Fluvial Sedimentology and Architecture of Two Latest Devonian Lower Huntley Mountain Formation Outcrops, North-Central Pennsylvania, USA

Home , Catskill Formation, Lock Haven Formation

Bucknell University Bucknell Digital Commons

Honors Theses Student Theses

Spring 2020

Evan W. Filion Bucknell University, [email protected]

Follow this and additional works at: https://digitalcommons.bucknell.edu/honors_theses

Part of the Geology Commons, Geomorphology Commons, Sedimentology Commons, and the Stratigraphy Commons

Recommended Citation Filion, Evan W., "Fluvial Sedimentology and Architecture of Two Latest Devonian Lower Huntley Mountain Formation Outcrops, North-Central Pennsylvania, USA" (2020). Honors Theses. 538. https://digitalcommons.bucknell.edu/honors_theses/538

This Honors Thesis is brought to you for free and open access by the Student Theses at Bucknell Digital Commons. It has been accepted for inclusion in Honors Theses by an authorized administrator of Bucknell Digital Commons. For more information, please contact [email protected].

iii

FLUVIAL SEDIMENTOLOGY AND ARCHITECTURE OF TWO LATEST

DEVONIAN LOWER HUNTLEY MOUNTAIN FORMATION OUTCROPS,

NORTH-CENTRAL PENNSYLVANIA, USA

Evan W. Filion

An Undergraduate Thesis

Presented to the Faculty of Bucknell University In Partial Fulfillment of the Requirements for the Degree of Bachelor of Science With Honors in Environmental Geosciences May, 2020

Approved:

Ellen P. Chamberlin, Ph. D. Thesis Advisor, Department of Geology and Environmental Geosciences

Jeffrey M. Trop, Ph. D. Thesis Advisor, Department of Geology and Environmental Geosciences

Christopher G. Daniel, Ph. D. Chair, Department of Geology and Environmental Geosciences

ACKNOWLEDGEMENTS

There are a lot of people who deserve mention in this acknowledgments section, not only for contributing to this scientific work, but also for making my experience as a geologist at Bucknell one of the truly unique and defining moments of my life. It has been an honor to be part of such a cohesive and encouraging academic department and a happy GeoPalz family! Great memories will live on in my heart and mind just as this thesis will live on in Carilee’s office and the online archive.

I want to begin by thanking Ellen Chamberlin and Jeff Trop, my two outstanding thesis advisors and mentors. You both challenged me to think deeper, write better, and grow as a person in ways I had not expected when signing up for this project. I am eternally grateful for your persistent support and steadfastness through my ups and downs with this project, and I am sure that I challenged your patience on more than one occasion. You have taught me so much and inspired a deep notion of scientific creativity.

Whether it was out climbing on the cliffs of US 15, chatting about the latest dataset or interpretation, or reading through my latest draft, I am so thankful for all the time and energy you invested into this project. I will never drive by another outcrop without thinking of the fun memories of walking along the highway with Ellen carrying the big

TLS case or driving up north and having Trop point out his favorite scour surfaces.

Next is a super big thank you and appreciation for Matylda Zaklicki. Matylda and

I have been in this project together since day one. Having a friend out there in Blossburg, at GSA, and back in the lab to share the funny moments and the load of tough deadlines has been one of the biggest joys of this whole project. Writing this thesis would not be

v the same without you, and I am so appreciative we are finishing this project together.

Thanks again for sharing your Blossburg Middle dataset, too!

I need to give an extra big shout-out to my main man Kyle Fouke for being the best friend a guy could ask for. You inspire me to always give my best and to be confident in who I am and what I believe. I would not trade (and certainly did not for sleep’s sake) a single minute of time we spent in the lab working on our theses side-by- side and having deep conversations, watching videos, playing ping pong, or winning

Geoguessr while not working on our theses. Thanks for helping me find a home in the

Geology Department, Kyle!

I also would like to thank Sam Jacob for his help out in the field measuring section and getting the GigaPan photos and lidar data for Blossburg West. It was a blast to spend a day off-roading, hiking, and collecting data with you and Professor Kirby in exchange for a few hours baking in the sun at the side of a highway! Sam, you also helped keep my morale up when you and I held the fort working late together in O’Leary

105 when everyone else would leave for the night.

I must acknowledge the superb service of the Bucknell University Transportation

Office. The excellent people in this office fulfilled over a dozen last-minute requests and played a crucial role in the efficiency and success of this project.

I am very grateful for the approval of my proposal and funding for this research from the Program for Undergraduate Research (PUR) at Bucknell University. I would also like to thank the Geologic Society of America Northeast Section for funding this research from the Stephen G. Pollock Undergraduate Student Research Grant.

I want to thank Dave Broussard for his collaboration on the project and for generously sharing a palynology report for samples of the Blossburg outcrops, prepared by Dr. Pierre Zippi of Biostratigraphy.com, LLC. I must extend a big thank you to Lily

Pfeiffer at the University of Oklahoma for running our samples through a particle size analyzer and compiling the grain sized data used in this project. I would also like to acknowledge Spectrum Petrographics for the preparation of excellent thin sections that, unfortunately, were not analyzed in this project. I must thank the Bucknell Department of

Geology and Environmental Geosciences for providing all the field equipment, software, and transportation funding necessary for this project. Thank you, Brad Jordan and Carilee

Dill, for helping out in a pinch with whatever I may have needed during the summer and throughout the school year! I would like to also acknowledge Michelle Beiler for her unwavering support and flexibility for my pursuit to finish this thesis.

My parents certainly deserve recognition for putting on their geologist hats and giving me super useful feedback on my grant proposals and thesis drafts. Thank you,

Mom, for letting me take over the home-office (filling it with rocks and fluvial sedimentology papers) just to finish my thesis after spring break! I would also, somewhat controversially, like to acknowledge COVID-19 for throwing in a pretty big curveball during my senior year and contributing, in a way, to the final development of this thesis.

Lastly, I must acknowledge the people who in small ways made this project special for me. From roommates to Terrace Room groupies to the goofballs in O’Leary and the Sustainability House, you know who you are. Putting together this thesis would not have been nearly as much fun without you! :)

vii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... iv

TABLE OF CONTENTS ...... vii

LIST OF TABLES ...... x

LIST OF FIGURES ...... xi

ABSTRACT ...... 1

INTRODUCTION ...... 3

Fluvial Sedimentology and Architecture Element Analysis ...... 4

Thesis Goals and Approach ...... 8

GEOLOGIC CONTEXT ...... 10

SEDIMENTOLOGY ...... 15

Methods ...... 16

Facies Observations and Interpretations...... 16

Channel Facies ...... 21

Proximal Floodplain Facies ...... 26

Distal Floodplain Facies ...... 29

Stratigraphic Changes from Blossburg South to Blossburg West ...... 31

viii

Channel Facies ...... 31

Proximal Floodplain Facies ...... 32

Distal Floodplain Facies ...... 32

FACIES ARCHITECTURE ...... 34

Facies Architecture Mapping Methods ...... 34

Terrestrial Laser Scanning (TLS) and Point Cloud Registration ...... 35

Outcrop Panoramic Photography ...... 38

Mapping Bar and Scour Surfaces ...... 39

Summary of the Architecture Maps for Two Blossburg Outcrops ...... 45

Facies Area Proportions ...... 49

Methods of Measuring Area Proportions ...... 49

Area Proportion Results ...... 55

PALEOCHANNEL DEPTH ...... 59

Methods ...... 61

Results and Interpretations ...... 61

GRAIN SIZE ANALYSIS ...... 64

Methods ...... 64

Results and Interpretations ...... 65

Stratigraphic Changes ...... 66

Sample Exclusion and Error Analysis ...... 69

PALEOSLOPE ...... 71

Methods ...... 71

Results, Interpretations, and Error Analysis ...... 74

DISCUSSION ...... 78

River Plan Form and Paleoenvironment ...... 78

Foreland Basin Variables Influencing Stratigraphic Changes ...... 81

Progradation of a Distributive Fluvial System ...... 83

Tectonic Uplift and Basin Subsidence ...... 88

Global Cooling, Glaciation, and Sea Level Drop ...... 89

Interpretation of Stratigraphic Changes and Future Directions ...... 93

CONCLUSIONS ...... 97

REFERENCES ...... 100

APPENDICES ...... 115

Appendix A: Comparison to Catskill Magnafacies in New York ...... 116

Appendix B: Palynology of the Blossburg Outcrops ...... 117

Appendix C: Supplementary Data Tables ...... 119

LIST OF TABLES

Table 1: Summary of lithofacies observations and interpretations ...... 19

Table 2: Summary of facies and associated lithofacies and environments ...... 21

Table 3: Lithofacies and facies proportions of Blossburg stratigraphic columns . 22

Table 4: Dimensional measurements of sand bodies and internal stories ...... 48

Table 5: Areal and linear facies proportions for Blossburg outcrops...... 56

Table 6: Summary of sedimentological and architectural outcrop characteristics 79

Table C1: Facies area measurements and calculations for Blossburg South...... 119

Table C2: Facies area measurements and calculations for Blossburg West...... 120

Table C3: Method 1 and 2 calculations for estimating paleoslope ...... 121

Table C4: Raw percentage output of sandstone grain size analysis ...... 124

Table C5: Cumulative percentage output of sandstone grain size analysis ...... 127

LIST OF FIGURES

Figure 1: Late Devonian paleogeographic map of Pennsylvania ...... 4

Figure 2: Block diagrams of meandering river deposits and facies architecture .... 6

Figure 3: Regional map showing Blossburg and other Late Devonian outcrops .. 12

Figure 4: Site map of Blossburg outcrops along US 15 and bedrock geology ..... 14

Figure 5: Stratigraphic column of Blossburg outcrops ...... 187

Figure 6: Representative photos of channel facies ...... 24

Figure 7: Representative photos of proximal floodplain facies ...... 27

Figure 8: Representative photos of distal floodplain facies ...... 30

Figure 9: Image of the TLS module used in this research ...... 36

Figure 10: Site maps of TLS and GigaPan positions ...... 37

Figure 11: Image of the GigaPan EPIC Pro V used in this research ...... 39

Figure 12: Example facies architecture map from Blossburg South...... 41

Figure 13: Example facies architecture map from Blossburg West ...... 42

Figure 14: Examples of applying mapping criteria to channel-belt-scale scours . 44

Figure 15: Blossburg South panorama and facies architecture map ...... 46

Figure 16: Blossburg West panorama and facies architecture map ...... 47

Figure 17: Blossburg South digital outcrop model in Faro® Scene ...... 50

Figure 18: Blossburg West digital outcrop model in Faro® Scene ...... 51

Figure 19: Visualization of the grid used in area proportion calculations ...... 52

xii

Figure 20: Rectangular area measurement grid overlaid on Blossburg South ...... 53

Figure 21: Trapezoidal area measurement grid overlaid on Blossburg West ...... 54

Figure 22: Facies area proportion box and whisker plots for Blossburg South .... 57

Figure 23: Facies area proportion box and whisker plots for Blossburg West ..... 57

Figure 24: Methodology for measuring bar heights ...... 60

Figure 25: Scatter plot of bar height measurements from Blossburg outcrops ..... 62

Figure 26: Grain size distributions (PDF and CDF) for sandstone samples ...... 67

Figure 27: Scatter plot of grain size medians (D50) for sandstone samples ...... 68

Figure 28: Scatter plot of paleoslope estimates for Blossburg paleorivers ...... 75

Figure 29: Prograding distributive fluvial system model ...... 84

Figure 30: Correlated stratigraphy for the Hangenberg global cooling interval ... 92

Figure B1: Composite palynozonation chart for late Famennian time ...... 118

ABSTRACT

Thick successions of river deposits accumulated in the north-central Pennsylvania region of the Appalachian foreland basin during Late Devonian time (~380-360 Ma). The properties and morphologies of these paleorivers are not well characterized. Latest

Devonian tectonic, climatic, and eustatic controls on river dynamics and basin infilling also remain unclear. This study assesses the sedimentology, facies architecture, paleochannel depths, and grain size of a 133 m thick section of fluvial strata exposed across two outcrops, Blossburg South (older) and Blossburg West (younger), mapped as lower Huntley Mountain Formation near Blossburg, Pennsylvania. Field-based lithofacies observations, high-resolution panoramic photography, terrestrial lidar scanning, and laser particle size analysis were used to build a stratigraphic column, map fluvial architecture, and estimate average paleoriver gradient.

Channel facies primarily consist of cross- and horizontally-stratified, fine-grained single- to multistory sand bodies with scours and lateral accretion surfaces. Proximal floodplain facies consist of gray thinly-bedded crevasse splay sandstones and massive levee siltstones and are more abundant upsection. Distal floodplain facies chiefly consist of red relict-bedded and fissile mudstone paleosols (with rootlets, slickensides, caliche nodules) and are more abundant downsection. Paleochannel depths (0.9-2.7 m) and bar deposit grain sizes (median diameters 142-221 μm) increase upsection. Estimated average landscape/river paleoslope for the latest Devonian alluvial plain near the studied

Blossburg locality (0.54 x10-4 – 5.2 x10-4) does not exhibit any consistent change over the

2 studied interval. The two outcrops are interpreted as a succession of suspended- to mixed- load meandering rivers within mobile channel-belts due to the predominance of lateral accretion surfaces, low median grain size (fine-grained sand range), and low paleoslope values (<10-3).

Several variables are explored to explain the observed stratigraphic changes. An upsection increase in channel facies proportion, channel depth, and grain size could be generated by a prograding distributive fluvial system, which has already been proposed to explain Late Devonian alluvial plain sedimentation. Evidence for rapid uplift, high sediment supply, and decreasing basin subsidence during Famennian time suggests a decrease in the ratio of accommodation-creation to sedimentation, which is consistent with system progradation. Biostratigraphic data interpret the age of Blossburg West strata to within the LN biozone (361-359 Ma), which is contemporaneous with the Hangenberg global cooling interval and an advance of lowland glaciers in the Appalachian foreland basin. Pro-glacial outwash and increased precipitation could potentially have influenced the deepening and sediment coarsening of Blossburg paleorivers, which subsequently would lead to an increasing channel facies proportion and amalgamation. However, improved relative age correlations between glacigenic strata in eastern Pennsylvania and fluvial strata in north-central Pennsylvania are necessary to confirm glacial influences on landscape/channel processes at the Blossburg locality.

