of maximum wind speed ous Type 1-a and it is doubtful if 4. Fire exposed to a steady-flow the wind speeds at high levels are convection wind. immediately above the fire as the “jet point” and states that the wind strong enough to shear off the con­ 5. Fire burning near a cell of verti­ vection column.” Type 3-a and 3-b cal air circulation. speed near the jet point for most dangerous fires will be 18 to 24 may be accompanied by strong miles per hour (29–39 km/h) for whirlwinds and rapid fire spread when jet point winds are 20 miles per hour (32 km/h) or more. The winds at the jet point level at Boise WBAS were below Byram’s mini- ma, but speeds must have been higher just above the fire. Fire crews reported spotting as much as a quarter of a mile (0.4 km) ahead of the fire Tuesday afternoon which would indicate some of the whirl­ wind activity mentioned by Byram.

On Wednesday the wind speed pro­ file resembles Byram’s Type 1-b, except that wind speeds in the fire zone were much below the limits shown. The strong winds above 10,000 feet (3,000 m) would tend to prevent formation of a convec­ tion column which might induce Figure 3—Daily winds aloft observations taken at Weather Bureau Airport Station, Boise, strong winds at the surface. Colson during period of Robie Creek Fire (upper graphs) compared with Byram’s wind speed pro­ (1956) states “the convection file types.

Fire Management Today 66 column will not attain great the Boise radiosonde observa­ This study indicates that the fore­ heights if the wind speed increases tions (fig. 2) it appeared that casters on large fires should con­ too rapidly with height. Too strong there must have been a supera­ sider carefully the wind speed pro­ a wind speed may cause the col­ diabatic lapse rate near the sur­ files and surface temperature dis­ umn to be broken away from its face at Idaho City and Shafer tribution as well as temperature energy source.” Butte on Monday and Tuesday lapse rates, surface weather charts, which was not nearly so pro­ and other observational material. If Byram’s Type 4-a resembles the nounced on Wednesday. This it were possible to dispatch a mo­ wind speed profile and also the fire superheating effect was at a bile radiosonde observational unit behavior on Thursday. Regarding maximum on Monday and to large fires the information Type 4-a Byram states “fires were Tuesday, was at a minimum on gained would be very valuable to intense and fast-spreading, but Wednesday, and gradually the forecaster in predicting fire they could not be considered dan­ increased again on Thursday and behavior. The cost of constructing gerous to experienced crews, nor Friday. and operating a mobile radiosonde were there any erratic and unusual 4. Minimum relative humidity was unit would be considerable, but in aspects to their behavior.” the same both days. view of the terrific property losses 5. Maximum temperatures were and suppression costs on large The speed profile at 1400 MST on the same at Boise and 5 °F to 8 fires, such a unit would be justi­ Friday closely resembles Byram’s °F (2.7–4.5 ºC) lower at Idaho fied. Pilot balloon observations Type 2-a, but other conditions City and Shafer Butte on would be impractical because of reduced the fire danger. Wednesday, but that change in visibility restrictions, and only very itself hardly seems great enough rarely does a large fire occur close The fire behavior on Monday, to be critical. enough to an upper air observa­ Tuesday, Thursday, and Friday fol­ 6. The winds aloft at Boise WBAS tional station to make the data rep­ lowed previously recognized pat­ show minor differences in direc­ resentative of conditions over the terns usually associated with the tion on the 2 days, but wind fire. prevailing weather variables. How­ speed profiles (fig. 3) varied con­ ever, the meteorological similarity siderably. Byram’s wind speed Acknowledgments between Tuesday and Wednesday profile types are different for the Our thanks to George M. Byram was remarkable while the fire 2 days and they offer a possible and Charles C. Buck of the U.S. behavior was very different. Follow­ explanation for the variation in Forest Service and to DeVer Colson ing is a comparison of the 2 days: fire behavior between the 2 days. of the U.S. Weather Bureau for their reviews and comments on the 1. Fuel conditions on Wednesday Conclusions first draft of this paper. Our thanks were essentially the same as on The principal objective in an analy­ also to the staff of the Boise Tuesday with fuel remaining on sis of this type is to develop means National Forest for their patience all sides of the fire. Lines had of improving forecast and warning in answering questions and supply­ been established on some of the techniques. Byram’s wind speed ing data. fire boundary, but the long run profiles have considerable merit, as the following day indicates that the evidence has shown, but, from References the spread potential was present. a forecaster’s standpoint, it would Arnold, R.K.; Buck, C.C. 1954. Blow-Up 2. Figure 2 indicates that stability be difficult to separate the blowup fires—Silviculture or weather problems? was not the prime differentiating days from the quiet days on the Jour. Forestry. 52: 408–411. factor. Byram, G.M. 1954. Atmospheric conditions basis of projected 0800 MST wind related to blow-up fires. Sta. Pap. 35. 3. When the maximum surface speed profiles. This is a field in Asheville, NC: USDA Forest Service, temperatures at Idaho City, which a further study seems war­ Southeastern Forest Experiment Station. Shafer Butte, and Boise were Colson, D. 1956. Meteorological problems ranted. associated with mass fires. Fire Control plotted on the tephigram with Notes. 17: 9–11. ■

Volume 63 • No. 3 • Summer 2003 67 A KEY TO BLOWUP CONDITIONS IN THE SOUTHWEST?*

Robert W. Bates

an minimum nighttime tem­ peratures be used in some A deadly one-two combination of an C areas as an indicator of one unusually warm night followed by a warm day type of blowup conditions? A pre­ may indicate blowup fire conditions. liminary study of several project fires occurring on the Tonto, Sitgreaves, and Prescott National action taken? Why can you reach Too, the charts on the 13 fires Forests in the years 1951 to 1961 some lightning fires while they are studied actually indicate a better showed that the night before each still in the tree, yet others explode tie-in using minimum rather than of these fires blew out of control into major fires? Why does a quiet maximum temperatures. was unusually warm. Of particular or apparently controlled fire sud­ significance is the fact that most denly act up? A look at relative Fire control organizations are not of them occurred following the humidity showed day-to-day fluctu­ fully aware of this change in condi­ warmest nights of the critical ations and seemed not to be an tions as it is not indicated in pres­ June fire period and often oc­ adequate answer to these ques­ ent fire-danger meters by any defi­ curred at a peak after several con­ tions. This study seems to indicate nite rise in the index. During June, secutive days of rapidly rising that a deadly one-two combination the Southwest is in extreme condi­ temperatures. For some fires of an unusually warm night fol­ tions already—so it might be said which occurred in July and Sep­ lowed by a warm day may be the that conditions have suddenly gone tember this also appeared to be key. from critical to supercritical. true. Only 4 of the 13 fires in the study failed to show this, but even If further study should prove this Critical Nighttime for those 4 the temperatures were to be reliable, we could determine Temperatures at or above what is believed to be more accurately when to increase An attempt to expand on this theo­ the critical point. Tempera-tures emergency fire forces and signal ry and determine an average date on the nights preceding the start, the start of intensive fire preven­ at which this temperature rise or blowup, of these fires varied tion. Following lightning, extra occurred proved futile. It can from a high of 81°F (27 °C) in the efforts to ensure early detection apparently happen almost any time semidesert to 52 °F (11 °C) in the could be undertaken. By taking in June in this area and may occur pine above the Mogollon Rim. 8:00 a.m. readings of the previous as early as May or as late as July in These temperatures were all night’s minimum temperature and other southwestern localities. In unusually high for the area where plotting them on a graph, it might 1956, the condition seems to have the fire occurred. be possible to spot the beginning of appeared in September on the potential blowup conditions. There Tonto National Forest on the One-Two Combination is usually a very sharp rise from Buckhorn Fire, coinciding with a Why, in June, are some fires con­ relatively cool nights to hot nights dry fall period. However, June 10 is trolled at small size while others over a period of only 2 or 3 days often mentioned as the breaking defy control no matter what the (figs. 1–4). Since this leaves very point on the Tonto. Each forest— little time to get ready, the use of possibly each district—would have nighttime temperatures may be a to chart this separately and watch When this article was originally published, better indicator than daytime tem­ for the start of the temperature Robert Bates was a district ranger for the peratures because it allows more rise. From this temperature study, USDA Forest Service, Tonto National time to prepare. Forest. I have arbitrarily said that night­ time temperatures above 45° (7° C) * The article is reprinted from Fire Control Notes 23(4) [Fall 1962]: 95–99.

Fire Management Today 68 are critical; and with those above 55° (13° C) blowup conditions exist. Cloudy nights keep night­ time temperatures high and may or may not be serious depending on whether the clouds disappear by morning. During June, there is probably only a small likelihood of nighttime clouds.

It is recognized that factors other than temperatures also contribute to fire size. Some areas have large fires during periods of high early spring winds; some fires are large because of organizational break­ downs, others because of topogra­ phy; California has its Santa Ana winds; and so on. Undoubtedly, Figure 1—Roberts and Hatchery Fires, June 1961. most of these fires would show temperature correlation only through coincidence. On the fires studied there was no attempt to make a complete analysis of all fac­ tors affecting the particular fire such as topography, wind, relative humidity, human error, time of day, fuels, and aspect. What was suggested by the study is that this one common denominator may provide a predictable basis for increased manning and a crash prevention effort during the criti­ cal periods. Assumptions and Recommendations Some assumptions and recommen­ dations that can be made from this limited study follow.

1. High nighttime temperatures do not of themselves cause fires to blow up, but under these condi­ Figure 2—Russell Gulch Fire, June 1951. tions, all other factors which tend to cause large fires are maximized.