INTRODUCTION

Foreland basin systems consist of marine and terrestrial depositional environments on continental crust in the foreland of fold-thrust belts (DeCelles, 2011).

Their deposits are among the largest accumulations of sediment on Earth and provide long-term records of the evolution of topography, environments, ecologies, and climate.

Sediments within the Appalachian foreland basin in eastern North America record collisional plate convergence and a series of mountain building events, or orogenies, during Ordovician-Permian time. During Devonian time, the Acadian orogeny was responsible for the accumulation of >3 km thick successions of siliciclastic sediment within the foreland basin, which was west of an active fold-thrust belt shaped by crustal shortening (Murphy and Keppie, 2005; Bradley and O’Sullivan, 2017) and glaciation

(Brezinski et al., 2010). Middle and Late Devonian river (fluvial) deposits comprise much of the basin fill (Sevon, 1985); these strata thin and fine westward and exhibit west- directed paleocurrent indicators (Fig. 1; Harper, 1999). These fluvial deposits host diverse fossil assemblages of terrestrial and aquatic organisms. The fossils and their host sediments record the environmental setting for the fin-to-limb transition that characterized the evolution of stem tetrapods from fish (Daeschler et al., 2009; Broussard et al., 2018). However, quantitative analysis of the fluvial deposit architecture, paleochannel depths, and paleoslope of latest Devonian foreland basin rivers is generally lacking, especially in north-central Pennsylvania.

Figure 1: Late Devonian depositional setting of the Appalachian foreland basin during the Acadian orogeny. Arrows depict the general westward direction of sediment transport, interpreted from lithofacies and paleocurrent trends (Harper, 1999). Map inferred to represent paleogeography >10 Ma prior to deposition of Blossburg strata in North-Central Pennsylvania. © 2013 Colorado Plateau Geosystems Inc.

Fluvial Sedimentology and Architecture Element Analysis

Fluvial deposits in north-central Pennsylvania contain information about the conditions for fluvial landscape processes, climate change, and basin infilling within the

Appalachian foreland basin during Late Devonian to Mississippian time. To access this information, both quantitative and qualitative sedimentological and architectural observations/reconstructions of cross-section exposures must be made. Outcrops provide a relatively rare and valuable dataset for this analysis. Fluvial sedimentologists have long

5 tried to elucidate the characteristics of fluvial facies (i.e., river deposits representative of specific depositional processes and locations with respect to the channel) and river morphology based on sedimentologic data (e.g., Leopold and Wolman, 1957; Schumm,

1960; Allen, 1965). The accuracy of this methodology is enhanced when emphasis is given to identifying patterns and features in cross-section views of fluvial architecture, including the spatial distribution of channel and floodplain facies and the geometry of each facies element expressed in outcrop (see Fig. 2 for a preview of characterizing fluvial deposit heterogeneity) (e.g., Allen, 1983; Friend, 1983). Many studies have used architectural element analysis, in which these patterns and features have been classified and attributed to specific elements and conditions within river systems (e.g., Friend,

1983; Miall, 1985; Bridge, 2006). The development of standard methods for mapping and quantifying fluvial architecture elements has enabled researchers to collect dimensional measurements of fluvial features (i.e. unit bars, channel fills, channel-belt systems, and scours) and to compare model stratigraphy to outcrop data (e.g., Bridge and Leeder,

1979; Miall, 1985; Heller and Paola, 1996; Mohrig et al., 2000; Holbrook, 2001). Many researchers have used architectural analyses to gain deeper insights into river planform and paleoenvironment reconstruction (e.g., Miall, 1985), the pattern and style of channel movements (e.g., Mohrig et al., 2000; Hajek and Edmonds, 2014; Chamberlin and Hajek,

2015, Chamberlin et al., 2016), accommodation-creation to sedimentation ratio (e.g.,

Bridge and Leeder, 1979; Heller and Paola, 1996; Colombera et al., 2015; Chamberlin and Hajek; 2019), and foreland basin tectonics and sediment transport (e.g., Gordon and

Bridge, 1987; Bridge, 2000).

story story -

scours at these these at scours

second bar bar second

ure displaying ure single both a

of deposition near an active channel: splay, splay, channel: active an near deposition of

channel architect channel

story sand bodies (SSBs) are detached from other from detached other (SSBs) bodies sand story are

ied from Chamberlin and Hajek, 2015. Chamberlin Hajek, and from ied

sand bodies (MSBs) have amalgamated channel deposits channel amalgamated and (MSBs) bodies have sand

channel scour. (Right) Multi scour. channel

story story

y y intra an

scale scour surfaces, representing significant avulsions, are depicted distal overriding and floodplain avulsions, representing are surfaces, significant scalescour

d architecture of meandering river deposits. (Left) (Left) deposits. meandering of Zones d river architecture

belt

ontinuous under multiple channel fills. multiple ontinuouschannel Modif under

Anatomy an Anatomy

story dominant character result a avulsions. of as character Single dominant story

: :

leochannel bar deposits, multi bar whereas leochannel

Figure Figure base raises deposit to up Previously a package channel the and floodplain, bar. bar deposited channel levee, separated b deposits are package; multi and pa Channel contacts. c remaining

An architectural element analysis provides detailed characterization of fluvial facies for an outcrop. Fluvial deposits can be divided into three primary facies groups: channel facies, proximal floodplain facies, and distal floodplain facies. Channel facies consist of primarily sand and gravel deposited within the margins of a channel. The gradual lateral migration of a river channel may produce laterally-accreted point bar deposits; these bars are commonly the basic structural elements of channel fills in fluvial architecture analysis (see left block diagram in Fig. 2). Proximal floodplains, the overbank regions directly adjacent to the main channel, are frequently inundated and contain levee and crevasse splay sediments (Fig. 2) (Allen, 1965; Friend, 1983). Distal floodplains are beyond the proximal floodplain and thus are infrequently inundated and can consist of fine-grained paleosols and pond/lake sediments (Allen, 1965).

As river channels within a foreland basin aggrade through time, they often migrate back over old channel fills or relocate rapidly to a new location on the floodplain in a process called avulsion (Mohrig et al., 2000; Slingerland and Smith, 2004). The left block diagram in Figure 2 depicts a channel that has aggraded and migrated within its margins; an intra-channel-belt scour surface separates active bar deposits from older underlying bar deposits. Channel-belt avulsions, however, physically separate a river’s active depositional location from recently abandoned channel deposits. This process can either produce single-story sand bodies (SSBs) or multistory sand bodies (MSBs) (refer to right block diagram in Fig. 2). An SSB forms when a channel avulses and re- establishes on top of fine-grained floodplain sediments, spatially isolating the active channel fill from previous channel fill deposits. An MSB forms when a channel avulses

8 and reoccupies an ancient channel fill, directly juxtaposing older and younger channel deposits; these channel-body sandstones are separated by a channel-belt-scale scour surface, representing a much larger temporal and spatial scale than that of an intra- channel-belt scour surface (Chamberlin and Hajek, 2015).

Architecture element analysis provides a complete map of facies elements and bounding surfaces in an outcrop. These surfaces, often expressed as linear features in outcrop, can be classified according to a hierarchy of erosional scours and accretionary surfaces within channels and/or channel belts (i.e., bar surfaces, intra-channel-belt scours, channel-belt-scale scours, etc.) (e.g., Allen, 1983; Friend, 1983; Miall, 1985; Holbrook,

2001; Van de Lageweg et al., 2016). This study serves as one of the first applications of fluvial architecture mapping to latest Devonian river deposits exposed in outcrop in north-central Pennsylvania. The mapping of bar surfaces and channel-belt-scale scours in these outcrops allows for quantitatively reconstructing channel body stacking, flow depth, and landscape slope for latest Devonian river systems in the northern Appalachian foreland basin.

Thesis Goals and Approach

The main goal of this thesis is twofold: 1) to characterize the sedimentology and facies architecture at two previously unstudied outcrops spanning a 133 m thick succession of the lower Huntley Mountain Formation near Blossburg, Pennsylvania; and

2) to gain preliminary insights on latest Devonian river systems, basin infilling, and environmental change in the northern Appalachian foreland basin. These outcrops, named

Blossburg South (older strata) and Blossburg West (younger strata), are part of a series of four Late Devonian outcrops exposed along U.S. Route 15 near Blossburg, Pennsylvania.

The first objective involves a combination of fieldwork, to collect sedimentology, imagery, point-cloud, and architecture mapping data; and computer work, to analyze the fluvial facies of each outcrop and to construct facies architecture maps. The second objective involves two discrete components: 1) using grain-size and paleochannel depth measurements to reconstruct river paleoslope (e.g., Lynds et al., 2014); and 2) interpreting facies architecture and dimensions through studies regarding river planform, landscape/channel processes, basin subsidence and tectonics, accommodation-creation and sedimentation rates, distributary fluvial system progradation, and climate change

(e.g., Leopold and Wolman, 1957; Bridge and Leeder, 1979; Faill, 1985; Miall, 1985;

Heller and Paola, 1996; Brezinski et al., 2010; Weissmann et al., 2013; Colombera et al.,

2015; Oest, 2015; Gardner, 2019). Data collected from the two Blossburg outcrops allow us to assess the evolution of these river systems over an extended interval of time. This assessment includes results from a companion study by Zaklicki (2020), which collected similar data from stratigraphically older strata exposed at the nearby Blossburg Middle outcrop. Detailed fluvial architecture studies and paleoslope reconstructions are generally lacking for Devonian sediments in the northern Appalachian foreland basin; notable exceptions include work from Gordon and Bridge (1987), Willis and Bridge (1988), and

Bridge (2000). See Appendix A for a comparative analysis of Gordon and Bridge (1987) data. Thus, this thesis provides an important contribution towards understanding river system and foreland basin evolution in Pennsylvania during Late Devonian time.

GEOLOGIC CONTEXT

The Appalachian foreland basin, extending over 2000 km along the eastern margin of North America, is shaped by a sequence of tectonic collisions from the

Grenville orogeny (~1.3 Ga) to the Allegheny orogeny (~265 Ma) (Ettensohn, 2008).

Beginning in Early to Middle Devonian time, the Acadian orogeny, representative of a transpressional collision between Avalonian terranes and the North American craton

(Laurussia), is responsible for the growth and westward progradation of the Acadian mountain front (Ettensohn, 2009). Models suggest orogenic deformation began in the north at the St. Lawrence promontory (near the current-day Bay of Fundy) and propagated southward along a convergent boundary as the terranes continued to accrete westward (Ettensohn, 1985; Faill, 1985). This orogeny progressed through Middle and

Late Devonian time (416-360 Ma) and likely ended with a final relaxational phase spanning the Devonian-Mississippian boundary (360-350 Ma), decreasing overall basin subsidence (Ettensohn 2008, 2009).

Uplifted bedrock was eroded from the Acadian mountains and transported westward in a series of rivers draining to the mid-continental epeiric sea (Fig. 1) (Sevon,

1985). These sediments accumulated in thick clastic wedges, now preserved in the Valley and Ridge and Appalachian Plateau regions of the eastern United States. Basin subsidence reached a maximum during Late Devonian time, accommodating up to

3500 m of clastic sediment in eastern Pennsylvania and Maryland (Faill, 1985;

Slingerland, 2009).

The Middle to Late Devonian Catskill clastic wedge is characterized by red sandstones and mudstones deposited in marine shelf, delta, and alluvial plain environments (Dennison, 1985; Sevon, 1985). These sediments can be viewed in outcrop throughout southwestern New York and northern Pennsylvania (Fig. 3) (Ver Straeten,

2013). In north-central Pennsylvania, the Lock Haven Formation and lower Catskill

Formation encompass the marine interval, and the middle to upper Catskill Formation encompasses fluvio-deltaic sedimentation (Harper, 1999; Slingerland et al., 2009).

Near the end of the Devonian and into the Mississippian, Catskill Formation red beds transition to the predominant gray to buff sandstones of the Pocono wedge, traditionally interpreted as meandering to braided river sediments on the alluvial plain

(Sheckler, 1986). The terrestrial sedimentary interval between these two wedges, spanning the Devonian-Mississippian boundary, is variable throughout different regions of Pennsylvania. In southern and eastern Pennsylvania, the Rockwell and Spechty Kopf

Formations are both characterized by gray sandstones and mudstones with a diamictite base (Berg, 1999; Brezinski et al., 2010).

In north-central Pennsylvania, the interval between distinct clastic wedges is recorded in the dominantly non-marine Huntley Mountain Formation. This formation has been characterized by Berg and Edmunds (1979) as primarily consisting of greenish-gray sandstones with minor red to gray mudstones, representative of both river channel and floodplain paleoenvironments. The depositional contact between the Catskill and Huntley

Mountain Formations is gradational and defined by the first appearance of gray-green sand bodies together with an upsection increase in sandstone proportion (Colton, 1968;

Figure 3: (A) Map showing Upper Devonian sedimentary strata (pale yellow) in Pennsylvania, New York, and New Jersey. (B) Geologic map showing outcrops studied in this research near Blossburg, Pennsylvania (red circle) and other Upper Devonian outcrops in north-central Pennsylvania (yellow circles). Bedrock geology from Pennsylvania Geological Survey. Adapted from Broussard et al. (2018).

Berg and Edmunds, 1979; Risser et al., 2015). Berg and Edmunds (1979) inferred this red to gray transition and the entire Huntley Mountain Formation succession to record a transition from meandering to braided river deposition through time, based on qualitative sedimentologic observations. However, quantitative fluvial architecture data characterizing river systems spanning the Catskill to Huntley Mountain Formation transition are lacking.

Recent highway widening of U.S. Route 15 created a series of roadcut outcrops in

Tioga County, Pennsylvania exposing Upper Devonian and Lower Mississippian strata

(Broussard et al., 2018). The two outcrops studied in this research near Blossburg,

Pennsylvania (Blossburg South and Blossburg West), are located in an area mapped by the Pennsylvania Geologic Survey (1980) as lower Huntley Mountain Formation near the mapped boundary with the Catskill Formation (Fig. 4). Timing of deposition for both outcrops is constrained to latest Famennian time (~363-359 Ma) based on age diagnostic palynomorphs (Appendix B). This study aims to utilize this sequence of new outcrop exposures to characterize the sedimentology and architecture of latest Devonian river systems and to interpret the evolution of larger-scale variables within the northern

Appalachian foreland basin.

Figure 4: Geologic map showing the location of outcrops near Blossburg, Pennsylvania for this study (Blossburg South and West) and a companion study by Zaklicki (2020) (Blossburg Middle). Refer to Figure 3b for map location. Geology from Pennsylvania Geological Survey.

SEDIMENTOLOGY

The sedimentology for each of the Blossburg South and West outcrops is characterized both generally, across the entire outcrop, and precisely, for a layer-by-layer analysis of a vertical section line through the stratigraphy. The general sedimentological analysis produces a lithofacies table (i.e., a table describing all of the lithologies present at a given locality) and a facies association table (i.e., depositional processes or environments interpreted from associated lithofacies). The stratigraphic analysis (layer- by-layer approach) produces a stratigraphic column, recording the relative changes in sedimentary properties through time, and provides an opportunity to quantify the proportion of each lithofacies or facies association along a linear profile.

The close proximity between the Blossburg South and Blossburg West outcrops allows for estimation of the stratigraphic distance between the top of the exposed strata at

Blossburg South (older strata) and the bottom of the exposed strata at Blossburg West

(younger strata). Correlating the stratigraphy of these two outcrops enables an analysis of upsection continuities and/or upsection changes, effectively establishing trends of sedimentation processes over an extended interval of time. Vertical stratigraphic sections through the stratigraphically-older Blossburg North and Blossburg Middle outcrops, compiled by Slane et al. (2009) and Zaklicki (2020), respectively, can also be correlated using this approach. Refer to Fig. 4 for geographic locations of these outcrops along U.S.

Route 15. Assessing stratigraphic changes across this extended section line enhances an analysis of the evolution of Late Devonian rivers.

Methods

Outcrop sedimentology was characterized with field notes, photographs, and a measured section spanning from the bottom of Blossburg South to the uppermost exposure of the Blossburg West outcrop (320-453 m interval in Fig. 5). The section was measured with a Jacob staff perpendicular to bedding (~5-15° SE dip). Physical properties of rocks along the stratigraphic section line and at other stations across the outcrop were recorded (e.g., composition, grain size, color, sorting, bed geometry, contacts with other deposits, sedimentary structures, and fossils).

The thickness of an unexposed stratigraphic gap between the top of the Blossburg

South section and the bottom of the Blossburg West section was estimated geometrically based on the orientation of bedding at both outcrops. A horizontal distance and elevation change between the outcrops was calculated with the Google Earth linear distance measurement tool. The resulting gap is estimated to be less than 5 m and is shown as a covered interval on the measured section (Fig. 5).

Facies Observations and Interpretations

The physical characteristics of all rock types found at Blossburg South and West are split into nine lithofacies (Table 1). The unique combinations of lithofacies present in each facies and subfacies are summarized in Table 2. The entire thickness of the stratigraphic section spanning Blossburg South and West (Fig. 5) has been interpreted and divided into the three main facies groups (C, P, and D) (Table 2). The percentages of the stratigraphic column classified by each lithofacies and separately by each facies group

Figure 5: Detailed measured stratigraphic section of Upper Devonian strata exposed in roadcuts along US 15 near Blossburg, Pennsylvania. Blossburg North section (0-115m) from Slane et al. (2009); Blossburg Middle section (192-270m) from Zaklicki (2020); 192-453m measured in this study.

West outcrops West

Summary of lithofacies observations and interpretations for the lower Huntley Mountain Formation at Blossburg Mountain at lower the Huntley Formation for observations and interpretations lithofacies Summaryof

Table Southand

Table 2: Summary of subfacies, lithofacies, and depositional environments associated with each facies group. Refer to Table 1 for lithofacies codes.

Facies Group Subfacies Depositional Environments Lithofacies

Channel bar, point bar, Sx, Sr, Sh, Sm Cb thalweg, crevasse channel Channel Ce Lower bar, thalweg Se

Ca Abandoned channel Fp, Fl

Ps Crevasse splay Sr, Sh, Sm Proximal Floodplain Pm Crevasse splay, levee Sc, Fp, Fl

Dm Well-drained floodplain Fm Distal Floodplain Dl Floodplain pond or wetland Fl

are calculated for both Blossburg South and West (Table 3). Three fluvial facies associations were identified at the two outcrops based on interpreted mechanisms of sediment deposition: channel facies (C), proximal floodplain facies (P), and distal floodplain facies (D). Subcategories for each facies further differentiate depositional processes and settings within each facies environment (e.g., Cb = channel bar deposit vs.

Ce = scour fill) (Table 2).

Channel Facies

Channel deposits are composed of sandstones and mudstones generated within the margins of a river channel. This facies group includes three sub-facies: Cb, Ce, and Ca

(Table 2).

Table 3: Summary of the percentage of each lithofacies observed and the percentage of each facies group interpreted in each outcrop stratigraphic column.

Percentage of Percentage of Lithofacies or Facies Group Blossburg South (%) Blossburg West (%)

Sx 17.6 30.5

Sr 0.0 0.9

Sh 35.0 23.0

Sm 15.3 23.7

Se 3.3 9.5

Sp 1.4 0.9

Sc 0.0 4.4

Fp 13.3 1.5

Fm 13.0 0.0

Fl 1.1 4.6

Channel Facies 72.9 90.1

Proximal Floodplain Facies 10.7 9.9

Distal Floodplain Facies 16.4 0

Facies Cb is characterized by amalgamated greenish-gray, fine- to medium- grained sandstones. These deposits can be massive (lithofacies Sm), ripple scale cross- bedded (<10 cm tall ripples) (Sr), dune-scale trough or planar cross-bedded (10-105 cm tall dunes) (Sx), or horizontally laminated with parting lineations (Sh) (Fig. 6b-c). Bed sets of cross-bedded or horizontally laminated sandstone are commonly separated by millimeter-thin curvilinear mud linings or minor scours, creating thin recessively weathering horizons filled with mudstone rip-up clasts (Fig. 6c).

Facies Cb is interpreted as channel bar, point bar, thalweg, and crevasse channel deposits. A combination of ripple- to dune-scale cross-stratification and horizontal lamination shows that sediment was transported under both lower and upper flow regimes. The thin mud layers and recessive weathering horizons, seen in cross section at the outcrop surface, are interpreted as accretionary bar surfaces, recording the migration of sandy bars downstream (Friend, 1983; Miall, 1985). Erosional surfaces are interpreted to represent scouring during lateral channel migration or during an avulsion as the channel down-cut through floodplain deposits (Holbrook, 2001). Abundant mudstone rip- ups along the scour surface boundary indicates reworking of mud within the channel or on the floodplain through incision.

Facies Ce differs from Cb in that it is composed of moderately sorted sandstone or intraformational conglomerate characterized by abundant mudstone rip-up clasts and reworked caliche nodules (Se). Channel-belt-scale and/or intra-channel scour surfaces are always present below and occasionally present above Ce deposits. Beds are up to 1 m

Figure 6: Representative photos of channel facies: (A) channel body sandstone (Sh) overlying distal floodplain mudstone (Fm) with scour contact; (B) amalgamated beds with dune-scale trough cross-bedded sandstone (Sx) (hammer bottom right); (C) tangential crossbeds soling into surface separating bedsets (dashed line); (D) mudplug facies (Fl) overlain by channel-belt-scale scour; (E) scour fill deposit (Se) overlying channel-belt- scale scour; (F) isolated scour fills (Se) between two bar deposits (Sx). Images A, B, C, E, and F from Blossburg South; image D from Blossburg West. Hammer 40 cm long for scale.

25 thick and are either separated by a scour surface or transition upsection into Cb facies via an abrupt decrease in the abundance of mudstone rip-up clasts (Fig. 6a, e-f).

Facies Ce is interpreted as a scour fill deposit (Miall, 1985). The abundance of mudstone rip-ups and reworked caliche nodules indicates that these deposits resulted of bank slumping and/or incision into a levee or floodplain with moderately- to well- developed paleosols (Allen, 1983; Plint, 1986). Abrupt upsection transitions from beds of

Ce to beds of Cb suggest that these erosive events were short-lived and infrequent or rarely preserved.

Facies Ca, unlike Ce and Cb, is composed of <3 m thick beds of red to gray mudstone and very-fine-grained sandstone (Fp and Fl). Deposits always overlie sandstones with a distinct U-shaped bedding contact, and the top of deposits are always truncated by a channel-belt-scale scour (Fl lithofacies in Fig. 6d).

Facies Ca is interpreted as mud-plug fill deposits within an abandoned channel or bar-top slackwater pools (Lynds and Hajek, 2006; Burge and Smith, 1999). The U- shaped lower margin of the deposit is representative of the active bankfull flow dimensions of the river channel prior to abrupt abandonment. Subaqueous settling of muds suspended in slack-water within an abandoned channel develops layers of mudstone within a channel facies element; this can occasionally undergo subaerial desiccation dry spells (Lynds and Hajek, 2006; Willis and Tang, 2010). The red color is interpreted to result from subaerial oxidation. Although it is expected to find desiccation cracks in subaerially exposed muddy sediment, no desiccation features were observed.

Proximal Floodplain Facies

Proximal floodplain refers to the extent of the floodplain that is episodically inundated by crevasse splays and flooding. Overbank sediments deposited in this reach contain muddy very fine-grained sandstones and siltstones with minimal pedogenic features. Crevasse splays occasionally deposit thin lenses of fine-grained sandstone interbedded with red, green, or gray muddy sandstone and siltstone deposits. This facies group is divided into two sub-facies based on dominant sandstone or mudstone lithology:

Ps and Pm (Table 2).

Facies Ps is characterized by thinly-bedded greenish-gray, very fine- to fine- grained sandstone lenses. The sandstone is commonly massive (Sm) yet occasionally contains ripple-scale cross-bedding (Sr) and/or horizontal lamination (Sh). Beds are commonly 10-40 cm thick and pinch-out laterally over tens of meters (Fig. 7b). Ps deposits always have an abrupt basal contact with mudstone below. At the top of each bed, sand fines upward into siltstone either abruptly or over a 5-10 cm range.

Facies Ps is interpreted as a crevasse splay sheet flood deposit created during a levee breach (Allen, 1965; Friend, 1983; Bridge, 2006). Ripple cross-bedding and horizontal stratification indicate that some crevasse splays exhibit lower and upper flow regime during a splay event. Fining upward sequences at the tops of some of these sandstone lenses indicate a gradual slowing of the unidirectional current as the proximal floodplain aggrades. Desiccation cracks and rooting are possible features found at the tops of crevasse splay deposits, yet none were observed at the two Blossburg sites.

Figure 7: Representative photos of proximal floodplain facies: (A) red silty sandstone of levee facies (Fp); (B) interbedded tan to buff crevasse splay sandstone (Sm, Sr) and gray overbank mudstone beds (Fp). Photo A from Blossburg South and photo B from Blossburg West.

Facies Pm is characterized by red and green siltstones and very-fine-grained sandstones with poorly developed pedogenic features and occasionally reworked mudstone rip-ups and caliche (Fp and Sc) (Fig. 7a). Pedogenic structures include rootlets, in-situ caliche nodules (<2 cm diameter), and rare convex-concave slickensides.

Intraclasts within occasional ~1 m thick beds include reworked caliche nodules up to

7 cm diameter and green and gray mudstone rip-ups <2 cm diameter. Beds are laterally continuous for tens of meters and may be vertically amalgamated to approximately 4 m thick. Sparse massive or laminated gray siltstone beds (Fl) are laterally equivalent to Sc deposits.

Facies Pm is interpreted as levee deposits aggraded through time via crevasse splays and levee-topping floods (Allen, 1965; Friend, 1983; Bridge, 2006). Reworked rounded caliche clasts and mudstone rip-ups indicate a turbulent incisional flow and/or bank slumping, yet the red very-fine-grained sandstone matrix (Sc) differentiates this proximal floodplain facies from greenish-gray fine- to medium-grained scour fill sandstones (Se) within a channel facies (Ce). The red color observed in lithofacies Sc likely indicates oxidation in a subaerially exposed environment; this further differentiates it from a thalweg or lower-bar scour fill deposit. The facies Pm is interpreted as deposition upon or beyond a channel levee while sand and reworked floodplain material is entrained in turbulent bankfull flow. Pedogenic structures suggest that sediments were inundated infrequently enough to develop soils. The occasional green mottling or massive coloration to these deposits is interpreted to represent persistent subaqueous exposure of

29 levee sediments. Gray siltstones (Fl) laterally equivalent to Sc deposits are interpreted as spatially limited regions of ponded water on or adjacent to the levee.

Distal Floodplain Facies

Distal floodplain refers to the lowest elevation floodplain beyond the extent of crevasse splay sandstone deposition. These deposits are composed entirely of mudstone with either well-developed or absent pedogenic features. This facies group consists of two subfacies: Dm and Dl (Table 2).

Facies Dm is characterized by red to red-gray relict-bedded to semi-fissile mudstone with well-developed pedogenic features, including rootlets, rhizoliths, concave-convex slickensides, and caliche nodules (Fm) (Fig. 8a-c). Occasional <4 cm thick horizons of light-green mudstone are interbedded locally. Mudstone bed sets are

1-5 m thick.

Facies Dm is interpreted as well-drained distal floodplain paleosols. The deep red color likely represents significant subaerial exposure and oxidation. Slickensides and caliche nodules suggest that the distal floodplain was very infrequently inundated, thus providing time for the diagenetic soil-building processes to overprint muddy sediment falling out of suspension in slack-water during floods (Bridge, 2006).

Figure 8: Representative photos of distal floodplain facies: (A) red relict-bedded mudstone (Fm); (B) caliche nodules and caliche horizon in Fm; (C) rootlets in Fm (color- enhanced to show feature). Photos A-C from Blossburg South.

Facies Dl is characterized by gray laminated silty mudstone in beds <1m thick

(Fl). No rootlets or pedogenic slickensides were observed in these deposits. This facies is interpreted as floodplain lake or pond deposits. The distinct gray color and lack of pedogenic features suggest persistent subaqueous conditions. The limited thickness of these deposits indicates that these lakes or ponds were likely not long-lived beyond the lifespan of a nearby channel-belt system.