Volume 63 • No. 3 • Summer 2003 69 2. If nighttime temperatures are rising, going fires must be secured before temperatures rise above the critical point. This is seen in the case of fires which blow up on the third or fourth day after start. 3. Fires occurring before and after temperature peaks are con­ trolled at small size; some of them under much worse rate-of­ spread conditions. 4. Spotting was a big factor in most of these fires although winds were not exceptionally strong. This fact was mentioned consistently in discussing these fires with people who had partic­ ipated in the suppression. Some of these fires became big even though firefighters were on Figure 3—Boulder and Pranty Fires, June 1959. them at the very start when they were only a few feet across. 5. In June, before the summer lightning period, temperatures can be used as the basis for increased prevention effort at all levels. 6. When lightning occurs, detec­ tion forces could be augmented, especially in the case of long aerial patrol routes where some areas are not covered until 3 or 4 hours after daylight. 7. It might be possible to develop new rules for prescribed burn­ ing. Limit burning to times when nighttime temperatures are less than 45 °F (7 °C) and to a time when the temperature trend is down. 8. In the June dry period, rapidly rising nighttime temperatures often seem to presage the first lightning storm of the season. Figure 4—Buckhorn Fire, September 1961. These high nighttime tempera­ tures usually occur from 24 to 48 hours before the lightning storms occur. These first storms are often dry. ■

Fire Management Today 70 A FIRE-WHIRLWIND IN ALABAMA*

Gordon Powell

n the afternoon of February 7, 1962, Forest Ranger George This smoke column was different from the others O Nunnelee and I were making in that it bent upwards and had a “mushroom” of routine equipment inspections in vapor that penetrated the stratum “barrier.” Covington County, AL. At approxi­ mately 2:30 p.m., while on higher elevations in the north end of the haze layer toward the southeast. road was the boundary on the county, I commented to Mr. Nevertheless, this column was dif­ north; firelines were maintained Nunnelee that a tall smoke in the ferent from the other visible on the east and south. The fire had south end of the county had the columns in that it was the only one burned about 500 yards (450 m) appearance of a potential “blowup that bent upwards and had a from the north–south road. fire.” “Blowup” seemed highly “mushroom” of vapor beginning to improbable because of the condi­ form immediately over the smoke Ribbon of Smoke tion of vegetation following a 0.1­ column that penetrated the stra­ At approximately 4:50 p.m. we inch (0.25-cm) rain two nights tum “barrier.” passed by about one mile (1.6 km) before. My comment was based on north of the fire and observed that the appearance of the smoke which As we neared the control fire the the eastern part of the burn was to me indicated adverse wind pat­ billowing smoke indicated that it the hottest. Driving southwest to terns over a control fire approxi­ was very hot. Approximately 2 approximately one-half mile (0.8 mately 21 miles (34 km) due north hours elapsed between the time we km) west of the fire we noticed of us. first noticed the smoke until we something unusual about the west­ arrived in the vicinity. The fire had ern part in that there was a Singular Smoke been set along a north–south road smooth appearing ribbon of Column to back eastward into the wind smoke, steady and lighter in color through some heavy logging slash The column of white smoke (fig. 1) than the surrounding smoke. This and grassy vegetation. An east–west formed an angle of approximately “ribbon” of smoke ran from the 75 degrees from the ground, rising toward the southwest. At approxi­ mately 5,000 feet (1,500 m) it bent to rise straight up to approximately 9,000 feet (2,700 m). At that alti­ tude the smoke column reached a stratum of haze and scattered flat­ tened clouds.

Like the smoke columns from other control fires in this vicinity (fig. 1), the column ascended to the stratum and flattened out in all directions. Smoke from the differ­ ent columns then drifted in the

When this article was originally published, Gordon Powell was a management forester for the Alabama Division of Forestry. Figure 1—Smoke columns over control fires as seen from a distance of 21 miles (34 km] to the north. Column B is the smoke with which we came in direct contact. Columns A * The article is reprinted from Fire Control Notes 24(1) [Winter 1949]: 20–24. and C are from fires several miles from B.

Volume 63 • No. 3 • Summer 2003 71 A smooth ribbon of smoke, ash was blown into it as the funnel steady and lighter in color, moved off.) As this center 15 to 20 inches passed over a smoking limb ran from the ground to the “vapor cloud” or log, it caused a sudden vigorous over the smoke column. spewing of flame in 4 directions, with each arm of flame resembling ground to the “vapor cloud” over whirling smoke, small in diameter, that from a blowtorch, and the 4 the smoke column. We felt that reaching downward through the arms forming a cross quartering a better judgment would tell us that column of smoke, but not bending circle by right angles of flames. this was a light refraction from the quite so abruptly as the other sun, yet it closely resembled a tor­ smoke, so that the base of the fun­ I stepped into the funnel and it nado funnel as it stood motionless nel was outside the fire and in the proved to be quite intensive in amid the boiling smoke around it old burn near the east–west road wind velocity. In fact, it filled my (fig. 2). Feeling as foolish as boys on the north edge of the burn. It clothes with particles of ash and trying to find the end of the rain­ was there that we found it. made my ears “pop.” bow, we drove toward the ribbon of smoke. From any distance the funnel was At about 5:20 p.m. there was a sud­ whiter than the other smoke, but den rush of cool wind from the The “hot part” of the smoke col­ close observation revealed particles south, then from the northeast, umn was rising from the burning of burned litter whirling vigorously and then from the west; and the fuel with the wind to a point over in the funnel. Some of these parti­ funnel was gone. Then a steady, the edge of the old burn and at cles were released at higher eleva­ gentle breeze caused the smoke in that point rising straight up to the tions so that many burned particles the old burn to drift northward at vapor mushroom and flattening were floating down from the smoke first and then to drift back into the out under the vapor in a very flat and settling over a wide area. The column of billowing smoke still layer. Starting at the vapor mush­ funnel was only about 6 to 10 feet rising as before from the burning room there was a funnel of (1.8–3 m) in diameter and ap­ tops of slash. peared to be the same diameter at the top as at the base. The smoke About 30 minutes elapsed between in the old burn slowly drifted the time we first saw the whirlwind counter-clockwise around the fun­ and the time that it disappeared. nel which was also whirling count­ The map in figure 3 shows the er-clockwise. general path of the funnel base and its position with respect to the fire. Funnel Behavior A close look at the funnel showed Unusual Firewhirl that it was not stationary as it first While watching this whirlwind, I appeared, because the base wan­ realized it was not the usual dered slowly around in an area “firedevil” or “dustdevil” type often about 50 feet (15 m) in diameter seen in or near forest fires; never­ inside the old burn. As the funnel theless, I had no idea that others wandered slowly around, we fol­ had seen and reported similar lowed closely on foot, studying its whirlwinds. However, having read actions. other accounts (Graham 1952, 1957), this whirlwind seemed dif­ While the base moved through the ferent in that acreage of the associ­ burn, it “sucked up” all the ash in ated fire was small, the terrain is the center 15 to 20 inches (38–51 relatively flat, and the base of the cm), exposing mineral soil. funnel was in a cool part of the old (However, it left no mineral soil burn. Passage over “hot spots” Figure 2—Appearance of funnel as seen exposed in the path because more seemed purely accidental. It also from the west.

Fire Management Today 72 As the center of the funnel was 70 percent, and the wind was passed over a smoking limb or log, from the northwest at 28 miles per hour (45 km/h). For what it may it caused a sudden vigorous spewing be worth, weather maps show that of flame in four directions. on the morning before the fire, there was a 100-mile-per-hour (160-km/h) jet stream over the area at 35,000 feet (10,700 m).

On the 8–100 meter type fire dan­ ger rating station at nearby Lawrence Tower site, the buildup index was recorded as 34, the high­ est fuel moisture percent was 6.5, and the burning index was 12 in Fire Class 3. More Information Needed Fire-whirlwinds have frequently been observed in all parts of the country where fires occur, and appear to have important effects on forest fires. It is likely that these funnels have often been observed but not reported in Alabama; or perhaps we are too busy fighting Figure 3—Map of the fire area. fire to see the whirls that are pres­ ent. There is hardly enough infor­ appeared to be of a different type Environmental Factors mation available to fully substanti­ from the small firewhirls produced ate the theories of the cause of fire The weather, too, seems worthy of whirlwinds, yet such knowledge on a model scale (Byram and mention in association with the Martin 1962). might answer many questions con­ occurrence of the fire-whirlwind. cerning unusual fire behavior that According to information gathered unexpectedly becomes peculiarly Worth mentioning is that had this from a nearby fire danger rating firewhirl wandered into the flame it hazardous to men and equipment. station and information furnished I join other observers in hoping probably would have drawn up by the Southern Forest Fire flame and burning particles rather that more detailed observations Laboratory, there are several inter­ will be reported. than burned particles. Also, it esting area weather features. might have increased considerably in violence. References At ground level the relative humid­ Byram, G.M.; Martin, R.E. 1962. Fire- ity was 22 percent, and the wind whirlwinds in the laboratory. Fire Explaining its triggering and cause was from the east and northeast at Control Notes. 23(1): 13–17. would be strictly a guess with me. 4 miles per hour (6.4 km/h). At Graham, H.E. 1952. A fire-whirlwind of tornadic violence. Fire Control Notes. Perhaps the intense heat of burn­ 5,000 feet (1,500 m) the relative ing logging slash, perhaps the 13(2): 22–24. humidity was 55 percent, and the Graham, H.E. 1957. Fire-whirlwind forma­ adverse wind patterns, or perhaps wind was variable up to 10 miles tion as favored by topography and upper the cumulus cloud cap might be winds. Fire Control Notes. 18(1): per hour (16 km/h). At 10,000 feet 20–24. ■ the key. (3,000 m) the relative humidity

Volume 63 • No. 3 • Summer 2003 73 THE FOREST FIRES OF APRIL 1963 IN NEW JERSEY POINT THE WAY TO BETTER PROTECTION AND MANAGEMENT*

Wayne G. Banks and Silas Little

n the spring of 1963, conditions conducive to severe forest fires In the spring of 1963, record low rainfall I prevailed rather generally and humidity and high winds laid the foundation throughout the Northeastern for a blistering fire season in New Jersey. States. Scant rainfall, low humidi­ ty, and high winds combined to produce high and extreme fire-dan­ Pennsylvania, homes were threat­ occurring again. The second seems ger ratings for prolonged periods. ened by numerous woods fires. particularly important because on On April 20 fire danger reached a Fast-moving fires were reported in April 20, when most of the damage peak in several areas. As a result, Maryland. West Virginia, Virginia, occurred in southern New Jersey, fast-moving fires of unusual inten­ and Kentucky were also hard fire suppression techniques and sity burned out of control. pressed to control their many fires. pre-suppression measures proved Whether these newspaper state­ woefully inadequate. The New England States were for­ ments were correct in all details is tunate in escaping really large probably not very important. But Weather Conditions fires; the largest was approximately what should be important to April 1963 was the driest April on 700 acres (283 ha) in northern foresters and the general public are record in New Jersey. Only 0.31 Maine. However, New England did the reasons for these disasters, the inches (0.79 cm) of rain fell during have many small fires. Massachu­ ways of preventing them, and the the first 29 days of the month, and setts, for example, had 4,861 forest probability of similar conditions 0.52 inches (1.32 cm) on the 30th. fires in April, a record for that State and possibly for any State.