Stratigraphic Changes from Blossburg South to Blossburg West

Channel Facies

Channel bar sandstone lithofacies (Cb) show a slight increase in maximum and minimum grain diameter range from the bottom of the Blossburg South stratigraphic column to the top of the Blossburg West stratigraphic column (from 320-453 m in Fig.

5). The lithofacies Sx, Sh, Sr, and Sm, as seen at Blossburg South, contain lower-fine- to lower-medium-grained sand, whereas the same lithofacies at Blossburg West contain upper-fine- to upper-medium grained sand.

The percentage of channel facies captured in the stratigraphic column increases upsection from 72.9% of the total exposed Blossburg South section to 90.1% of the total exposed Blossburg West section (Table 3). Scour fill facies (Ce) are more common upsection at Blossburg West in comparison with Blossburg South. There are only eight beds of lithofacies Se observed in Blossburg South sand bodies, whereas 15 or more beds are observed throughout Blossburg West, which has a smaller overall stratigraphic

32 thickness. This change is over a three-fold increase in scour fill deposit density across

~130 m of stratigraphic thickness.

Proximal Floodplain Facies

At Blossburg South all greenish-gray crevasse splay sandstone beds (Ps) are consistently isolated between layers of red siltstone or silty sandstone. Splay sandstones throughout Blossburg West are primarily tan to buff and interbedded with olive-gray siltstone and silty sandstone. Additionally, the lithofacies Sc is not present at Blossburg

South.

Distal Floodplain Facies

The presence of well-drained floodplain paleosols at the outcrops decreases significantly upsection; lithofacies Fm comprises 13.0% of the Blossburg South stratigraphic column and 0.0% of the Blossburg West column (Table 3). Similarly, the measured percentage of distal floodplain facies for each outcrop decreases from 16.4% to

0.0% from Blossburg South to West (Table 3).

At outcrop locations laterally adjacent to the measured section, small unmeasured pockets of Fm lithofacies at Blossburg West display a more red-gray color compared to the lighter red color of Blossburg South paleosols. No significant in-situ caliche nodules were observed in these Blossburg West deposits. Consistent with the observations of grayer mudstone upsection, the stratigraphically lowest observed appearance of a definitive olive-gray or dark-gray mudstone occurs at 388m, near the top of the Blossburg

South section. Above this measure, greenish-/olive-/dark-gray mudstone becomes the dominant sediment color. This could potentially suggest soil drainage decreased through time across this interval, yet further paleopedologic analysis is required to definitively assert an upsection change in floodplain drainage (e.g., Oest, 2015).

FACIES ARCHITECTURE

The steepness of the road-cuts that produced the Blossburg South and West outcrops exposes a cross-section view through lower Huntley Mountain Formation strata and enables us to map, measure, and quantify sedimentary facies and fluvial surfaces present in the rocks. Lithofacies observations from stations across the outcrops were used to infer facies associations (e.g., channel, proximal floodplain, and distal floodplain) and to overlay facies polygons onto outcrop cross-section images, thus characterizing the general deposit architecture. Fluvial surfaces (e.g., bar surfaces, intra-channel scours, and channel-belt-scale scours) were identified and mapped. The resulting architecture maps provide qualitative and quantitative data regarding the stacking of channels, identify proximal floodplain deposits, and enable a more accurate facies proportion calculation for a cross-section area, instead of merely a linear measurement provided by the measured stratigraphic section. Whole outcrop architecture is compared to analyze stratigraphic trends.

Facies Architecture Mapping Methods

Constructing facies architecture maps for Blossburg West and Blossburg South involved three steps: 1) collecting and registering outcrop point cloud data with a terrestrial laser scanner and 3D visualization software; 2) creating high-resolution photographic images of both outcrops; and 3) drawing facies polygons, bar surfaces, and scours onto the high-resolution outcrop image through field mapping and remote analysis in the computer lab.

Terrestrial Laser Scanning (TLS) and Point Cloud Registration

Terrestrial laser/lidar scanning (TLS) is a novel tool for remotely collecting spatial measurements and digitally modeling outcrops. TLS modules collect both crude panoramic photographs and detailed point cloud data with a camera and a rotating mirror that bounces a depth-sensing laser pulse off surfaces. These pixel and spatial point datasets can be merged in a 3D analysis software to create a digital outcrop model

(DOM) where the image is overlaid on a point-cloud. This model provides the opportunity to accurately and remotely measure the geometry of outcrop features. This tool allows for quantitatively assessing the architectural elements of an inaccessible outcrop or when limited time is available for field measurements.

The TLS module used in this study is a FARO® FocusS 350/350 Plus (Fig. 9).

This model has a scanning range up to 350 m, a panoramic camera built-in, and a GPS unit to track scanning locations. In the field, the TLS was mounted on a tripod and positioned 30 to 100 m away from the base of the outcrop. Three scan locations were used at Blossburg South and two at Blossburg West (Fig. 10). At each location two scans were taken: a low-resolution 360° scan to provide a base-map for point registration and a higher-resolution scan limited to the extent of the outcrop. The product-specific software,

FARO® Scene, was used to register point clouds and overlay panoramic photographs taken by the TLS. Combining the point clouds provided the ability to visualize the outcrop with multi-angle imagery, effectively generating a DOM.

Figure 9: Faro® FocusS 350/350 Plus TLS module mounted on a tripod. (Image from https://www.faro.com/products/construction-bim/faro-focus/)

Registration settings were manipulated to allow for processing of far-distance lower resolution scans.

One challenge with collecting data spanning a highway (only necessary at

Blossburg South) is the occasional inclusion of a large vehicle in a scan or panoramic photo; this affected the image quality for the Blossburg South scan. To rectify the issue, the individual panoramic photo-overlays were re-colorized to provide higher contrast.

This editing enabled important geologic features to appear more prominently for accurate measurement, and it did not add any potential error for distance values obtained from the remastered DOM.

Figure 10: Site map of TLS and GigaPan data collection locations for (A) Blossburg South and (B) Blossburg West. 3D visualizations show TLS and GigaPan points in relation to outcrop topography for (C) Blossburg South and (D) Blossburg West. Map extent for (A) and (B) in Figure 4. (2D and 3D basemap source: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community)

Outcrop Panoramic Photography

High resolution panoramic photography can be useful for remotely analyzing the architecture of an outcrop. In this project one goal is to develop facies maps of the exposed deposits for both Blossburg South and West. These geometric and sedimentary characterizations of the outcrop are useful for inferring river planform, avulsion dynamics, basin-infilling parameters, and volumetric estimation for the facies composition of an outcrop or geologic interval (Friend, 1983). To develop these maps, one must first have a high-resolution base-map image of the outcrop, and second have the ability to draw in surfaces and boundaries representative of fluvial architecture elements and transitions between sedimentary facies. This requires a combination of both in-the- field and remote measurements and mapping judgements.

To enable remote surface mapping, high-precision photographs were collected from the outcrops spanning up to 100+ m width and 50+ m height. To create these images, photos were taken with a Canon EOS Rebel T3 camera through a Canon 70-300 mm lens (set to 300 mm magnification) mounted on a GigaPan EPIC Pro V from 50-100 m away from the outcrop. The GigaPan module is specially designed robotic camera mount for taking panoramic photos (Fig. 11). Once given geometric instructions to capture the entire outcrop at a given magnification, it automatically orients the camera to the positions in a grid and takes a photo at each location. Each photomosaic contains

300+ individual photos. These photos were stitched together in the software GigaPan

Stitch after inputting the grid information. The panoramic photos were then downloaded locally and uploaded to www.gigapan.com where they can be viewed publicly

Figure 11: GigaPan EPIC Pro V with a mounted camera. (Image from http://www.omegabrandess.com/products/Gigapan/GigaPan-EPIC-ProV)

(Blossburg South: http://gigapan.com/gigapans/216101; Blossburg West: http://gigapan.com/gigapans/216743). The Adobe Photoshop RAW format images generated in GigaPan Stitch were then compressed into multiple files representing varying levels of detail for printing and graphic work in Adobe Illustrator.

Mapping Bar and Scour Surfaces

Bar and scour surfaces were interpreted and mapped using a similar architecture mapping methodology to Mohrig et al. (2000), Chamberlin and Hajek (2015), and

Chamberlin et al. (2016). Scours identified at the Blossburg outcrops are representative of

4th-6th order erosional surfaces characterized and mapped by Holbrook (2001). Two

40 scales of scour surface are characterized for this study: intra-channel/intra-channel-belt scours (4th and 5th order) and channel-belt-scale scour surfaces (6th order).

A bar surface is defined as the upper bounding surface of a clinoform sandy bar bed-set (Mohrig et al., 2000; see Paleochannel Depth section). Bed-sets are comprised of a single sandstone lithofacies (e.g., Sx, Sh, Sm, etc.) and typically contain either cross- or horizontal-bedding surfaces. Inclined surfaces can occasionally appear nearly horizontal depending on the orientation of the outcrop cross-section view. Figure 12 demonstrates the relationship between mapped cross-beds (white lines) and bar surfaces (blue lines) and exemplifies both horizontal and inclined stacking of bed sets. Bar surfaces are typically either near-linear or gently concave-up. Figure 13 shows examples of both surface geometries. As seen in outcrop, the base and top of each curvilinear bar surface must terminate at a scour surface, representing either the base of a channel bar or a truncation surface.

Intra-channel scours result from the lateral migration of an active channel within a channel belt system over older bar deposits. Intra-channel scour surfaces are typically overlain by a thin (<10 cm thick) bed of intraformational conglomerate densely packed with mudstone rip-up clasts. Intra-channel scours typically continue laterally for 5-30 m and cut across bar and bedding surfaces. The upper end of an intra-channel scour surface is always truncated by either a channel-belt-scale scour or another intra-channel scour

(Figs. 12 and 13). These scours can either be smooth and sub-horizontal or have an irregular geometry with up to 2 m of relief. Some scours are overridden by thin deposits

(<1 m thick) of scour fill facies, Se.

s s

) )

(Bottom

). ).

ig. ig.

not F not

Selected area of Blossburg South outcrop area Selected

(Top)

apping facies polygons, scour surfaces, bar surfaces, and surfaces. polygons, scour facies apping

Figure Figure location image 1 (from panoramic of the area selected integration demonstrate to of map Facies in contacts lithofacies, proces the surface geometry, and bed m of

utcrop utcrop

Facies map of selected area to to of selected area map Facies

bar surfaces. bar

(Bottom) (Bottom)

Selected area of Blossburg West o area Selected

s in the process of mapping facies facies mapping of process the in s polygons,

: (Top) (Top) :

panoramic image. image. panoramic Figure Figure lithofacies, geometry, integration and bed of the demonstrate contact surface and surfaces, scour

Channel-belt-scale scour surfaces indicate avulsion of an entire channel-belt system. For this research, a strict criterion was set for interpreting channel-belt-scale scours: only scours outlining a contact between different facies and extending over 30 m are interpreted to represent a channel-belt avulsion. These scours could separate either two channel subfacies (i.e. scour fill overlying channel bar or channel bar overlying channel abandonment) or channel facies overriding floodplain facies. Scour fill deposits above a channel-belt-scale scour are typically at least 1 m thick, distinguishing this pattern from a smaller-scale intra-channel scour. Figure 14 provides three close-up examples of how scours that meet these criteria may appear in outcrop.

There are varying degrees of certainty in surface mapping accuracy depending on outcrop exposure. Occasionally, it is challenging to identify and distinguish between the three surface categories. For example, some bar surfaces appear to grade laterally into intra-channel scour surfaces; these segments were mapped independently as bars or scours. Long surfaces that track laterally into poorly-exposed intervals and surfaces identified remotely in areas with poor image resolution are mapped with dashed lines.

Figure 14: Examples of applying mapping criteria to channel-belt-scale scours: (A) scour surface visibly continues for >30 m; (B) scour defines the contact between channel bar facies below (Sx) and scour fill facies above (Se); scour fill deposit >1 m thick; (C) scour defines the contact between a preserved pocket of distal floodplain facies below (Fm - red mudstone with well-developed pedogenic features) and scour fill or channel facies above (Se, Sx). Images A and B from Blossburg West and image C from Blossburg South. Lithofacies codes are from Table 1.

Summary of the Architecture Maps for Two Blossburg Outcrops

Figure 15 shows the Blossburg South facies architecture map. Four spatially isolated sand bodies are preserved at Blossburg South. A summary of dimensional measurements for sand bodies and internal stories is present in Table 4. Levee deposits, in the shape of channel “wings,” laterally bracket a thin (<3 m thick) single-story sand body in the lower left of the outcrop. The other three sand bodies extend for >150 m across the entire outcrop. Inclined lateral accretion bar surfaces are present throughout the outcrop (see Fig. 12). Poor exposure toward the top of the outcrop prevented detailed surface mapping of the upper-most sand body.

Figure 16 shows the Blossburg West facies architecture map. Only one continuous amalgamated sand body is identified at Blossburg West; this is divided by five channel-belt-scale scour surfaces into six stories. A summary of the dimensional measurements from this singular MSB is also present in Table 4. Distal floodplain facies is only present in one location as a <10 m wide, <2, m thick lens isolated within proximal floodplain facies. Channel-belt-scale scours frequently overlie <2 m thick lenses of channel-abandonment facies. Lateral accretion bar surfaces are also present throughout

Blossburg West (see Fig. 13). These are commonly truncated by scour surfaces and exhibit no bar-top rollover.

Figure 15: (Top) Panoramic image of the Blossburg South outcrop taken on 04/29/2019 with the aid of a GigaPan EPIC Pro; image can be found at http://gigapan.com/gigapans/216101. (Middle) Outcrop image mapped with facies polygons, channel-belt-scale scours, intra-channel scours, bar surfaces, and rock sample locations. (Bottom Right) Legend explaining colored lines, overlays and symbols.

Figure 16: (Top) Panoramic image of the Blossburg West outcrop taken on 07/12/2019 with the aid of a GigaPan EPIC Pro; image can be found at http://gigapan.com/gigapans/216743. (Middle) Outcrop image mapped with facies polygons, channel-belt-scale scours, intra-channel scours, bar surfaces, and rock sample locations. (Bottom Right) Legend explaining colored lines, overlays and symbols.

Table 4: Summary of dimensional measurements of sand bodies (SB) and internal stories (IS) from Blossburg South and West. Measurements are based on facies maps from Figures 15 and 16.