Fire disasters made the headlines of many newspapers. In New Jersey, newspapers reported more than 200,000 acres (80,000 ha) burned and 458 buildings destroyed (fig. 1). These reports listed 7 persons dead, many injured, and 2,500 evacuated—of whom 1,000 were left homeless. New York newspa­ pers reported that a brush fire on Staten Island covered 10 square miles (25.9 km2) and destroyed 100 homes. In the suburbs of Philadelphia, and elsewhere in

When this article was originally published, Wayne Banks and Silas Little were research foresters for the Northeastern Forest Experiment Station, Newtown Square, PA.

* The article is reprinted from Fire Control Notes 25(3) Figure 1—New Jersey home and vehicle destroyed by the April 1963 fires. [Summer 1964]: 3–6.

Fire Management Today 74 On 22 days, maximum wind veloci­ Foresters and the general public ties at Trenton were 20 to 40 miles should ask why the fires occurred, per hour (32–64 km/h). In the 30 how can they be prevented, and what is days prior to April 20, precipitation deficiency amounted to 3 inches the likelihood of similar weather conditions (7.6 cm). occurring in the future.

Relative humidity on that day tions that approached those of mally they are considered insuffi­ dropped from 50 percent at 6 a.m. April 20, 1963. In early May 1930, cient to maintain a fire. Owing to to 23 percent at 10 a.m. and when fires were rampant in South intensity, rapid spread, and ability remained between 20 and 23 per­ Jersey, surface burning conditions to carry across very light fuels, cent until nearly 5 p.m. Temper­ on two days approached those of suburban fires were difficult to ature was 80 °F (27 °C) at midday, April 20. The chief difference was suppress, and many buildings were dropping to 53 °F (12 °C) at mid­ that, in the 30 days preceding April lost. night. Fuel moisture indicator 20, 1963, there had been an inch sticks at two fire-danger stations (2.5 cm) less precipitation than in In the heavier and more flamma­ showed 3.5 and 4.1 percent fuel the 30 days preceding May 2 and 4, ble fuels of the New Jersey Pine moisture at 2 p.m. At both stations 1930. Region, the wind-driven fires the buildup index was 100 and the burned with great intensity and burning index was 200 on the 8–0 However, previous seasons have caused severe damage to both oaks meter. had conditions comparable to April and hard pines. Fires spread rapid­ 20 in both wind velocity and ly across upland sites where there The estimated average wind veloci­ drought. Since 1913 there have was relatively little fuel, as on ty for April 20 was 20 miles per been six fall days of apparently areas where prescribed burning hour (32 km). The average of the similar conditions, and one sum­ had been done 1 or 2 years earlier. maximum wind velocities reported mer day and four fall days when On such sites a very light cover of from the three nearest Weather conditions approached those of pine needles was sufficient to Bureau offices was 33 miles per April 20. However, because of less maintain a fire. Oak leaves, where hour (53 km), and gusts were prob­ wind within stands in summer and present, were blown across bare ably as high as 50 miles per hour early fall, the shorter days of fall, spots so that fires advanced rapidly (80 km). Turbulence prevailed at and less fresh leaf litter, we doubt even in scattered fuels. low levels, and many small whirl­ that any of these days actually pro­ winds developed. Prevailing wind vided burning conditions as critical Because fuels contained so little direction veered during the day as on April 20, 1963. moisture and winds were so from northwest to west, then back strong, the rate of spread of fires to northwest, and to almost north On several other days of that April on April 20 was extreme almost late at night. there were high winds. Fuel mois­ regardless of fuel type. One of the ture was low, and at one danger larger fires, which started just Comparable Conditions station the buildup index regis­ north of the Lebanon State Forest, in the Past tered 100 on 10 days. But at no advanced about 3-1/2 miles (5.6 Because April was such a black day time did all the elements of fire km) in 2 hours and 9 miles (14.5 for fire protection in New Jersey danger combine to create condi­ km) in 6 hours. Probably the rate and in sections of neighboring tions so severe as those on April of spread on April 20, 1963, has states, we attempted to determine 20, although April 29 was fairly been matched or even exceeded by the past frequency of such weather close. other fires for short periods. conditions. Weather Bureau However, foresters and wardens records for the previous 49 years Fire Behavior with many years’ experience in indicated that the spring fire Because of the drought, low fighting South Jersey fires could weather was never quite so bad as humidity, and high winds, some of recall no case where the sustained in 1963. During that half century the fires of April 1963 started and rate was as high as on April 20. only four spring days had condi­ spread in fuels so light that nor­ The Forest Fire Service of the New

Volume 63 • No. 3 • Summer 2003 75 This whirlwind seemed different In general, prescribed burning 1 or in that acreage of the associated fire more years before the wildfires of April 20 did not facilitate suppres­ was small, the terrain is relatively flat, sion appreciably, especially where and the base of the funnel was in oak litter was an important compo­ a cool part of the old burn. nent of the fuel complex. In these areas the 1963 fires were not stopped under fuel conditions that Jersey Department of Conservation accumulation of litter, and that had permitted the suppression of and Economic Development pro­ mostly pine needles. Tank trucks earlier wildfires. More recent burns vided data from 1924–63 that and handtools were useful in con­ that left some surface fuel remain­ showed only 1930 to be compara­ trolling spot fires in 1-year needle ing only reduced the damage, and ble to April 1963 in number of litter, but in oak leaf litter neither others that removed nearly all the large fires and their rate of spread. was enough. For example, at about fuel did not stop the fire. On one The two fires with the greatest area 1:45 p.m. between New Lisbon and firebreak where the 1962–63 win­ burned per hour were in 1963 and Route 70, a spot fire started along a ter burn had consumed only the covered about twice as much road within 50 feet (15 m) of a tank top litter, the fire burned with suf­ ground per hour as any of the truck, its crew, and several men ficient intensity to kill many of the much publicized 1930 fires for with handtools. At that particular oaks and severely scorch the which complete data are available. moment, a small whirlwind spread crowns of the pines (fig. 2). The data also emphasize the this fire for 100 to 200 feet (30–60 importance of April and May in m). High winds forced the aban­ Rapid combustion of wind-tossed local protection problems. donment of suppression attempts, dry fuels in the April 20 fires creat­ even though the area had only a ed extreme temperatures and Suppression year’s litter since the last pre­ greater damage on prescribed burn Difficulties scribed burn. areas than in other years. On areas The New Jersey Forest Fire Service with 1 year’s accumulation of litter uses a combination of suppression The extremely dry and windy con­ after periodic burns, head fires techniques and several kinds of ditions caused much difficulty in killed most of the oaks but not the machines. The latter include backfiring. Attempts to backfire overstory pines. Strong flank fires trucks of various sizes up to 500­ and hold the line along sand, grav­ on such areas killed about half of gallon (1,893-L) tank trucks el, and even blacktop roads had to the overstory oaks. Damage to oaks equipped with 4-wheel drive; air­ be done slowly and carefully to pre­ in areas with 2 or more years’ craft equipped to drop 150 to 200 vent the backfires from jumping accumulation of litter was usually gallons (568–757 L) of retardant; the road. Backfires along a State about as severe as in stands with and tractor and plow units. Back­ highway crossed the road even no previous prescribed burning. firing and handtools are also used, though the cleared strip in that However, any reduction of fuel was and backfiring plays a large part in area ranged from 75 to 120 feet apparently effective in reducing stopping head fires and tying in (23–37 m) wide. In some places damage to pines. flank fires. head fires arrived before backfiring could be completed. Preventing Similar On April 20 none of the suppres­ Disasters Effectiveness of sion methods proved effective. For Prevention. One of the major fires example, only one of the three Prescribed Burns of April 20–21 reportedly started pilots employed for firefighting Prescribed burning in the winter to where a debris burner had a permit was willing to fly, considering the facilitate suppression of fires in the for night burning. The fire held 40-mile-per-hour (64-km/h) winds South Jersey Pine Region has long over in a brush pile and broke out and the low-level turbulence. been advocated. However, this on April 20. At the nearest fire- Effectiveness of tractor and plow measure too proved less effective danger station the buildup index units on April 20 was confined to on April 20, 1963, than in previous had reached 59 on April 19, areas with no more than 1 year’s wildfires. climbed steadily to 100 on the

Fire Management Today 76 Figure 2—The stands on both sides of this road had been prescribe-burned in the winter of 1962–63. A good burn had been obtained on the left side, and here the fire of April 20, 1963, burned only a few scattered patches. On the right side, only the top litter had been burned by the winter fire, and much damage was done in April.