Blossburg Blossburg Measurement South West

Number of definitive sand bodies 4 1

Number of internal stories >5 >5

SB max thickness (m) 13 >30

SB min thickness (m) 2 ?

SB median thickness (m) 6 (n=3) >30 (n=1)

IS max thickness (m) 10 7

IS min thickness (m) 2 2

IS median thickness (m) 3 6

IS mean thickness (m) 5 5

Facies Area Proportions

Using the digitized facies polygons from the architecture maps and linear distance measurements from 3D models of the two Blossburg outcrops, the areal proportion of each facies for both outcrops was calculated and compared to the linear proportion calculated using the stratigraphic column within the measured area.

Methods of Measuring Area Proportions

To use the facies architecture maps digitally drawn on the panoramic images, multiple length scales needed to be used to compare area measurements from the center and edges of the maps. To standardize area measurements across the image, each outcrop architecture map was broken into a six-tile rectangular/trapezoidal grid using a digital outcrop model (DOM). Images of both models can be seen in Figure 17 (Blossburg

South) and Figure 18 (Blossburg West). The vertices of the rectangular/trapezoidal grid lines were anchored along the surface of the outcrop at a fixed horizontal and vertical distance away from each other (Fig. 19). These vertices were located on the panoramic architecture maps, and the points were connected to re-draw the grids. Figure 20 shows the rectangular grid projected onto the Blossburg South outcrop map, and Figure 21 shows the trapezoidal grid projected onto the Blossburg West outcrop map. In both of these grid overlays, the stratigraphic column path is provided for the interval covered by the area measurement grids.

The true area of each grid cell was calculated based on height and width measurements on the DOM. The drawn area of each grid cell polygon was measured and

terrestrial laser scanner (TLS) stationed at multiple locations. Labeled stationed scanner multiple Labeled (TLS) laser at locations. terrestrial

m a a m

and yellow line used to create 50 50 yellow bar. create to and scale m used line

from Faro® Scene displaying a front view of the Blossburg South digital outcrop model (DOM). model of displaying view Blossburg Scene Southdigital the front fromoutcrop Faro® a

Screenshot Screenshot

: :

Figure Figure fro data point cloud stitched with generated DOM with measurement black dashed distance

tal outcrop model (DOM). (DOM). model outcrop tal

ned at multiple locations. Labeled ned multiple Labeled at locations.

Screenshot from Faro® Scene displaying an oblique view of the Blossburg West dig displaying the of Scene Blossburg view oblique from West Screenshot Faro® an

: :

Figure Figure statio from scanner data (TLS) laser point cloud stitched with terrestrial a generated DOM 50 yellow bar. create to and scale m with used line measurement black dashed distance

the the bottom

the outcrop and outcrop the and

inset at inset

outh. Measured lines the in Measured outh.

he grid was measured. The measured. was grid he

show how t how show

view in Faro® Scene of the of the in used Scene viewgrid area Faro® in rectangular

Example Example

: :

Figure Figure Blossburg S at for calculations facies each percentage been have corners two all dashed are rectangles yellow and black; and program to highlightedorange in points two of between overall distance the shows surface the on left m) distance This line. distance (~20 represented the by was vertical vertical the measurement height grid for used

projected )

are used in this image and in the area calculation. The purple line represents represents The line calculation. purple area are the in this in and image used

column path measured through the outcrop. The facies area grid is representative of of representative is grid area facies The outcrop. the through measured path column

South GigaPan image. The vertical and horizontal dimensions of selected segments from the grid are are grid the dimensions segments from vertical The horizontal and selected of image. GigaPan South

. Facies polygons from Fig. polygons Fig. Facies . from

Facies polygons measured to estimate the relative facies composition of the Blossburg South outcrop. Black SouthBlack outcrop. facies Blossburg estimate to compositionpolygons the of relative the Facies measured

: :

lined polygons represent the rectangular grid established in Faro® Scene on the digital outcrop model (Fig. grid outcrop digital the on Scene (Fig. model establishedrectangular the Faro® in represent polygons lined

an overlay of the stratigraphic of the overlay an Figure out Blossburg the onto white in labeled m 325 from 366 to stratigraphy

nto the the nto

) projected o ) projected

(Fig. 19) (Fig.

Blossburg West outcrop. Black Black outcrop. West Blossburg

BlossburgSouth

e outcrop. The facies area grid is representative of stratigraphy from from stratigraphy of grid representative is facies area The outcrop. e

l and horizontal dimensions of selected segments from the grid are labeled in in labeled are grid the segments from and l dimensions selected of horizontal

development for used method

re used in this image and in the area calculation. The purple line represents an overlay overlay an line purple represents The calculation. used area the in this in and re image

column path measured through th measured columnpath

epresent the rectangular, trapezoidal, and triangular grids established in Faro® Scene on the digital digital the on trapezoidal, triangular Scene grids and Faro® in rectangular, established the epresent

followinggrid same the

(

Facies polygons measured to estimate the relative facies facies estimate to compositionpolygons the of relative the Facies measured

igraphic

: :

the strat the

of of Figure Figure outlinedr polygons model outcrop vertica GigaPan The image. BlossburgWest from polygons Fig. Facies white. to m 424 m. 452

55 divided by the true area to standardize measurements of facies polygon area across the outcrop. Facies polygons were cropped to the extent of each grid cell, and the drawn area of each polygon was measured.

Area Proportion Results

The relative proportions of channel facies, proximal floodplain facies, and distal floodplain facies across the entire measured area at Blossburg South and West are summarized in Table 5. These values are compared to the relative facies proportions calculated from the equivalent stratigraphic column interval captured by the area measurement grids. Figures 22 and 23 show box and whisker plots with the distributions of facies area percentages for both outcrops. These distributions are based on the relative area percentages calculated for each grid cell (n = 6) and demonstrate the range in percentage values resulting from heterogeneous fluvial deposits. Calculations to get these values and outcrop percentage totals are shown in Tables C1 and C2 in Appendix C.

The facies area proportion calculations demonstrate an upsection change in the character of floodplain sedimentology over the stratigraphic interval 325-452 m. Distal floodplain deposits increase upsection from being the dominant overbank facies (26.5%) to being very uncommon (0.5%). Conversely, proximal floodplain deposits increase from

5.9% to 24.3% of the total area at the top of the section. Channel facies increase from

67.5% to 75.2% over the interval.

Comparing the linear sedimentological observations along the stratigraphic column with areal facies observations demonstrates the level of potential error incurred

366 m (Blossburg South) and and (Blossburg m South) 366

325

centages calculated for intervals intervals for calculated centages

s, and area grids shown in Appendix C (Tables C1 and C2). and C1 shown grids C area (Tables Appendix in and s,

mn per mn

portional composition of both outcrops for each facies. Area percentages with Area calculated facies. composition for each outcrops both portional of

. Raw data, data, calculation Raw .

the pro the

Summaryof

452 m (Blossburg m 452 West)

mapped facies polygons, and stratigraphic colu stratigraphic and polygons, facies mapped Table5 424

Figure 22: Box and whisker plot displaying the distribution of area percentages for each facies group generated as a sample from each of six grid polygons across the Blossburg South outcrop. Bar within the box represents median area percentage; X within the box represents mean percentage.

Figure 23: Box and whisker plot displaying the distribution of area percentages for each facies group generated as a sample from each of six grid polygons across the Blossburg West outcrop. Bar within the box represents median area percentage; X within the box represents mean percentage; circle represents an outlier.

58 when measuring heterogeneous fluvial deposits only in one dimension. Table 5 reveals that stratigraphic column facies proportions likely both underestimate the percentage of channel facies at Blossburg South and overestimate channel facies at Blossburg West.

The reverse relationship is true with respect to proximal floodplain facies. Distal floodplain measurements, however, are very similar between the one-dimensional and two-dimensional proportions. These results both suggest a significant upsection decrease in distal floodplain preservation.

PALEOCHANNEL DEPTH

Channel bars frequently aggrade to the bankfull-flow free water surface of a river

(e.g., Mohrig et al., 2000). This makes it possible to estimate bankfull paleoflow depth for ancient rivers based on measurements of the heights of bar surfaces exposed in outcrop. As a lithified deposit, the geometry of a laterally-accreted channel bar is preserved in cross-section as a series of curvilinear clinoform surfaces within a channel sand body (e.g., Bridge, 2006). The upper margin of a fully preserved bar deposit commonly contains finer-grained sand, ripple-scale cross-bedding, interbedded mudstones, and distinctive bar-top rollover (Chamberlin and Hajek, 2019).

Five bars at Blossburg South and seven bars at Blossburg West were confidently identified and measured. These depths were used to both characterize the morphology of latest Devonian rivers through time and to estimate river paleoslope (Lynds et al., 2014).

Bar surfaces can have a variable level of preservation based on observable bar-top rollover geometry (fully-preserved) or truncations by scours (partially-preserved)

(Chamberlin and Hajek, 2019). It is important to consider the potential error accumulated from underestimating bankfull paleoflow depth from measuring partially-preserved bars as a minimum flow depth. Figure 24 provides examples of both fully and partially preserved bar surfaces and how a paleochannel depth measurement can be gathered from each of these traces.

istance istance

urement urement

screenshot of a linear of linear a screenshot

ved bar surfaces. Fully vedsurfaces. bar

(C)

); );

example of applying bar height bar applying of example meas

(B) (B)

depicted in (B) and its vertical and horizontal components in Faro® Scene Scene components depictedin horizontal its Faro® vertical and and (B) in

top rollover, and partially preserved bars scours. by truncated d vertical preserved and The are partially rollover, top

w and high point is measured as bar as height; point high is measured and w

section model of a multistory sand body with both fully and partially preser partially both with body and fully sand multistory sectiona of model

preserved bar in the Blossburg West outcrop (image adapted from adapted from outcrop Fig. bar (image Blossburg West the in preserved

Cross

outlining the same bar surface outliningbar surface same the

(A)

: :

howing a two dimensional view of a 3D digital outcrop model); official height estimate = = m. 2.7 estimate height official model); 3D outcrop digital a of view dimensional two a howing

ment

(image s (image Figure Figure exhibit surfaces bar bar preserved lo surface bar a between partially a to method measur

Methods

Bar heights were identified in the field but measured remotely with a linear distance measurement tool within the FARO® Scene software, used for stitching and displaying the DOM generated with TLS point cloud data. Measuring all bar heights remotely this way, both physically accessible and inaccessible ones, controls any biases between using the digital linear distance tool and using a physical Jacob staff in the field.

Remote bar height measurements rely on the locations of bar and scour surfaces included in the facies maps drawn on GigaPan images (Figs. 15 and 16). Figure 24c demonstrates how the height, or relief, of a bar surface was measured in FARO® Scene.

The preservation potential of each bar (i.e., fully- or partially-preserved) was assessed graphically and sedimentologically. Bars that exhibited upward fining and/or bar-top rollover were classified as fully-preserved. Bars that did not meet these criteria and were truncated by scours were classified as partially-preserved.

Results and Interpretations

The heights of five bars from Blossburg South and seven bars from Blossburg

West were measured and recorded. The sampled bars at Blossburg South have a median height of 1.6 m and range from 0.9-1.6 m. Sampled bars at Blossburg West have a median height of 2.0 m and range from 1.5-2.7 m. Bar heights from these two outcrops and bar heights reported from the Blossburg Middle outcrop (Zaklicki, 2020) are plotted with respective stratigraphic position in Figure 25.

Figure 25: Measured bar heights from Blossburg Middle, South, and West outcrops plotted with stratigraphic height through measured section (Fig. 5). Color boxes represent stratigraphic extent of each Blossburg outcrop. Two trendlines and R2 values are based on linear regression, either including all points (dashed line) or excluding one outlier (dotted line; outlier marked in red). Closed circles represent data from fully-preserved bars, and open circles represent data from partially-preserved bars. Blossburg Middle data from Zaklicki (2020).

To perform an accurate comparison of each outcrop dataset, outliers must be considered. One of the points in the Blossburg Middle dataset is a statistical outlier

(measured height of 3.5 m > 1.5*IQR + Q3 = 3.1 m). Two trendlines and two Wilcoxon rank-sum tests were performed to investigate stratigraphic changes in bar height. For the rank-sum test, the null hypothesis (Ho) represented no significant association between bar height and stratigraphic level; the alternate hypothesis (Ha) represented a significant association; the significance threshold value (α) was set to 0.10. For the dataset including the outlier, linear regression returned an R2 value of 0.1076, and the rank-sum test returned a p-value of 0.0575 (< α = 0.10), rejecting the null hypothesis. For the dataset excluding the outlier, linear regression returned an R2 value of 0.4143, and the rank-sum test returned a p-value of 0.0087 (< α = 0.10), rejecting the null hypothesis with more confidence.

These results suggest it was likely that Late Devonian rivers deepened on average during this representative interval of time. The range in channel depth within each single outcrop suggests that there was local variability in flow depth near the Blossburg locality, which is common in alluvial river systems. Additionally, the average channel depth for each outcrop will likely be a slight underestimate due to partial bar surface preservation.

Three out of five bars at Blossburg South and four out of seven bars at Blossburg West are partially-preserved. This does not alter the interpretation of overall channel deepening, yet it precludes analysis of the amount of deepening.

GRAIN SIZE ANALYSIS

The coarseness of the sediment transported by a river is fundamentally linked to several channel variables. Analyzing the grain size distributions of channel bar deposits enables us to estimate paleoriver properties, such as flow velocity, sediment transportation (i.e., suspended- vs. bed-load dominant), and channel slope. The coarsest sediment transported through an alluvial network will be deposited at the base of channel bars. Measuring the particle size distributions of sandstone samples from the lower reaches of bar deposits thus enables a controlled quantitative analysis of the coarse fraction of grain sizes in the flow of several paleoriver systems. Capturing the coarsest grains prevents bias of underestimating flow velocity, which is estimated as a function of the median grain size when reconstructing river paleoslope (Lynds et al., 2014). The steps involved for grain size analysis include: rock sampling, sample disaggregation, laser particle size analysis, and statistical analysis of grain size data.

Methods

Five bars at both Blossburg outcrops were sampled for grain size analysis (n=10).

The locations on the outcrops for all samples are recorded in Figures 15 and 16. Roughly a 10 cm diameter sphere of minimally-weathered rock was retrieved from the bottom of each bar.

For the disaggregation process, a 2-5 g piece of the original rock sample was gently tapped and moved circularly in a mortar with a pestle; this carefully broke up all

65 sand grain clusters. The disaggregated sample was periodically checked under a microscope to verify the grains were minimally fractured. Grains were sequentially sieved (500-125 μm sieves) and re-mixed to ensure all grains were isolated. The re-mixed samples were then split six times to produce a 50-200 μg aliquot to be used in the laser diffraction particle size analyzer.

Subsamples were analyzed by Lily Pfeifer using a Malvern Mastersizer 3000 laser diffraction particle size analyzer at the University of Oklahoma. The analyzer utilizes the physical property and assumption that decreasing grain size will increase the diffraction angle of a laser beam (McCave et al., 1986). The mechanism involves a laser, a sample chamber, a series of lenses and mirrors to direct and expand the light path, and a receiver for light signals. As particles pass through the laser path, the volume percentage for each grain size bin (e.g., 400-454 μm grain diameter) is calculated. The probability density function (PDF) curve and cumulative distribution function (CDF) curve for each sample was calculated with these volume percentages. The median grain diameter (D50) for each sample was calculated as the value at 50% on the CDF curve. An exact value was estimated based on linear interpolation between the maximum and minimum grain-size bins bracketing the 50% mark.

Results and Interpretations

The five channel bar sandstone samples, representative of lithofacies Sx, Sh, Sm, and Se*, were labeled Bar 1 – Bar 5 for each outcrop (* see Sample Exclusion and Error

Analysis section). The observed grain sizes from the sandstone samples range from

66 approximately 1-600 μm with modes ranging from 163-272 μm and medians ranging from 41-221 μm. A PDF plot (Fig. 26a) and a CDF plot (Fig. 26b) were generated from

Blossburg South and West raw grain size data (Tables C3 and C4 in Appendix C).

Median grain sizes (D50) were plotted against stratigraphic position to create Figure 27.

Data from the Blossburg Middle outcrop (Zaklicki, 2020) was included in this scatter plot to extend the sampled stratigraphic interval at Blossburg.

Stratigraphic Changes

Geometric average grain size medians (D50avg) were calculated with Equation 1. A geometric average was chosen over a linear average because grain size data is presented on a logarithmic scale and organized logarithmically by orders of magnitude (see

Krumbein phi scale in Krumbein, 1934). The geometric average grain size for channel bar deposits at Blossburg South is 170 μm (n = 3**) and at Blossburg West is 173 μm

(n = 5) (** see Sample Exclusion and Error Analysis section).

Equation 1: Calculation of geometric average grain size with grain size median (D50) and sample size (n) values.

푛 퐺푒표푚. 퐴푣푔. 퐺푟푎푖푛 푆푖푧푒 (퐷50푎푣푔) = √∏ 퐷50

To determine if any significant stratigraphic change in grain size across the three outcrops is observed, a rank-sum test was performed with grain size median from

Blossburg Middle and West. The null hypothesis represents no change in grain size

Figure 26: (A) Volume percentage and (B) cumulative volume percentage particle size distributions for 8 channel bar sandstone samples from Blossburg outcrops. Points of intersection for curves and black dashed line represent D50 values summarized in Figure 27. For raw volume percentage data, see Tables C3 and C4 in Appendix C.

Figure 27: Scatter plot of median grain size (D50) for 14 channel bar sandstone samples from Blossburg Middle, South, and West. Blossburg Middle data from Zaklicki (2020). Linear regression trendline plotted with black dashes. Open circles excluded from trendline calculation (D50 < 62.5μm).

69 median from Blossburg Middle to Blossburg West, set with α = 0.10. The test returned a p-value of 0.0217 (< α = 0.10), rejecting the null hypothesis. This result suggests it is likely that the median particle size of the sediment being transported and deposited in latest Devonian rivers near the Blossburg locality increased through time.

Sample Exclusion and Error Analysis

Two channel bar sandstone samples from Blossburg South and one sample from

Blossburg Middle (data from Zaklicki, 2020) were excluded from particle size distribution and stratigraphic analyses:

* Bar 3 from Blossburg South initially appeared to be an Sm deposit, but was later re-interpreted as an Se deposit, which is not directly usable for paleoslope estimation.

Reinterpretation was aided by observations of a thin-section slide prepared from the original Bar 3 sample. The lithology confirmed from thin-section analysis was intraformational conglomerate (Se). It was thereby excluded from the grain size analysis.

** Bar 4 of Blossburg South was excluded because it has a calculated D50 of

40.61 μm, which is lower than the threshold 62.5 μm, the minimum diameter of very- fine-grained sand. One bar from Blossburg Middle was also excluded based on a D50 value below this threshold (see red cells in Table C5 of Appendix C).

It is unlikely that the methods performed for this grain size analysis would produce an overestimate of the true median grain size. It is possible that a small portion

(likely much less than 5%) of the disaggregated fine- to medium-grained sand particles

70 could have been fractured during the disaggregation process. However, this was not observed in more than a few grains when checking the sieved samples under the microscope. The process was also uniformly applied to all samples; thus, any error induced from disaggregation can be assumed to be constant across all samples, further validating any relative grain size comparisons.

The possible small underestimation induced from disaggregation does not account for the anomalously low D50 values calculated for Bar 4 of Blossburg South and the other excluded bar sample from Blossburg Middle. It is likely that the samples themselves were composed of finer grained material from an inter-bar fine deposit or a pocket of very muddy sand. It is also possible that significant experimental error was introduced when the sample was analyzed in the laser particle size analyzer.

PALEOSLOPE

Reconstructing the average gradient for latest Devonian river systems is a powerful tool for interpreting landscape and channel dynamics within the alluvial plain of the northern Appalachian foreland basin. Lynds et al. (2014) demonstrates the application of geomorphic relationships between grain size, channel depth, and river slope for estimating the paleoslope of ancient rivers. In this project, bar heights, measured as a proxy for flow depth, and median bar deposit grain size (D50), representing dominant sediment particle size, were used to estimate paleoslope and additional geomorphic parameters regarding particle shear stress and flow velocity. The mathematical calculations, estimated slope values and their accuracies, and potential sources of error are discussed in this section.

Methods

Two methods from Lynds et al. (2014) were used to estimate the paleoslope of river deposits preserved in the Upper Devonian lower Huntley Mountain Formation.

These formulae are developed for sandy suspended-load-dominant rivers. Late Devonian

Appalachian foreland basin paleorivers are inferred to have abundant suspended sediment given the presence of muddy overbank deposits and the grain size median range confined to very-fine- to fine-grained sand (Laursen, 1958; Friend, 1983). The applicability of the equations from Lynds et al. (2014) to our river systems is, however, an assumption in our methodology. The paper outlines three methods for estimating paleoslope, two of which are used in this study.

Method 1 provides a simple slope approximation based on the assumption of an ideal and constant fluid shear stress applied to river sediment at bankfull flow (Shields

* 3 3 number, τ bf = 1) and quartz sediment in water (ρs = 2.65 g/cm ; ρ = 1.00 g/cm ). This method requires the input of a median grain diameter (D50b) and a channel depth at bankfull flow (Hbf), estimated by bar heights (Eq. 2):

Equation 2: Method 1 calculation of river surface paleoslope (S) estimated with particle density (ρs), fluid density (ρ), R = (ρs - ρ) = 1.65 for quartz sand in water 0-30°C, median bar deposit grain diameter (D50b), and bar clinoform surface height (Hbf) (Lynds et al., 2014).

푅퐷 푆 = 50푏 퐻푏푓

The resulting paleoslope is dimensionless, representing a rise over run value. In cases where partially-preserved bar heights are the only available flow depth data, paleoslope values are overestimated.

Method 2 provides a refinement to method 1 by accounting for both the change in bankfull flow Shields number with varying particle Reynolds number (Eq. 3) and the

* skin-friction shear velocity (u sf) to particle settling velocity (ws) ratio constraints required for suspended-load-dominant sediment transportation (Lynds et al., 2014). The result of these additional variables is a multiplier function factored into Equation 2. While additional calculations are necessary, the same input variables, median grain diameter

(D50) and bankfull flow channel depth (Hbf), are required. Equations 5 and 6 provide the necessary auxiliary calculations to solve for method 2 paleoslope. The transition to

* suspended-load-dominant sediment transportation has been observed when u sf/ws = 2.0

(Julien, 2010). Lynds et al. (2014) suggests that for rivers predominantly depositing very-

* fine- to fine-grained sand (62.5-250 μm), a u sf/ws value of 3.1 produces optimum results.

Equation 3: Particle Reynolds number (Rep) calculation based on D50b, R, gravitational acceleration (g), and kinematic viscosity of water (ν) (assumed 20°C) (Lynds et al., 2014).

퐷 √(퐷 푅푔) 푅푒 = 50푏 50푏 푝 휈

Equation 4: Method 2 paleoslope formula combining Equation 2 with additional * constraints: shear velocity vs. settling velocity (u sf/ws), dimensionless settling velocity * * (W ), and particle Reynolds number (Rep); u sf/ws set to 3.1 (Lynds et al., 2014).

1 2 ∗ ∗ 3 푅퐷 푢푠푓 푊 푆 = 50푏 ( ( ) ) 퐻푏푓 푤푠 푅푒푝

Equation 5: Dimensionless settling velocity (W*) calculated as an empirically-derived function of dimensionless particle size (D*) (Lynds et al., 2014).

log 푊∗ = −3.76715 + 1.92944(log 퐷∗) − 0.09815(log 퐷∗)2 − 0.00575(log 퐷∗)3

+ 0.00056(log 퐷∗)4

Equation 6: Dimensionless particle size (D*) calculated as a function of particle density (ρs), fluid density (ρ), gravitational acceleration (g), median grain diameter (D50b), and kinematic fluid viscosity (ν) (assumed 20°C) (Lynds et al., 2014).

(휌 − 휌)푔(퐷 )3 퐷∗ = 푠 50푏 휌휈2

Results, Interpretations, and Error Analysis

The bar heights and median grain sizes of three bars from Blossburg South and five bars from Blossburg West with qualifying D50 values (>62.5 μm) were used to calculate paleoslope. Paleoslopes are plotted against stratigraphic height in Figure 28.

Data for Blossburg Middle from Zaklicki (2020) is included to extend the sampled stratigraphic interval at Blossburg. Among all three outcrops, individual paleoslopes range 0.60 x10-4 – 6.4 x10-4.

Geometric average paleoslopes were calculated with Equation 7. A geometric average was chosen over a linear average because slope data is presented on a logarithmic scale, is organized by orders of magnitude, and ranges in this study over a factor of ten from minimum to maximum. Bars from Blossburg South record a geometric average paleoslope of 2.1 x10-4 (method 1) or 2.3 x10-4 (method 2) (n = 3). Bars from

Blossburg West record a geometric average paleoslope of 1.4 x10-4 (method 1) or

1.5 x10-4 (method 2) (n = 5).

Equation 7: Calculation of geometric average paleoslope with paleoslope (S) and sample size (n) values.

푛 퐺푒표푚. 퐴푣푔. 푆푙표푝푒 (푆푎푣푔) = √∏ 푆

Figure 28: Paleoslopes calculated with (A) method 1 and (B) method 2 from Lynds et al. (2014) plotted against stratigraphic position. Color boxes represent stratigraphic extent of each Blossburg outcrop. Closed circles represent data from “fully-preserved” bars, and open circles represent data from “partially-preserved” bars (see Fig. 24 and Paleochannel Depth section). Red plus-sign and error bars represent geometric average slope for each outcrop with an uncertainty factor of 1.3. Blossburg Middle data from Zaklicki (2020).

To determine if any significant stratigraphic change in paleoslope across these three outcrops is observed, a Wilcoxon rank-sum test was performed with the method 1 and method 2 datasets with a significance threshold value (α) set to 0.10. The null hypothesis (H0) represents zero change in paleoslope from Blossburg Middle to

Blossburg West. The alternative hypothesis (Ha) represents a significant association between paleoslope and stratigraphic level. The method 1 dataset returned a p-value of

0.556 (> α = 0.10), not rejecting the null hypothesis. The method 2 dataset returned a p-value of 0.556 (> α = 0.10), also not rejecting the null hypothesis. These results suggest it is likely that the overall land surface gradient of the alluvial plain near the Blossburg locality did not exhibit a gradual increase or decrease during the studied interval in the

Late Devonian.

In Lynds et al. (2014), river slopes estimated from the deposits of modern rivers

(using method 2) and the true land surface gradients for those rivers are compared. Error analysis suggests that individual method 2 slope reconstructions can have an error of up to an order of magnitude; however, aggregated estimations (e.g., geometric averages) for a single locality are generally accurate within an error factor of 1.3 (Lynds et al., 2014).

The variability factor of 1.3 is shown in Figure 28 as error bars bounding the geometric average slope for each outcrop.

Utilizing this variability factor, we can confidently infer that the true average gradient of latest Devonian paleorivers in north-central Pennsylvania near Blossburg,

Pennsylvania was in the range of 0.98 x10-4 – 5.2 x10-4 during the Blossburg South sedimentation interval and 0.65 x10-4 – 3.4 x10-4 during the Blossburg west

77 sedimentation interval. Including data from Blossburg Middle (Zaklicki, 2020), we suggest that the average gradient of late Famennian paleorivers in north-central

Pennsylvania was within the range of 0.54 x10-4 – 5.2 x10-4.

* The use of a u sf/ws value of 3.1 based on D50 values within the very-fine- to fine- grained sand particle size range enhances the accuracy of the method 2 calculation with the assumption of full sediment suspension (Laursen, 1958). The median estimation error

* reported in Lynds et al. (2014) for using the generalized u sf/ws value of 2.0 is -8.1%; however, the reported median estimation error for very fine- to fine-grained sand deposits

* when a u sf/ws value of 3.1 was used is +0.93%. This suggests that based on data from

Lynds et al. (2014), this methodology is likely to produce a series of paleoslope estimates that are, on average, 0.93% above the true value. Although this overestimation has been documented, the values from the present study have not been adjusted.

Paleoslope estimated by both methods 1 and 2 can be highly affected by the level of precision and error inherent in channel depth and grain size measurements. All slope values calculated for partially-preserved bars have the potential of being significant overestimates. The fact that the fraction of partially-preserved bars out of the total number measured ranges only from 0.50-0.67 suggests a similar factor of overestimation is likely present in the paleoslope values calculated for all three outcrops. This supports the significance of any comparisons between average paleoslopes for each outcrop.