17th, dropped to 97 on the 18th, Camping should be prohibited at have brought a head fire, quite but was back at 100 on April 19th. remote sites when the buildup possibly before they could have index reaches 60, and at all areas been evacuated. We suggest that no burning per­ when the index is 80 or more. mits, for either day or night, be Prohibiting camping may meet Another important prevention issued when the buildup index is resistance; yet such a measure is activity is reduction of fuel 60 or more according to the system needed as much for the safety of through prescribed burning dur­ now in use in the Northeast. Any the campers as for fire prevention. ing the winter in types where permits issued when the buildup On April 20, 1963, a large group of these burns are silviculturally index is less than 60 should be so Boy Scouts were camped in the desirable. Earlier recommenda­ limited in time that they will expire Lebanon State Forest, where only a tions for the South Jersey Pine before the index reaches 60. slight shift in wind direction would Region appear to remain sound:

Volume 63 • No. 3 • Summer 2003 77 1. For maximum protection of Presuppression. What can be done Suppression. What can be done to improved property, burns at 1­ in presuppression to help ensure control fires under conditions such or 2-year intervals be used. initial-attack success under fire as prevailed in New Jersey and 2. For extensive forested proper­ conditions such as had developed other parts of the Coastal Plain in ties, barrier zones be prepared in April 1963? We suggest broaden­ April 1963? When the high winds by the prescribed burning of ing the scope of the working agree­ eliminated the small airplane as a belts of upland sites, which ments between the New Jersey working tool, suppression forces would reinforce swamps or other Forest Fire Service and other State found themselves back to conven­ natural firebreaks. agencies, companies, and individu­ tional weapons—tanker trucks, 3. Eventual development of a als to furnish equipment when it is plows, and hand crews—which checkerboard pattern on upland needed. Needed equipment from were inadequate. Perhaps larger sites in the larger unimproved outside sources should be on aircraft, carrying heavier loads and holdings, i.e., a pattern of young standby whenever the fuel and effective under windier conditions, unburned stands and of older, weather conditions indicate a con­ and larger tanker trucks with mul­ periodically burned stands. flagration threat. tiple pumping units might be feasi­ Prescribed burns at 4- or 5-year ble. Although the latter might not intervals are considered essential Protection agencies might also be adequate for such fires as in a protection program. consider providing tanks of 500­ occurred in April 1963, they should gallon (1,893-L) capacity or larger prove effective against many fires In years like 1963, only recently and equipped with their own that cannot now be attacked direct­ burned areas will be effective bar­ pumping units. These tanks could ly. They could also be a valuable aid riers against fire. But in view of the be strategically located, and stored to backfiring. rarity of such extreme fire danger, in such a way that they could be an annual and more costly fuel mounted on flat-bed or dump Also the use and coordination of reduction seems justified only near trucks and put into operation equipment could be improved. buildings or other improved prop­ quickly. Much difficulty was experienced in erty. holding backfires, even along wide The responsibilities of most forest cleared rights of way. Could tanker Management for pine over oak, fire protection agencies today units of the type available, support­ besides favoring the production of extend to much more than protect­ ed by large tank trucks for refilling timber, can facilitate fire control ing woodlands alone. The extension them, adequately fireproof the fuels under certain conditions. Periodic of residential building and industry on the opposite side of the roadway prescribed burning in areas with into rural wooded areas, the rever­ to permit rapid and safe backfiring few oaks results in less rapid com­ sion of farmland to forest, and the from such roadway? This type of bustion of the rather compact nee­ development of forest recreational operation might require planning dle litter. In April 1963 the burning areas are now making high-value and practice. It might very well needles were not carried long dis­ improvements and even lives resemble the “one-lick” method tances by the wind as oak leaves dependent on the efficiency of for­ used by hand crews, with several were. Suppression was therefore est fire suppression. Public recog­ tankers proceeding in tandem at a easier in stands that had few oaks. nition of these increasing responsi­ reasonably good speed, each one bilities must be encouraged if pro­ spraying a designated portion of tection agencies are to receive the the fuels. Studies to determine the financial support that they need. feasibility of this approach should be initiated. ■

Fire Management Today 78 THE HARROGATE FIRE—MARCH 15, 1964*

B.J. Graham

he Harrogate Fire started at about 2:30 p.m. in Section Where possible, the flanks were worked T 1938, Hundred Kanmantoo, on by fire crews, and the head fire was Mr. Brice’s property (fig. 1). The confined largely to a front of 90 chains. fire traveled mostly east through valuable grazing land and burned approximately 1,600 acres (650 ha). The exact cause of the fire is the flanks were worked by fire attacked the southern boundary not known, but after investigations crews, and the head fire was con­ and quickly obtained control. it is thought to have started from a fined largely to a front of 90 chains spark from the exhaust of a chain (5,940 feet [1,811 m]). Several miles of fencing were dam­ saw. aged, and approximately 1 mile By 3:05 p.m. the fire had spread for (1.6 km) of it must be replaced. Most of the area consisted largely approximately 70 chains (3,960 feet Some fences of sawn hardwood and of annual grassland; there was a [1,408 m]) east. By 3:12 p.m. it had iron droppers withstood the fire, scattering of Eucalypt trees. The spread for another 60 chains (3,960 but the wire components will last winter and spring preceding the feet [1,207 m]), but the flanks were for a much shorter period because summer of 1964 were very wet, being controlled and the head fire of the deterioration of the galvaniz­ and the summer was mild. continued east. ing content. Therefore, pasture fuels in this area were abundant and completely At 3:20 p.m. a pincer movement by Effective Suppression cured. The temperature reached units working on both flanks was The organization and method of 92 ºF (33 ºC) in the afternoon of becoming effective, and the head attack employed by fire controllers March 15, 1964, with a light wind fire, still moving east, was nar­ was extremely efficient and result­ blowing west to northwest. The rowed to a 70-chain (3,960-foot ed in the saving of large areas of winds were consistent; the approxi­ [1,408-m]) front. The fire had then heavily grassed land. Seventeen mate mean velocity was 10 miles traveled approximately 2 miles (3.2 Emergency Fire Service units, per hour (16 km/h). km). At about this time the capably supported by private units, Brukunga Unit was destroyed, and attacked the fire on the first day, Fire Behavior one man (the driver) was badly and a lesser number of conducted Commencing at Section 1938, burned. The farm buildings at patrolling and mopping-up opera­ Hundred Kanmantoo, the fire White Hut, which were in the fire’s tions on March 16. swept generally east at 2.5 miles path, were saved. per hour (4 km/h). The rate of The confining of this fire to spread of the head fire was affected At 3:30 p.m. the head fire hit the approximately 1,600 acres (650 ha) by such topographical features as Nairne–Harrogate road. It jumped is even more praiseworthy when creeks and ridges. These features over, but it was controlled after one considers the rocky nature of and the country’s rocky nature burning 50 acres (20 ha). The rest this inaccessible country and the restricted access. Where possible, of the front was controlled along fact that except for the Nairne– the road, and the fire was consid­ Harrogate road no natural or artifi­ ered under control. The time was cial barrier was near the fire. ■ When this article was originally published, 3:50 p.m. The fire units then B.J. Graham was a bushfire protection advisor in Australia.

* The article is reprinted from Fire Control Notes 27(1) [Winter 1966]: 7, 15. It was originally adapted from the South Australian Emergency Fire Services Manual, 1964.

Volume 63 • No. 3 • Summer 2003 79 Figure 1—Map showing progress of the Harrogate Fire.

Websites on Fire* Wildfire Mitigation Programs find by searching a State or program index or by using the advanced keyword option. Developed by This national database is a clearinghouse for infor­ Louisiana State University and the USDA Forest mation on State and local programs for mitigating Service’s Southern Research Station, the site uses the risk of wildland fire on private lands. Users can sources ranging from statutes and agency regula­ get information on more than 129 nonfederal poli­ tions to State, county, and local fire personnel. cies and programs. Information about regulatory “Featured programs” have included the Colorado programs, educational efforts, homeowner assis­ State Forest Service’s Wildfire Hazard Mitigation tance, and community recognition events is easy to Program, Texas Urban Wildland Interface Exhibits, * Occasionally, Fire Management Today briefly describes Websites brought to and Kittitas County Junior Firewise Educational our attention by the wildland fire community. Readers should not construe the description of these sites as in any way exhaustive or as an official endorsement Program. by the USDA Forest Service. To have a Website described, contact the managing editor, Hutch Brown, at USDA Forest Service, Office of Communication, Mail Stop 1111, 1400 Independence Avenue, SW, Washington, DC 20250-1111, Found at 202-205-1028 (tel.), 202-205-0885 (fax), [email protected] (e-mail).

Fire Management Today 80 THE FIRE BEHAVIOR TEAM IN ACTION: THE COYOTE FIRE, 1964*

John D. Dell

n a large wildfire, the fire boss bases much of his strategy on Santa Ana winds, surfacing at night, had swept O information provided by his an almost-controlled brush fire across firelines staff and other assistants. In 1964, into heavily populated residential areas. fire behavior teams furnished advice at several forest fires in southern California. Each team, tory was imminent. However, dur­ ready for the plans chief before directed by a fire behavior officer, ing the day shift the fire ceased to each shift so fireline overhead gathered and analyzed vital infor­ threaten the residential areas as could be thoroughly briefed. Also, mation on weather, fuels, and the Santa Ana winds returned aloft. any sudden deviations were topography. The team concept (see The sea breeze and upslope wind explained immediately to the fire the sidebar), originally described by caused the fire to spread into the boss, and a revised forecast was Countryman and Chandler (1963), mountains. made. proved an effective method for evaluating the behavior of fast- At the Coyote Fire camp, the fire Team members, meanwhile, began moving fires. This article briefly behavior officer was briefed by the taking pilot balloon observations of describes how a fire behavior team U.S. Weather Bureau’s fire weather winds aloft, making ground obser­ operated on the 67,000-acre forecaster on existing and antici­ vations, studying preattack maps, (27,000-ha) Coyote Fire that pated weather. The fire behavior and talking with fireline overhead burned on the Los Padres National officer made a behavior forecast in order to better understand fire Forest in September 1964. Coyote Fire Role of the Fire Behavior Team At 11 p.m. on September 22, 1964, a fire behavior team was dispatched Team observations and the fire behavior officer’s interpretation com­ from the Riverside Forest Fire bined with reports of scouts and line overhead are extensively used in Laboratory to the Coyote Fire at fire control planning. the request of the Los Padres National Forest. Santa Ana winds, An important function of the fire surfacing at night, had swept an behavior team is fire weather almost-controlled brush fire across observation. By working closely firelines into heavily populated res­ with the U.S. Weather Bureau’s idential areas. By early morning, fire weather forecaster, the team the fire, then out of control, had saves much time and avoids dupli­ burned more than 600 acres (240 cation of effort. Information pro­ ha). One of the most devastating vided by the fire weather forecast­ conflagrations in recent local his- er includes maps of the latest syn­ optic weather transmitted by fax from the Weather Bureau. When this article was originally published, John Dell was a fire research technician Team members may also make Fire behavior team takes a pilot balloon for the USDA Forest Service, Pacific observation with theodolite to determine Southwest Forest and Range Experiment upper air soundings of humidity patterns of local winds aloft. Photo: USDA Station, Berkeley, CA. and temperature or measure the Forest Service. winds aloft. * The article is reprinted from Fire Control Notes 27(1) [Winter 1966]: 8–10, 15.