Any error introduced by grain fracturing from sample disaggregation during the grain size analysis process would result in an underestimation of true paleoslope by a small margin. This error is assumed to be a negligible for paleoslope estimations.

DISCUSSION

River Plan Form and Paleoenvironment

Latest Devonian strata at Blossburg South and West record fully terrestrial fluvial environments with no evidence for tidally-influenced deltaic or shallow marine facies.

Single-story sand bodies and stories within multistory sand bodies are interpreted as the deposits of migrating rivers and bars within a channel belt system. Table 6 summarizes the sedimentological and architectural characteristics of the Blossburg outcrops and general stratigraphic changes.

The architecture of Blossburg South is interpreted as a succession of mixed-load meandering river systems within a mobile channel belt. Low-angle bar surfaces with consistent dip directions above channel-belt-scale scours suggest lateral/downstream accretion through channel migration (e.g., lateral accretion surfaces in Fig. 13). Scour fill and/or bank-slump facies deposited over intra-channel scours record lateral incision and scouring into cut-bank and floodplain sediments. The moderate proportion of channel deposits (<70%) relative to floodplain deposits, sedimentary characteristics, and overall architecture is most consistent with the “classic sandy, mixed-load meandering river” model from Miall (1985). The stratigraphic column and facies map for Blossburg South is also consistent with block diagram and stratigraphic column models for a mixed-load meandering river in Friend (1983). The low stream gradient estimated for Blossburg

South (~2.3 x10-4) is consistent with slopes from several modern meandering rivers

(0.5-5 x10-4), such as the Kansas River, the Minnesota River, and the Milk River of

Table 6: Summary of sedimentological and architectural characteristics of three Blossburg outcrops and their corresponding interpreted upsection change (Blossburg Middle data from Zaklicki, 2020).

Blossburg Blossburg Blossburg Upsection Characteristic Middle South West Change Geometric average median 138 170 173 Increase bar deposit grain size (μm)

Area proportion of channel ~40% 67.5% 75.2% Increase facies

Area proportion of proximal ~50% 5.9% 24.3% N/A floodplain facies

Area proportion of distal ~10% 26.5% 0.5% Decrease floodplain facies

Dominant proximal Crevasse Crevasse Levee N/A floodplain facies splay splay

Sand body maximum 6.2 13 >30 Increase thickness (m)

Median thickness of internal 1.5 3 6 Increase SB stories

Median bar height 1.5 1.6 2.3 Increase (paleochannel depth) (m)

Geometric average method 2 Not 1.3 2.3 1.5 paleoslope (x10-4) significant

Montana (Lane, 1957; Leopold and Wolman, 1957). Additionally, floodplain cohesion, associated with the formation of meandering rivers, can be inferred from rootlets present in Blossburg South overbank sediments, indicating bank-stabilizing riparian flora.

Evolutionary advances in plant root systems throughout the middle Paleozoic potentially increased bank stability significantly by Latest Devonian time (Gibling and Davies, 2010;

Ielpi and Lapôtre, 2020).

The architecture and sedimentology of Blossburg West is interpreted to represent deeper meandering rivers, coarser sediment, and increased incision and sediment reworking relative to Blossburg South. A high proportion of channel facies (>70%; i.e., high net-to-gross ratio) and amalgamation of sand bodies is commonly associated with braided river stratigraphy (e.g., Miall, 1977, 1978; Martin, 1993). However, the abundance of inclined bar surfaces suggests lateral migration and deposition in bank- attached bars were dominant processes. The low average paleoslope of Blossburg West rivers (~1.5 x10-4), lower even than that of Blossburg South, is also consistent with a meandering planform. The observed upsection increases in channel bar facies, scours, and intraformational conglomerates with mudstone rip-up clasts spanning the Blossburg

South and West sections indicates sediment reworking from cut-bank erosion and scouring from avulsions incising new channels through floodplain material (e.g.,

Slingerland and Smith, 2004; Hajek and Edmonds, 2014).

Although there is sedimentological evidence for increased reworking through time, architectural analysis shows four out of seven bars are partially preserved at

Blossburg West, whereas three out of five are partially preserved at Blossburg South

(Fig. 25). This small change in bar preservation does not support the interpretation of increased sediment reworking through time (Chamberlin and Hajek, 2019); however, the very small sample size and non-randomized bar selection process suggest this result is inconclusive until further evidence is gathered.

Foreland Basin Variables Influencing Stratigraphic Changes

The primary upsection changes across the Blossburg Middle to Blossburg West interval include an increase in grain size, an increase in channel depth, an increase in channel facies proportion (i.e., increase in overall sandiness), an increase in sand body amalgamation, and a reduction in distal floodplain preservation (Table 6). These changes in sedimentology and architecture can be explained by the interplay of three controlling factors: 1) accommodation-creation rate, 2) sediment supply, and 3) landscape/channel dynamics. Shifts in these variables could each independently contribute to the observed stratigraphic changes, yet they often shift together in response to basin-scale processes. It is likely these three variables contributed collectively to the changes in fluvial deposition throughout the Latest Devonian Appalachian foreland basin.

Accommodation refers to the volume within an alluvial system available for the aggradation of sediment (Posamentier and Vail, 1988; Jervey, 1988). If the accommodation-creation rate is lower than a system’s volumetric sediment supply, then erosion and deposit reworking will occur within the basin fill. Sediment supply refers to the actual characteristics of sediment being transported downstream within an alluvial network. This includes the specific lithology, grain size distribution, provenance, and rate

82 at which the sediment is eroded, entrained in flow, and transported. Fundamental changes in provenance and/or weathering may influence sediment supply and deposit architecture

(e.g., unroofing and erosion of muddier or sandier source material; Hajek and Heller,

2012).

The upsection transition from isolated single-story sand bodies to amalgamated multistory sand bodies documented across the Blossburg stratigraphy is consistent with fluvial architecture models for decreasing accommodation-creation rate relative to sediment supply rate (e.g., Bridge and Leeder, 1979; Heller and Paola, 1996). An upsection decrease in floodplain preservation and paleosol development is also consistent with hypothetical sedimentation models testing a decrease in accommodation-to- sedimentation (Kraus, 2002). Although these changes can be predicted by geometric models isolating accommodation-to-sedimentation ratio as the sole independent variable, landscape/channel processes can also change during periods of changing accommodation, thus increasing the complexity of the system. Furthermore, Colombera et al. (2015) shows that fluvial systems responding to changes in aggradation rates do not always produce predictable deposits reflective of synthetic stratigraphy models; the study suggests that caution should be used when inferring variations in accommodation or sediment supply from fluvial architecture analyses. It is important to consider the possible relationships between changing deposit characteristics and changing landscape/channel dynamics for any basin sedimentation reconstruction.

Recent stratigraphy models have demonstrated several channel processes that could build stratigraphy with increasing sandiness through time. Primary

83 landscape/channel dynamics include changes to avulsion style or pattern, migration/ avulsion rate, plan form, and channel geometry (e.g., width-to-depth ratio). Under constant accommodation and sediment supply rates, increasing channel depth has been shown to increase deposit reworking and channel body amalgamation (e.g., Hajek and

Heller, 2012). Channels prone to rapid migration and/or frequent avulsion should also produce sandier and more reworked deposits in comparison to those generated by less mobile channel-belts (e.g., Heller and Paola, 1996; Straub and Esposito, 2013).

Additionally, this same increase in sandiness could result from a shift from random avulsion relocations to a clustered avulsion pattern (Chamberlin and Hajek, 2015).

The observed changes in latest Devonian fluvial stratigraphy and inferred depositional/erosional processes may reflect myriad basin-scale variables. The possible impacts of three variables are discussed here: 1) basinward progradation of a distributive fluvial system; 2) tectonic uplift and subsidence rates; and 3) the Hangenberg global cooling interval, including a glacial advance and rapid sea level drop. These processes will be discussed individually, yet their influences are likely contemporaneous and interconnected.

Progradation of a Distributive Fluvial System

Distributive fluvial systems (DFSs) are generally classified as broadly fan-shaped depositional regions within an alluvial plain hosting bifurcating and mobile channel systems sourced from an apex (Fig. 29a) (Hartley et al., 2010; Davidson et al., 2013;

Weissmann et al., 2013; Oest, 2015). Channel size generally decreases downstream from

Figure 29: Generalized sedimentology and fluvial architecture model for a distributive fluvial system (DFS). (A) Plan view of bifurcating meandering channel systems. Idealized architecture models present for different facies tracts (e.g., proximal, medial, distal). (B) Idealized cross-section of a prograding distributive fluvial system succession along with expected stratigraphic changes. Adapted from Weissmann et al. (2013), Oest (2015), and Broussard et al. (2018).

85 proximal regions to distal regions of the DFS due to bifurcations (Weissmann et al.,

2010). Fluvial architecture also displays variations across proximal to distal channel deposits. Higher sandiness and larger, more amalgamated multistory sand bodies are expected in proximal regions, whereas thinner, spatially isolated single-story sand bodies are expected in medial to distal areas (e.g., Nichols and Fisher, 2007; Hartley et al., 2010,

Weissmann et al., 2013; Owen et al., 2015). Floodplain sediments in proximal regions are less abundant and contain well-drained paleosols; however, floodplain deposits from medial to distal reaches of the DFS are more abundant, better preserved, and generally contain poorly-drained paleosols (Hartley et al., 2013). The architectural characteristics predicted for channels in the upstream or downstream (i.e., proximal or distal) regions of a distributive system do not share similarities for the characteristics predicted for a non- distributive, conventional tributary fluvial system; in fact, several deposit characteristics for tributary fluvial systems show a reversed association with location along the channel profile in comparison to DFS models (see Table 1 in Oest, 2015).

Progradation of a DFS represents the basinward migration of DFS facies tracts

(Weissmann et al., 2013). Stratigraphic models for a prograding system would contain medial or distal fluvial deposits at the base of the section (rivers far from the fan apex) and proximal deposits higher in the section (rivers close to the apex). As seen in Figure

29b, a model cross-section through a prograding DFS succession predicts an upsection increase in sandiness, channel size, and sand body amalgamation. Lateral variation across the extent of a DFS has been shown to produce different juxtapositions and variable deposit thicknesses for each facies tract (i.e., proximal, medial, or distal), yet an

86 upsection increase in inferred proximity to the apex is almost always observed (e.g.,

Owen et al., 2015; Batezelli et al., 2019).

Late Devonian clastic sediments within the Catskill wedge record a large-scale westward progradation of marine and terrestrial environments in the Appalachian foreland basin (Ettensohn, 1985; Faill, 1985; Sevon, 1985; Harper, 1999). Recent studies have hypothesized that the presence of a prograding DFS in the Appalachian alluvial plain could be responsible for developing the Latest Devonian sedimentary successions throughout north-central Pennsylvania (e.g., Oest, 2015; Broussard et al., 2018). This hypothesis, currently applied to deposits of the Catskill Formation, is evaluated here for the transition from the Catskill Formation to the lower Huntley Mountain Formation.

The fine-grained sand bodies separated by laterally continuous floodplain deposits at the Blossburg Middle outcrop (uppermost Catskill Formation; architecture data from

Zaklicki, 2020) and Blossburg South outcrop (lowermost Huntley Mountain Formation) are consistent with medial/distal DFS deposit architecture. Higher in the section at the

Blossburg West outcrop (Huntley Mountain Formation), the high channel facies proportion and sand body amalgamation is more consistent with proximal/medial DFS deposit architecture. The additional upsection changes of increasing sandstone grain size and increasing channel depth are consistent with a progradational stratigraphic succession.

Although the fluvial architecture of latest Devonian Blossburg strata is consistent with progradational models for a transition to more proximal DFS facies tracts, architecture, paleocurrent, and age data from more stratigraphic sections is required to assert the presence of a distinctly fan-shaped prograding sediment accumulation in the alluvial plain. Oest (2015), Broussard et al. (2018), and Treaster et al. (2018) began the process of geographically arranging Late Devonian outcrops in north-central

Pennsylvania within the context of a northwest-prograding DFS; however, enhanced temporal and paleogeographic correlations between these outcrops is necessary for evaluating the viability of a Late Devonian DFS model for the Appalachian foreland basin (see correlation/quantification examples in Owen et al., 2015a, 2015b; Batezelli et al., 2019).

Additionally, further sedimentology and fluvial architecture studies are needed to evaluate evidence for progradation and a distributive fluvial system. The paleopedology for Blossburg South and West has not been characterized in detail (see Oest, 2015 for a detailed regional paleopedology study). According to a DFS progradation hypothesis,

Huntley Mountain Formation stratigraphy should contain better-drained paleosols than those of the Upper Catskill Formation. Furthermore, latest Devonian paleoslopes estimated during the present study (Table 6) are all within the range of river slopes for several modern DFSs (e.g. Hartley et al., 2010; Davidson et al., 2015; Kosi megafan in

Gaurav et al., 2015), yet a deeper analysis into the implications of an unchanging paleoslope during a DFS progradation event should be undertaken.

Tectonic Uplift and Basin Subsidence

Basin accommodation is commonly controlled by the rate of tectonic uplift, subsidence rate, and/or sea level change (Jervey, 1988; Viseras et al., 2003).

Landscape/channel processes adapt directly to fluctuations in river base level and uplift of the sediment source area, ultimately developing a new stream equilibrium profile

(Miall, 1996). This base level can be referred to as “relative sea level,” which represents a combination effect of fluctuating sea floor elevation, from basin subsidence, and true sea level, influenced by climate (Posamentier and Allen, 1999). Sediment aggradation will occur when an equilibrium profile shifts upward, as a result of relative sea level rise, and/or shifts basinward, as a result of uplift of the source terrane (Posamentier and Vail,

1988). Contrastingly, when base level drops, sediment erosion and reworking can occur, often producing incised valleys.

An interval of rapid uplift of bedrock source terranes east of the Appalachian foreland basin is inferred for Famennian time (370-360 Ma) based on thermochronologic datasets (Keppie and Dallmeyer, 1995; Murphy and Keppie, 2005; Kunk et al., 2005;

Wintsch et al, 2010; Proctor et al. 2013; Gardner, 2019). Evidence for rapid exhumation of metamorphic micas, supporting the interpretation of high uplift/exhumation rates, suggest high sediment input to the foreland basin alluvial networks during Famennian time (Broussard et al., 2018; Gardner, 2019). Regional stratigraphic studies suggest that the northern Appalachian basin experienced a transition from Late Devonian high subsidence rates to Early Mississippian low subsidence rates, resulting from isostatic rebound and orogenic quiescence (Faill, 1985; Chestnut, 1991; Ettensohn, 2008, 2009).