Volume 63 • No. 3 • Summer 2003 81 conditions. Equipped with its own The fire behavior officer made a behavior forecast transportation, radio net, and ready for the plans chief before each shift so fire- instruments, the team was able to disperse to various locations on the line overhead could be thoroughly briefed. fire. A communications net linked the team with the fire behavior These winds continued most of the In the fire behavior forecast for officer, and pertinent information night. At 2 a.m. (September 24), September 24, prepared early that was sent regularly. Ground obser­ temperatures as high as 92 ºF morning, it was predicted that vations were made of temperature, (33 ºC) and humidities as low as strong Santa Ana conditions would humidity, windspeed, and wind 10 percent were reported at the continue at high levels, but would direction; type, density, and condi­ Weather Bureau’s fire weather weaken at low levels during the tion of fuels in the path of the fire; mobile unit at the fire camp. Fire day. It was reported that fuels were topography and aspect; current fire again swept through residential very dry and hot runs could be behavior; and trouble areas. The areas in the foothills. expected where favored by wind or fire behavior team made a thor­ topography. Atmospheric instabili­ ough surveillance of the area above By morning the fire had burned ty would favor the development of and to the flanks of the main front. along the entire Santa Barbara large convection columns, spot­ It noted that strong northeasterly front, over the ridge, and down ting, and firewhirls. Santa Ana winds were still aloft, although the into the Santa Ynez River drain­ winds were again likely to surface. layer had been rising since morn­ age. Personnel were added to the ing. The odds were against these fire behavior team to increase cov­ Fire Size Doubled winds surfacing as they had the erage of the growing conflagration. During the next few shifts the night before, but the forecast indi­ The fire behavior officer and part Coyote Fire nearly doubled in size. cated this might happen. of the team were requested to More than 3,000 firefighters from observe and advise on a critical all parts of the Western United Wider Fire Front backfiring operation on the west States joined in the battle. Zone Early on the evening of September side of the fire. Another team fire camps were set up at several 23, a team member was sent to the member made an upper air sound­ points around the fire in order to ridgetop above the fire to observe ing by helicopter, in order to deter­ place the manpower where it could its behavior and to look for any mine moisture in the lower atmos­ be shifted most efficiently. This, of signs of unusual changes. At 7 phere and atmospheric stability p.m., he noted that the humidity over the fire. had dropped to 14 percent. Light and variable winds gradually devel­ oped into a strong northeasterly blast that gusted up to 35 miles per hour (56 km/h). The fire began to intensify on the upper slopes. These factors indicated the fire would probably resume the same pattern as on the previous night— but on a wider front. The fire behavior team immediately report­ ed these observations to the fire behavior officer at fire camp. He, in turn, notified the fire boss so crews on the line would be warned. By 7:40 p.m., these winds were felt in Santa Barbara. The huge Coyote Fire near Santa Barbara, CA (September 22 to October 1, 1964), required the use of three fire behavior teams. The fire burned more than 67,000 acres (27,000 ha) of brush-covered watershed. Photo: USDA Forest Service.

Fire Management Today 82 course, created a difficult commu­ Another fire behavior team flew were taken during the 6-day period nications problem for the fire from the Northern Forest Fire the fire behavior teams were behavior team. Sending fire behav­ Laboratory at Missoula, MT. Two assigned to the Coyote Fire. ior forecasts to all these zone more radio-equipped vehicles and From helicopter, truck, jeep, and camps over already overloaded extra backpack radios were brought by walking the fireline, team communications systems was diffi­ in from Riverside. The fire behavior members observed and reported cult, yet very essential. Every effort officer reorganized and divided his conditions on the fire almost was made to bridge the communi­ enlarged team and located the continually. cations gap and to relay forecasts shifts at four points around the fire to all camps, since many of the line (fig. 1; also see the sidebar). Effective Team personnel from distant forests were Approach unfamiliar with some of the condi­ About 270 ground weather obser­ The Coyote Fire provided a good tions that affect the behavior of vations, 40 pilot balloon observa­ example of the effectiveness of the southern California fires. tions, and 7 upper air soundings team approach in fire behavior

Figure 1—As the fire spread north, the fire behavior officer divided and relocated his team for better coverage. Photo: USDA Forest Service.

Volume 63 • No. 3 • Summer 2003 83 coverage of conflagrations. With Fire Behavior Team Locations information flowing in from vari­ ous locations, the fire behavior On the 67,000-acre (27,000-ha) Coyote Fire, the fire behavior team worked from various locations to improve the information flow (fig. 1). officer has more time to appraise the situation, consult with the Potrero Seco Camp. The fire behavior officer and fire weather fore­ weather forecaster, and furnish a caster were headquartered here with the mobile weather unit. Team 1 more complete and accurate fire was assigned to take pilot balloon observations and upper air sound­ behavior forecast. With team ings, and to make ground observations and reconnaissance ahead of members in constant communica­ the fire. tion, he can be alerted to areas requiring his special attention. Los Prietos Camp. Team 2 was assigned ground observations and reconnaissance on the west and northwest sides of the fire. The fire weather forecaster, sup­ plied with frequent weather obser­ Pendola Camp. Team 3 was assigned ground observations, pilot bal­ vations taken by the team, benefits loon observations, and reconnaissance in the upper drainage of the by having more detailed local Santa Ynez River. information on which to base his forecasts. Polo Camp. One meteorologist was stationed here to take helicopter upper air soundings, ground observations, and reconnaissance on the The fire behavior team, as south side of the fire. employed in southern California Relay. Here a pickup truck equipped with radio relayed fire behavior during 1964, is certainly not the observations from the roving teams to the fire behavior officer. The final answer to fire control prob­ relay system, though makeshift, provided fairly good coverage of the lems. It is, however, the best fire, supplied more information on which to base forecasts, and made it approach yet developed for keep­ easier for the teams to circulate forecasts to the various camps. ing abreast of the behavior of a forest fire. Equipped with the instruments, tools, and trained personnel to do the job, fire behavior teams offer a service that—if properly used—can contribute to more effective fire control operations. Reference Countryman, C.M.; Chandler, C.C. 1963. The fire behavior team approach in fire control. Fire Control Notes. 24(3): 56–60. ■

Fire Management Today 84 “GLEASON COMPLEX” PUTS UP HUGE “PLUME”: A TRIBUTE TO PAUL GLEASON

Paul Keller

n May 3, 2003, a unique “wild­ fire” occurred above the spec­ What would you want your legacy to be? O tacular rock walls overlooking Paul Gleason: I suppose I would want my legacy the Crooked River in fire-prone to be that firefighters begin to realize the impor­ central Oregon. Although this inci­ dent never appeared in a situation tance of being a student of fire and that I was report, in the hearts of the wild- able to make that happen.* land fire community, it put up quite a plume. From wherever they came, they Hotshotting Shaped Almost 100 men and women—no were all from the same original His Core clan: brothers and sisters in the strangers to fire shovels and Any member of the national wild- family of fire—more specifically, pulaskis—passed through the crew land fire community—from Paul Gleason’s family of fire. That’s check-in area on that morning. ground-pounder to fire behavior why they had come from so far and Mostly “individual resources,” they analyst and on up the food chain of so wide to the memorial “Gleason hailed from all over the Pacific the Incident Command System— Complex.” Northwest, California, Idaho, who hasn’t heard of Paul Gleason Montana, Colorado—even as far has obviously been in a deep Rip Yes, the heat from this “fire” was away as Michigan, Massachusetts, Van Winkle coma. and Maine. All for a single-day fire the compassion for a fallen com­ rade. The torching was from the event, a different kind of planned Gleason’s wildland fire career bonding of pure emotions. The ignition. spanned almost 40 years—from his flareups were the outbursts of teenage days as lead brush hook on appreciation and gratitude, coupled Gleason’s Family of southern California’s Dalton with the pangs of so many bitter­ Fire Hotshots in the early 1960s, to his sweet memories. Their volunteer dispatch orders for time as an adjunct professor for this unforgettable day had been wildland fire science at Colorado All for Paul Gleason. delivered a few weeks in advance. State University, until the cancer When they got the word—despite burning within him bumped him all extenuating personal circum­ from us forever in February 2003. stances—they came. They included nurse practitioners, laborers, artists, city planners, teachers, What are the biggest ranchers, business owners, and a improvements you have gamut of folks working for the witnessed in the wild- USDA Forest Service and other agencies involved in wildland fire land fire service? management. Paul Gleason: I think that would have to be the increased focus on Paul Keller served under Paul Gleason as a firefighter safety. squad boss on the Zigzag Hotshot Crew from 1985 to 1989. He is now a contract Paul Gleason watching over his crew during writer/editor for the USDA Forest Service’s a prescribed fire. Photo: Tom Iraci, USDA Fire and Aviation Management Staff, Forest Service, Pacific Northwest Region, * From an interview with Paul Gleason in February Washington Office, Washington, DC. Portland, OR, 1988. 2003. For the full interview, see p. 91.

Volume 63 • No. 3 • Summer 2003 85 The Gleason Complex A Celebration of Paul’s life & The Fire Community of the Northwest

Please join the Zigzag Hotshots (1976-2003) for a celebration of Paul’s life and all of our connections in the fire community of the Northwest… Date: Saturday, May 3, 2003 Location: Roby’s Farm at Smith Rock, Oregon (5 miles north of Redmond, 3 miles east of Highway 97 at 9990 NE Smith Rock Loop, Terrebonne) 1300 Crew Check-In Drop off your gear, set up your tent, add your photos to the PLANS BOARD! 1500 Briefing at the “Barn” (A 5-minute walk from Roby’s Farm. If driving, instead of turning right to go to Roby’s Farm, continue past the gravel road 1/4 mile to the last house/barn on the left…signs will be in place) Slide show by Tom Iraci Guest Speakers Open Mike… share a story about Paul If you would like to speak or have slides to contribute, please coordinate w/Dennis Ghelfi at [email protected] or call (503) 622-3191 Ext 638 1700 Dinner please RSVP BY APRIL 20 to [email protected] or call (503) 622-3444 Burgers, Dogs, Garden Burgers provided (contributions accepted!) Bring a side dish to share Bring your own Beverages and lawn chairs (coolers/ice provided) 2000 Night Crew Briefing Tactics: Feed the bon fire and tell enough old fire stories to last your lifetime…it might get deep, wear your boots! Demob: Sunday AM campers please RSVP BY 4/20 to [email protected] Coffee provided Logistics: To get to Roby’s Farm: Go to Smith Rock State Park, turn right on Smith Rock Loop, a gravel road w/30 mailboxes and mobile home park just past the parking area. Proceed down the gravel road to the second house on the left… signs will be in place. Lodging: Lots of camping space available on site. Many motels and restaurants in Redmond, 6 miles south. See Ya There!

Along the way, he excelled as a dis­ But the core of Paul Gleason’s leg­ people who had battled fire along­ trict fire management officer and endary wildland fire expertise came side Paul on his hotshot crew. It forest-level fire ecologist for the from his two decades of hotshot was his third public memorial. In Forest Service. Then the National firefighting. Beginning in 1979, March, two others were held: in Park Service wooed this mathemat­ Paul devoted 12 years as superin­ southern California—near the site ics graduate over to their side of tendent of the Zigzag Interagency of the 1966 fatality Loop Fire, in the wildland fire equation. In the Hotshot Crew in the Forest which Paul lost 12 of his fellow late 1990s, Gleason retired his Service’s Pacific Northwest Region. firefighters—and west of Denver at Forest Service badge to become a the foot of the Rocky Mountain wildland fire specialist and deputy The Gleason Complex, appropriate­ Front Range. fire management officer for the ly based on a fire camp theme, was National Park Service’s Rocky for the most part a gathering of Mountain Region.

Fire Management Today 86 National Contributions Legend in the Fire Community Gleason is widely recognized for On Paul Gleason, from Michael Mexico. In the world of Western his various contributions to the Thoele’s book Fire Line— wildfire, only two or three hot- national wildland fire service. Chief Summer Battles of the West:* shot bosses were seen as his among these are: equal. … “At forty-four, though he would • Developing the LCES (lookouts, not have claimed it for himself, “He had marched through all the communications, escape routes, he ranked as one of the lesser fire ranks in the infantry of forest fire. and safety zones) concept that gods. A Pacific Northwest hotshot He had paid his dues and earned has become a foundation for crew boss, he was something of a his spurs. He was known as a man wildland firefighter safety; legend in his own time, a tough, who quoted Chinese philoso­ • Helping to pioneer the wildland aggressive, intellectual firefighter phers, read books on the art of firefighter professional tree who was the stuff of stories told warfare, and, in the off-seasons, falling program; and in fire camps from Alaska to New was a rock climber who took on • Relentlessly pursuing the devel­ * Golden, CO: Fulcrum Publishing, 1995. the big walls all over the West.” opment of improved fire behavior training.

At the 2001 National Hotshot Workshop—before anyone knew If you were to pick the most important character cancer had ignited deep inside trait for an effective leader, what would that be? him—Gleason was given the Golden Pulaski Lifetime Paul Gleason: It would be mindfulness. That abili­ Achievement Award for his ty to take in your surroundings and sort out the “longevity, leadership, versatility, important stuff, to be aware, to be vigilant. and hard work” in dedicating his life to wildland fire management.

“Paul lives and breathes fire,” ad students came into contact with child. The tragic impact of single­ explained Mike Hilbruner, the him, either directly or indirectly, handedly finding the six dead fire­ national applied fire ecologist for through fire behavior courses fighters on the Dude Fire no doubt the Forest Service, at the presenta­ S–290 to S–590. Paul literally convinced Gleason of LCES’s value. tion ceremony. “His years with the “wrote the book” for all or parts of But this promoter of firefighter hotshot program, together with his these national courses. safety had been driving around Forest Service and [National] Park with “LCES” on his pickup’s cus­ Service work in fire ecology, exhib­ A Gleason Complex attendee tom license plates for the prior 4 it an outstanding versatility.” remembered how much he enjoyed years. Gleason’s fire critiques—his class­ “Now,” Hilbruner continued, “he is room dissections of incidents han­ As a tribute to Paul’s penchant for expanding his diversity of experi­ dled by his crew after they were teaching—and learning—“The ence by becoming a university over—almost as much as the actu­ Paul Gleason Wildland Fire instructor—helping to prepare the al firefighting under Paul. Scholarship Fund” has been estab­ next generation of fire managers lished at Colorado State University and fire researchers.” There is no question that Paul’s to help support undergraduate stu­ aggressive, heads-up actions saved dents pursuing studies in wildland Gleason Forte: severely burned firefighter Geoffrey fire science. “This will help contin­ Teaching Hatch’s life on the 1990 fatality ue the work Paul was so dedicated Teaching was always one of this Dude Fire. A popular misconcep­ to doing,” explains his wife Karen expert firefighter’s prime fortes. In tion is that this experience planted Miranda Gleason. the last decade of Paul’s life, myri- the seed for Gleason’s LCES brain­

Volume 63 • No. 3 • Summer 2003 87 Supporting Women on What makes you want to follow someone? Fireline Paul Gleason: Confidence, knowing for certain Often overlooked in Paul’s lengthy that the person making the call has your safety wildland fire legacy is his promo­ foremost in their mind. And knowing that the job tion and support of women on the you are about to take on is the right thing to do, fireline. One of the first women in the Nation to penetrate the historic that it makes sense. all-male hotshot ranks—on the Zigzag Hotshot Crew in 1976— confided recently that she couldn’t Mount St. Helens, and Mount Trickster, Too have survived and endured without Rainier—all climbed by Gleason Paul’s underlying, often irreverent Paul Gleason’s encouragement and (he summited Rainier at age 16)— and mischievous humor—yes, he optimism. can be viewed. had a good dose of trickster in his soul, too—was also celebrated at A true teacher, he could instill in “Paul loves to challenge himself the Gleason Complex. But the others not only his wildland fire mentally and physically,” Mike Gleason traits recollected most by savvy, but also his self-assurance Hilbruner said at the Golden this part of his “fire family” were and wily determination to try to Pulaski award ceremony in 2001. his remarkable tenacity and grit. understand and overcome all “Those who have worked with Paul obstacles. have seen him fall back into cut­ Paul was diagnosed with colon and ting order [dig fireline with the liver cancer last spring. Next came Indeed, as early as 1983, the icono­ crew] whenever he could—without 6 months of intensive chemothera­ clastic Gleason had hired 7 women compromising his crew’s safety. As py. At the Gleason Complex, the on his 20-person hotshot crew. By superintendent, he didn’t have to people who helped take care of 1985, often to the bewilderment of do that. But it’s typical Gleason— Paul as his body and energy slowly other hotshot crews at the time, demonstrating an around-the­ atrophied—but never his positive his lead pulaski was—watch out, clock, assertive, hands-on ground- spirit—all marveled at this man’s fellas—a woman. pounder work ethic.” willpower and courage.

Gleason Trademark: Paul Gleason was never afraid of Strength death; he never gave up his hope to Gleason’s sometimes seemingly continue living. unbridled physical strength, prowess, and resilience were Tough Loss to Absorb another of this man’s vibrant trade­ To the incredulity and sorrow of marks. “In 1978, I remember see­ the national—even international— ing Gleason do 45 good pullups,” wildland fire community, Paul remarked a Gleason Complex Gleason died on the morning of attendee. In addition to being a February 27, 2003. He was 57 years luminary in the world of wildland old. fire, Paul was also a nationally rec­ ognized rock climber. He and his “I remain astonished at how vividly brother Phil claimed several first Paul lives on in our talk, his non­ ascents of western precipices. stop wisdom, and his sense for the beauties of more action and less A memorial stone for Paul is being talk,” recalls Dr. Karl E. Weick, placed near the entryway to the Rensis Likert Distinguished Japanese Gardens in Portland, OR. University Professor of Organiza­ It is a special vantage point from Paul Gleason reaching out for a fire tool to tional Behavior and Psychology at which the majestic, glaciated faces extinguish a hotspot inside a snag. the University of Michigan Business of Mount Hood, Mount Adams, Photo: John Gale, USDA Forest Service, School. Weick is internationally LaGrande Hotshots, LaGrande, OR, 1987.

Fire Management Today 88 renowned for his research on safety The Awesome Power of Fire and decisionmaking in many envi­ On December 12, 2002, Paul Paul has become more introspec­ ronments, including wildland fire. Gleason’s wife Karen Miranda tive now, spending much time in Gleason sent an e-mail to friends spiritual practices of Buddhism, “Paul Gleason made a big impact and family commenting on Paul’s which he finds comforting, focus­ on my life, despite a mere handful deteriorating condition: ing on the tenets of patience and of contacts,” said Dr. Weick. “I can compassion. Like a forest return­ only wonder in awe at the even Many of you have known Paul as ing over time after a burn and the greater impact he may have had on a supervisor, a teacher and train­ cycles of nature, so does Paul those who shared much longer, er, or just a great example of a believe now in rebirth of all living more intimate moments with him. positive person and firefighter beings. … I know that the whole [wildland who has lived life to the fullest in fire] community must find this a many interesting circumstances. In short, now, without much abil­ tough loss to absorb.” … But when people refer to Paul ity to fight the fire, he is becom­ as “a great role model,” “an amaz­ ing the wise old master sitting on Dr. Marty Alexander, a senior fire ing person,” even “a hero of sorts the rock in a purple robe after behavior research officer with the in the fire community,” he climbing to the top of the moun­ Canadian Forest Service in doesn’t understand this very well. tain, watching the awesome Edmonton, Alberta, ranks Gleason He doesn’t think he is special— power of fire burning the land­ with a handful of “historic fire fig­ just an ordinary guy who always scape, which he will become part ures” in the last century in North wanted to do a good job. … of one day. America and Australia. “Unfortun­ ately, Paul’s time was far too short,” Alexander laments. “But he definitely has had an effect on my “Why would such a legendary fire Professor Allgöwer recalls how the outlook on wildland fire.” person as Paul find time to come pony-tailed Gleason “fell in love” to a place where wildland fires defi­ with the Swiss National Park’s Paul Lives On nitely do not head the list of natu­ forests. A portion of this ral hazards?” asks Britta Allgöwer, Switzerland sojourn—in which In 2001, at age 55, under the senior research scientist and proj­ Paul’s wife Karen was able to join mandatory Federal firefighter ect leader in the Department of him—included pursuing his series retirement rules, Paul was Geography at the University of hands-on zeal for mountaineering. required to retire. Still ablaze with Zurich. “Maybe that’s just it,” “Paul loved to explore everything enthusiasm and desire to pursue answers Allgöwer, who hired around him,” confirms Allgöwer. his wildland fire vocation, he Gleason to come to her country to “And he also loved to explore his wasn’t about to downshift gears. So evaluate the fuel load and help in limits. Like trying to solo a 30­ he sprinted directly into his fire her quest to develop an ecosystem foot-high [9-m] 5.9 [difficult] sciences teaching post at the uni­ management strategy for the Swiss climb late in the day all by him­ versity level. He also hit the open National Park. “Paul was genuinely self—without being sure whether road as a wildland fire science con­ interested in everything that con­ or not he could ‘do the last couple sultant. One of his first stops was cerns fire. He literally and truly of moves’—as he confessed to me Switzerland. lived fire.” later!”

“To work with Paul was a revela­ What do you consider your strengths to be? tion,” says Allgöwer. “Although the Paul Gleason: [Pause.] Probably endurance. great wildland fire expert, he never superimposed his reasoning and And more specifically, enduring adversity and opinions to his surroundings here. using that experience to make something good Together, ideas and hypotheses come out of it. were formulated and passionately debated.”

Volume 63 • No. 3 • Summer 2003 89 The class and grace for which Paul was known shone through in his written words of farewell to his friends:

Paul Gleason rock climbing at one of his favorite spots near Mt. Hood, OR. Photo: Jeff Hunt, Breckenridge, CO, 1982.

“Paul lives on,” Allgöwer insists. “Whenever his name is mentioned, I always notice a warm glance in people’s eyes—even more so, now.” Class and Grace “Clearly, Paul was suffering beyond what any human is prepared to endure,” said friend Merrill Kaufmann, a research plant physi­ ologist for the Forest Service’s Rocky Mountain Research Station, on the severity of Paul’s aggressive, fatal cancer.

“Though, I can truly say that Paul lived through his recent illness with the same class and grace we have always known of him.” ■

Fire Management Today 90 INTERVIEW WITH PAUL GLEASON*

Jim Cook and Angela Tom

aul Gleason’s career as a fire­ fighter spanned parts of five During his career, Paul Gleason was front and P decades. He started as an 18­ center on three well-known fires of the modern year-old crew member on a south­ era—the Loop Fire in 1966, the Dude Fire in ern California hotshot crew and culminated his career as a profes­ 1990, and the Cerro Grande Fire in 2000. sor of wildland fire science at Colorado State University. superintendent in 1979. He role until he lost his battle with Paul grew up in southern remained in that role until 1992. cancer in 2003. California, the son of a traveling He then transferred to the Arapa­ evangelist preacher. He became an ho–Roosevelt National Forest as a During his career, Paul Gleason accomplished rock climber in his district fire management officer was front and center on three well- teens and continued to climb and eventually became the forest’s known fires of the modern era— throughout his entire life. In 1964, fire ecologist. the Loop Fire in 1966, the Dude he got his first job as a firefighter Fire in 1990, and the Cerro Grande on the Angeles National Forest, CA. Paul’s next move was to another Fire in 2000. His role on these He continued to work there on the fire agency in 1999 as the deputy three touchstone fires gave rise to Dalton Hotshot Crew through fire management officer for the his passion for firefighter safety 1970, with the exception of a 1­ Rocky Mountain Region of the and the “student-of-fire” philoso­ year stint in the U.S. Army. From National Park Service. In 2001, phy that he crusaded for. He was a 1971 to 1973, he went to college mandatory retirement at age 55 leader of firefighters and a leader and earned a degree in mathemat­ took Paul away from the Federal for the wildland fire service. ics. fire service and into academia. For the next 2 years, Paul was adjunct Paul’s contributions are far reach­ During this time, he also traveled professor for the wildland fire sci­ ing. He teamed up with D. Douglas and climbed extensively. He ence program at Colorado State Dent to pioneer the professional returned to work as a firefighter in University. He remained in this tree falling program for wildland 1974 as the assistant foreman for a 20-person regional reinforcement crew on the Okanogan National Forest, OR. Then, in 1977, he took the job as the assistant superin­ tendent of the Zigzag Interagency Hotshot Crew on the Mount Hood National Forest, OR, moving up to

Jim Cook is the training projects coordina­ tor for fire operations safety for the USDA Forest Service, National Interagency Fire Center, Boise, ID; and Angie Tom is a 10­ year seasonal wildland firefighter now serving on the Midewin Interagency Hotshot Crew based on the Midewin National Tallgrass Prairie, Wilmington, IL.

* Reprinted from the Website (http://fireleadersip.gov) of the Wildland Fire Leadership Team, chartered by the National Wildfire Coordinating Group’s Training Paul Gleason ensuring that his crew members safely cross the Illinois River on the Working Team. The interview was conducted in Denver, Siskiyou National Forest in Oregon during the 1987 Silver Fire. Gleason crossed last. CO, on February 26, 2003. Photo: John Gale, USDA Forest Service, LaGrande Hotshots, LaGrande, OR, 1987.

Volume 63 • No. 3 • Summer 2003 91 firefighters. He developed the Formal Publications by Paul “lookouts, communications, escape Gleason* routes, safety zones” concept that has become the modern foundation Finney, M.A.; Bartlette, R.; Colorado. Albuquerque, NM: of firefighter safety (Gleason 1992). Bradshaw, L.; Close, K.; USDA Forest Service, He was very involved in the devel­ Gleason, P. 2002. Report on Southwest Region: 77–82. opment of fire behavior training, fire behavior, fuel treatments, [see: http://www.fireleader­ with a focus on taking the scientif­ and fire suppression. In: ship.gov/toolbox/documents/l ic aspects of extreme fire behavior Graham, R.T., tech. ed. ces_gleason.html] and making them understandable Hayman Fire case study Gleason, P. 1994. Unprepared for concepts for every firefighter. He analysis. Interim draft. Fort the worst case scenario. reached outside the fire service and Collins, CO: USDA Forest Wildfire. 3(3): 23–26. collaborated with experts, such as Service, Rocky Mountain Gleason, P. 1994. Sun Tzu and Dr. Karl Weick, who were doing Research Station: 21–149. appropriate suppression research in the realm of decision- [See: .] 27–29. zations. Gleason, P. 1991. LCES—A key to Gleason, P.; Robinson, D. 1998. safety in the wildland fire After the fire a still small In the final tally, Paul was always a environment. Fire voice. In: Close, K.; Bartlette “student of fire”—a role model for Management Notes. 52(4): 9. (Hartford), R.A., eds. Fire others. To the very end of his life, [See also: Wildfire News & Management Under Fire he engaged in teaching and learn­ Notes. 5(2): 1.] (Adapting to Change), ing about fire. The opportunity to Gleason, P. 1994. Appendix G: Proceedings of the 1994 ask Paul the following questions Lookouts, communications, Interior West Fire Council about leadership came the day escape routes, safety zones— Meeting and Program; 1994 before he died and at his insis­ “LCES.” Orig. doc. June 1991. November 1-4; Coeur d’Alene, tence. In: Russell, W.L., Jr. An ID. Fairfield, WA: Analysis—Learning from our International Association of Cook/Tom: What makes you want mistakes: Lessons gleaned Wildland Fire: 191–199. to follow someone? from the 1994 shelter deploy­ Greenlee, J.M.; Thomas, D.; Gleason: Confidence, knowing for ments in the Southwest area Gleason, P. 1995. From cos­ certain that the person making the and the South Canyon Fire, mos to chaos. Wildfire. 4(1): call has your safety foremost in 7–11. * Compiled by the issue coordinators, Martin E. their mind. And knowing that the Alexander and David A. Thomas. job you are about to take on is the right thing to do, that it makes sense. Gleason teamed up with D. Douglas Dent to Cook/Tom: Who do you think is a pioneer the professional tree-falling program for leadership role model and why? Gleason: Chuck Hartley, who was wildland firefighters. the superintendent when I first went to work on the Dalton Cook/Tom: If you were to pick the tant stuff, to be aware, to be vigi­ Hotshots. Why … because he most important character trait for lant. Then take all that informa­ instilled that confidence. When I an effective leader, what would tion, put it together, and see if it worked for Chuck, I never doubted that be? makes sense to you. Another part for a minute that our safety was Gleason: That’s a hard one to of that mindfulness concept is the always the first thing in his mind. answer. [Pause.] But if there has to ability to relate to all types of peo­ Plus, Chuck ran the crew in a way be one, it would be mindfulness. ple and see what they can con­ that allowed us to have a sense of That ability to take in your sur­ tribute. confidence in ourselves and in our roundings and sort out the impor­ own capabilities as well.

Fire Management Today 92 Cook/Tom: Are leaders born or made … explain? Gleason: I think they are born, and I know we might disagree on this some. Certainly, many impor­ tant leadership skills can be devel­ oped, but I feel that trait of mind­ fulness is an innate capability that someone either has or doesn’t have.

Cook/Tom: Regarding leadership, what quote comes to mind? Gleason: “Those are my people. Wherever they go I must follow, for I am their leader.”

Cook/Tom: Thinking back to your youth, what influences helped you become a leader? Gleason: My father traveled for his work, frequently with the whole family. In our travels we camped out a lot and I think that was a big part of why I have always been drawn to the outdoors. Growing up Paul Gleason (left), with Merrill Kaufmann, a research forest ecologist for the USDA in southern California, I remember Forest Service, Rocky Mountain Research Station, standing on the site of the Snowtop many summers and falls where the Fire a day after the fire. On July 10, 1993, the fire burned to within 4-1/2 feet (1.4 m) of hills around the Los Angeles area Kaufmann’s cabin in Colorado. Gleason was at the cabin during the fire and helped pro­ would be on fire and especially the tect it. The cabin survived because Kaufmann had covered it with a metal roof, thinned the surrounding trees, and removed nearby fuels. Photo: Evelyn Kaufmann, 1993. memory of long strands of fire moving across the hills at night. I endurance, and more specifically Cook/Tom: Since you started in think I knew I wanted to be a fire­ enduring adversity and using that 1964, what are the biggest fighter by the time I was in 5th experience to make something improvements you have witnessed grade. Climbing was another big good come out of it. in the wildland fire service? influence on me. I remember one Gleason: I think that would have time my brother Phil and I did a Cook/Tom: How about your weak­ to be the increased focus on fire­ climb late in the fall and we nesses? fighter safety. weren’t able to finish the whole Gleason: I didn’t know I had any! climb before dark. So we had to [Laughs. Pause.] My biggest weak­ Cook/Tom: What do you consider spend the night up there suspend­ ness is patience, wanting to see the worst changes you have seen ed in our slings with just the stuff things happen too quickly or get in the wildland fire service? we had on us, which wasn’t much. changes in place right away. Not Gleason: Lack of aggressiveness Man, it got cold when the sun went having the patience to let things on the fireline. This might sound down, I think it was in November. develop. Sometimes I’m that way like a contradiction to the last with the people I work closely answer, but I don’t think so. What I Cook/Tom: What do you consider with. My expectations of their time mean here is that there seems to your strengths to be? and commitment could be unreal­ be a lot of indecisiveness on select­ Gleason: Let me think about that istic on occasion. ing a strategy and getting with it to for a minute. [Pause.] Probably make it happen.

Volume 63 • No. 3 • Summer 2003 93 Gleason was very involved in fire behavior training, was, what we did, why we did it. with a focus on making the scientific aspects of But I have to live with those deci­ sions because at that time it was extreme fire behavior understandable for every my responsibility. firefighter. Cook/Tom: Do you think a legacy is important and, if so, what do Cook/Tom: Describe a few of the began to find the others on the want your legacy to be? toughest decisions or dilemmas Perryville Crew … but that’s a Gleason: If you choose to lead oth­ you have faced. whole other story. ers you will have a legacy. But that Gleason: The first thing that legacy will be determined by those comes to mind is the Dude Fire … Cook/Tom: What helped to guide that follow you. [Pause.] I suppose and especially the decision to leave you in that decision? I would want my legacy to be that the subdivision and go down into Gleason: [Pause.] I don’t know firefighters begin to realize the Walk Moore Canyon. I passed a exactly why I did that, but it just importance of being a student of number of people coming up the seemed like the right thing to do. I fire and that I was able to help line in a hurry as I was going just knew I would have to live with make that happen. down. I talked to a couple of the my decision. It’s like the Cerro other hotshot superintendents. Grande thing. That was another Reference Everyone thought there was no tough situation. We made our deci­ Gleason, P. 1992. LCES and other way that anyone working in the sions in good faith and using our thoughts. National Wildfire Coordinating bottom of that canyon could make best judgment based on what we Group, Wildland Fire Leadership it out alive. Even so, a few of us knew. I remember how difficult it Committee, Website

leadership.gov/toolbox/documents/lces_g continued on down into the was to go to talk to the people in leason.html>. ■ canyon. We met Hatch and then Los Alamos and tell them who I � � \Ç à{x y|ÇtÄ àtÄÄç? ctâÄ ãtá tÄãtçá t ÂáàâwxÇà Éy y|Üxʉt ÜÉÄx ÅÉwxÄ yÉÜ Éà{xÜáA �

Fire Management Today 94 GUIDELINES FOR CONTRIBUTORS

Editorial Policy Paper Copy. Type or word-process the manu­ Clearly label all photos and illustrations (figure Fire Management Today (FMT) is an interna­ script on white paper (double-spaced) on one 1, 2, 3, etc.; photograph A, B, C, etc.). At the end tional quarterly magazine for the wildland fire side. Include the complete name(s), title(s), affil­ of the manuscript, include clear, thorough fig­ community. FMT welcomes unsolicited manu­ iation(s), and address(es) of the author(s), as ure and photo captions labeled in the same way scripts from readers on any subject related to well as telephone and fax numbers and e-mail as the corresponding material (figure 1, 2, 3; fire management. Because space is a considera­ information. If the same or a similar manuscript photograph A, B, C; etc.). Captions should make tion, long manuscripts might be abridged by the is being submitted elsewhere, include that infor­ photos and illustrations understandable without editor, subject to approval by the author; FMT mation also. Authors who are affiliated should reading the text. For photos, indicate the name does print short pieces of interest to readers. submit a camera-ready logo for their agency, and affiliation of the photographer and the year institution, or organization. the photo was taken. Submission Guidelines Submit manuscripts to either the general man­ Style. Authors are responsible for using wild- Electronic Files. See special mailing instruc­ ager or the managing editor at: land fire terminology that conforms to the latest tions above. Please label all disks carefully with standards set by the National Wildfire name(s) of file(s) and system(s) used. If the USDA Forest Service Coordinating Group under the National manuscript is word-processed, please submit a Attn: April J. Baily, F&AM Staff Interagency Incident Management System. FMT 3-1/2 inch, IBM-compatible disk together with Mail Stop 1107 uses the spelling, capitalization, hyphenation, the paper copy (see above) as an electronic file 1400 Independence Avenue, SW and other styles recommended in the United in one of these formats: WordPerfect 5.1 for Washington, DC 20250-1107 States Government Printing Office Style DOS; WordPerfect 7.0 or earlier for Windows 95; tel. 202-205-0891, fax 202-205-1272 Manual, as required by the U.S. Department of Microsoft Word 6.0 or earlier for Windows 95; e-mail: [email protected] Agriculture. Authors should use the U.S. system Rich Text format; or ASCII. Digital photos may of weight and measure, with equivalent values in be submitted but must be at least 300 dpi and USDA Forest Service the metric system. Try to keep titles concise and accompanied by a high-resolution (preferably Attn: Hutch Brown, Office of Communication descriptive; subheadings and bulleted material laser) printout for editorial review and quality Mail Stop 1111 are useful and help readability. As a general rule control during the printing process. Do not 1400 Independence Avenue, SW of clear writing, use the active voice (e.g., write, embed illustrations (such as maps, charts, and Washington, DC 20250-1111 “Fire managers know…” and not, “It is graphs) in the electronic file for the manuscript. tel. 202-205-1028, fax 202-205-0885 known…”). Provide spellouts for all abbrevia­ Instead, submit each illustration at 1,200 dpi in e-mail: [email protected] tions. Consult recent issues (on the World Wide a separate file using a standard interchange for­ Web at mat such as EPS, TIFF, or JPEG, accompanied Mailing Disks. Do not mail disks with elec­ ) for placement of the author’s name, title, For charts and graphs, include the data needed will be irradiated and the disks could be ren­ agency affiliation, and location, as well as for to reconstruct them. dered inoperable. Send electronic files by e-mail style of paragraph headings and references. or by courier service to: Release Authorization. Non-Federal Tables. Tables should be logical and under­ Government authors must sign a release to USDA Forest Service standable without reading the text. Include allow their work to be in the public domain and Attn: Hutch Brown, 2CEN Yates tables at the end of the manuscript. on the World Wide Web. In addition, all photos 201 14th Street, SW and illustrations require a written release by the Washington, DC 20024 Photos and Illustrations. Figures, illustrations, photographer or illustrator. The author, photo, overhead transparencies (originals are prefer­ and illustration release forms are available from If you have questions about a submission, please able), and clear photographs (color slides or General Manager April Baily. contact the managing editor, Hutch Brown. glossy color prints are preferable) are often essential to the understanding of articles.

Volume 63 • No. 3 • Summer 2003 95