Sea level regression, punctuated by brief transgressive intervals, characterized Late

Devonian time (Woodrow and Sevon, 1985; Bridge, 2000; Ettensohn 2009). An alluvial system with high sediment supply and decreasing accommodation-creation resulting from lower subsidence and overall sea level regression would be expected to produce a progradational geometry from basinward migration of the stream equilibrium profile and fluvial facies tracts (Myers and Milton, 1996). Although, this notion supports the hypothesis for DFS progradation within the Appalachian foreland basin, a decrease in accommodation-to-sedimentation alone could produce the increases in sandiness and amalgamation observed across the Blossburg stratigraphic interval (Bridge and Leeder,

1979; Heller and Paola, 1996). The observed upward coarsening of bar deposits at

Blossburg could also be independently explained by rapid uplift/exhumation of bedrock sources and associated basinward migration of eroded sediments (Gasparini et al., 2004).

Global Cooling, Glaciation, and Sea Level Drop

The effects of the Hangenberg global cooling event at the end of the Devonian

(~361-359 Ma) include rapid sea level fluctuation, mass extinction of marine organisms, and worldwide glacial advance (Kaiser et al., 2006; Brezinski et al., 2010; Kaiser et al.,

2015; Lankin et al., 2016). Miospore data from two gray/black shales at Blossburg West suggest that both samples are placed in the Retispora lepidophyta - Verrucisisporites nitidus palyzone of the Late Famennian (LN biozone), which is contemporaneous with

Hangenberg climate change (see appendix for palynological data from Zippi, 2019 and

Fig. A1). In the northern Appalachian foreland basin, glacigenic diamictites within the

Spechty Kopf and Rockwell Formations in eastern and southern Pennsylvania record the advance of coalescing lowland glaciers in the proximal alluvial plain (Brezinski et al.,

2008, 2010). The glacio-lacustrine and braided fluvial stratigraphic succession overlying diamictite facies in the Spechty Kopf and Rockwell Formations is indicative of pro- glacial outwash streams choked with a high sediment supply (Sevon, 1985; Bjerstedt and

Kammer, 1988; Miller, 1989; Brezinski et al., 2008). Additionally, paleosol analyses of floodplain sediments coeval with Appalachian glacial deposits suggest a possible increase in precipitation within the northern foreland basin (e.g., Cecil et al., 2004; Retallack,

2006; Brezinski et al., 2010).

Deeper channels, resulting from increased paleodischarge from upstream glacial outwash and/or increased precipitation, could alone produce the increase in sandiness, amalgamation, and channel bar grain size observed through the Blossburg interval (Hajek and Heller, 2012). Increases in the depth of Pleistocene paleorivers have been attributed to a similar glacial advance and sea level drop (Flint, 1971; Reusser et al., 2004;

Brezinski et al., 2010). Glacial erosion often prompts high sediment supply (e.g., Hallet et al., 1996), which could also independently account for the observed upsection increase in sandiness and amalgamation observed in the Blossburg stratigraphy. Further studies into Huntley Mountain Formation sedimentation rate, the stratigraphic/temporal extent of channel deepening, and the duration of Appalachian glaciation are required to infer whether the entire >43 meter thick Blossburg West succession could have been influenced by pro-glacial outwash.

Although palynological analyses of Blossburg West sediments and glacial deposits of eastern Pennsylvania correlate the deposits to within the same biozone (LN) and within <2 Ma of each other, age correlations could be enhanced with detrital muscovite and zircon samples from Blossburg strata. Data from Gardner (2019) suggest detrital muscovite lag times were <6 Ma at the end of the Devonian, which can be used to more accurately infer true depositional age. Detrital zircons recovered from Spechty Kopf

Formation diamictites provide a maximum depositional age of 366 Ma for eastern

Pennsylvania glacial deposits (Phipps, 2020). Detailed detrital muscovite and zircon sampling from Spechty Kopf and Rockwell Formation diamictites and the Blossburg outcrops would provide enhanced age correlations between proximal glaciation and more distal fluvial deposits within the foreland basin during the end of the Devonian.

Rapid sea level fall at the beginning of the Hangenberg cooling event (LN biozone) is responsible for the formation of incised valleys (Fig. 30). Latest Devonian

Appalachian sediments deposited in these foreland basin incised valleys include the

Cussewago and Berea Sandstones (Dennison et al., 1986; Streel, 1986; Bjerstedt and

Krammer, 1988; Isaacson et al., 1999; Sandberg et al., 2002; Brezinski et al., 2010).

Incised valley fills are typically classified through sequence stratigraphy methods and identified by a lower-bounding valley fill surface (e.g., Posamentier and Vail, 1988;

Myers and Milton, 1996; Blum and Tornqvist, 2000; Holbrook, 2001). Although the observed upsection increase in sandiness and amalgamation is consistent with an expected fluvial response to sea level regression, no incised valley bounding surface was identified at either Blossburg South or West.

Figure 30: Approximated sea level fluctuations and correlated latest Devonian and Early Pennsylvanian stratigraphy for the Appalachian foreland basin and European successions. Miospore zonation used for age correlations. Refer to Figure A1 for estimated ages of latest Devonian biozones for Blossburg stratigraphy. Reproduced from Brezinski et al. (2010).

Furthermore, upsection increases in channel depth, grain size, sandiness, and amalgamation are present across all three studied Blossburg outcrops (Blossburg Middle data from Zaklicki, 2020). It is unlikely that a glacial advance and rapid sea level drop during LN biozone time is primarily responsible for the changes in sedimentology and fluvial architecture in Blossburg strata, yet these variables likely contributed to channel deepening and sand body amalgamation within a restricted stratigraphic interval, likely near or within the Blossburg West outcrop extent due to a similar inferred depositional age to that of glacial sediments. Incised valley fills commonly disconformably overlie older fluvial or marine strata (Posamentier and Vail, 1988). Enhanced age dating and stratigraphic/alluvial surface analysis throughout the Blossburg South to Blossburg West section would identify any missed disconformities in this study, which would be important for Latest Devonian sequence stratigraphy and interpreting effects of rapid sea level regression on landscape/channel processes.

Interpretation of Stratigraphic Changes and Future Directions

The effects of three basin-scale variables have been discussed: progradation of a

DFS, tectonic uplift and subsidence, and a global cooling interval. Upsection increases in channel depth, bar deposit grain size, channel facies proportion, and sand body amalgamation support the previously proposed hypothesis for a progradational DFS succession within the northern Appalachian foreland basin alluvial plain (Oest, 2015;

Broussard et al., 2018). The prograding DFS model was proposed for the Upper

Devonian Catskill Formation; this study hypothesizes that DFS progradation extended

94 into latest Devonian time during the deposition of the Huntley Mountain Formation.

Evidence for rapid source area uplift/exhumation and inferred decreasing basin subsidence during the Famennian (370-360 Ma) supports the model for progradation of the alluvial plain. During this time, sedimentation rates are inferred to be high while accommodation-creation rate is inferred to have decreased. This also likely contributed to the overall increase in sandiness and amalgamation observed in the Blossburg stratigraphy.

Although pro-glacial outwash, increased precipitation, and rapid sea level drop during the Hangenberg global cooling interval (~361-359 Ma) is consistent with increasing deposit sandiness, channel deepening, and the creation of incised valleys in coeval strata west of the Blossburg locality, direct evidence for changing sedimentology or fluvial architecture of Blossburg stratigraphy during this discrete glaciation time interval was not definitively observed in this study. The effects from global cooling and regional glaciation on landscape/channel processes could produce similar fluvial architecture to that observed at Blossburg West, yet most of the major stratigraphic changes analyzed in this study are present across three Blossburg outcrops and, hence, not confined to a discrete change at the beginning of the global cooling interval. As a result, the influence of the Hangenberg interval on Blossburg fluvial sedimentation cannot be confidently inferred.

Detailed sampling of detrital mineral grains at the base of each identified sand body story throughout the Blossburg section and throughout the upper Catskill and

Spechty Kopf Formation succession would enhance age constraints and relative timing of

95 deposition for eastern glacially-influenced strata and western fluvial strata within

Pennsylvania. Maximum depositional ages from single-grain analyses would add to a growing body of published Late Devonian foreland basin detrital ages and could be used for estimating absolute and relative depositional ages from known metamorphic muscovite lag-times (e.g., Gardner, 2019; Phipps, 2020). Petrographic analysis of thin sections from samples obtained throughout the Blossburg Middle to Blossburg West section would enhance published interpretations of sediment provenance for north-central

Pennsylvania fluvial successions (Gardner, 2019).

A higher resolution sedimentology and architecture dataset and a larger sample size for grain size analysis, bar height measurement, and paleoslope estimation throughout the Blossburg section would increase confidence in the observed upsection changes reported in this study. Low sample size due to poor outcrop exposure and resource limitations for the study presents a challenge for establishing conclusive trends.

Paleoslope estimations, in particular, would significantly benefit from increased sample size and enhanced error analysis, factoring in bar preservation. The lack of any upsection trend could represent variations in landscape morphology through time, yet it is not conclusive for supporting or denying the interpretation of a prograding DFS, rapid tectonic uplift, and/or relative sea level regression. It is suggested that the fluvial sedimentology and architecture of additional lower Huntley Mountain Formation outcrops in north-central Pennsylvania be studied to enhance the limited datasets for channel depth, grain size, and paleoslope datasets from the Blossburg outcrops.

Geographically correlating the sedimentology of these outcrops will have the benefit of

96 evaluating the DFS hypothesis, which is limited to a handful of outcrops (e.g., Oest,

2015; Broussard et al., 2018; Treaster et al., 2018). Refer to Owen et al. (2015a) for an example of a DFS reconstruction based on correlated stratigraphy among numerous outcrops. It is also recommended that a randomized study of bar preservation be added to the analysis of new Huntley Mountain Formation outcrops. This variable will enhance our understanding of channel deposit reworking and avulsion dynamics for latest

Devonian river systems (Chamberlin and Hajek, 2019).

CONCLUSIONS

The sedimentology and facies architecture of two lower Huntley Mountain

Formation outcrops in Blossburg, Pennsylvania record the evolution of latest Devonian river systems in the northern Appalachian foreland basin. Three major fluvial facies were identified through lithofacies analysis: 1) channel facies, composed of cross- and horizontally-stratified fine-grained single- and multistory sand bodies containing scours and lateral accretion surfaces; 2) proximal floodplain facies, composed of gray thinly- bedded crevasse splay sandstones and massive levee siltstones; and 3) distal floodplain facies, predominantly composed of red relict-bedded and fissile mudstone paleosols (with rootlets, slickensides, caliche nodules). The proportion of proximal floodplain deposits increases upsection, and distal floodplain deposits are only abundant downsection. Using high-resolution panoramic photography and a digital outcrop model built with terrestrial lidar point cloud data, fluvial facies architecture was mapped, identifying bar surfaces, intra-channel scours, and channel-belt-scale scour surfaces. Miospore ages from green to dark gray shales at both outcrops suggest the stratigraphy at Blossburg South is interpreted as late Famennian in age, in the VH biozone (~364-362 Ma); the stratigraphy at Blossburg West is interpreted as latest Famennian in age, in the LN biozone (~361-359

Ma).

Data from the stratigraphically older Blossburg Middle outcrop from Zaklicki

(2020), which has been correlated and appended to the stratigraphic column developed through both Blossburg South and West, is included in an analysis of overall stratigraphic

98 changes observed through Blossburg strata. Total sand body thickness, internal story thickness, and sand body amalgamation all increase upsection. The heights of bar surfaces (ranging from 0.6-3.5 m, median = 1.6 m), measured as proxies for paleochannel depth, and median bar deposit grain size (median diameters ranging from 99-221 μm) both increase upsection, suggesting channel deepening and sediment coarsening during this latest Devonian time interval. The average land surface slope of the latest Devonian alluvial plain near the Blossburg locality, interpreted from the geometric averages of paleoslope values, is inferred to be within the range of 0.54 x10-4 – 5.2 x10-4. Calculated paleoslopes do not exhibit any significant stratigraphic change, suggesting it is unlikely that the alluvial plain exhibited a gradual slope change during this interval.

Blossburg South and West architecture is representative of a succession of suspended- to mixed-load meandering rivers within mobile channel-belts due to the predominance of lateral accretion surfaces and low paleoslope values (<10-3). The stratigraphic changes observed in sedimentology and architecture are interpreted to result from controlling factors, such as accommodation-creation rate, sediment supply, and landscape/channel dynamics, responding to basin-scale variables evolving during latest

Devonian time. The variables discussed in this study include: 1) basinward progradation of a distributive fluvial system (DFS); 2) tectonic uplift and subsidence rates; and 3) the

Hangenberg global cooling interval, including a glacial advance and rapid sea level drop.

An upsection increase in channel depth, grain size, and channel facies proportion is consistent with sedimentation models for the progradation of proximal DFS facies tracts over medial to distal facies. These observations support the tentative hypothesis of

Oest (2015) and Broussard et al. (2018) for Late Devonian progradation of a DFS within the northern Appalachian foreland basin alluvial plain. Evidence for rapid uplift, high sediment supply, and decreasing basin subsidence during Famennian time suggests a decrease in accommodation-to-sedimentation, consistent with DFS progradation. A decrease in accommodation-to-sedimentation could independently contribute to an increase in channel facies proportion and sand body amalgamation. Future work developing sedimentological and architectural analyses of additional Huntley Mountain

Formation outcrops in north-central Pennsylvania is encouraged. In addition to increasing the sample size for channel depth, grain size, and paleoslope, the geographic and stratigraphic correlation of several measured sections is necessary to confirm the broad fan-shape geometry and facies succession of DFS deposits.

A period of global cooling at the end of the Devonian, inferred to overlap in age with Blossburg West strata (LN biozone), was characterized by the advance and retreat of lowland glaciers within the northern Appalachian foreland basin. Increased channel discharge and sediment supply from proglacial outwash and a wetter climate could potentially have influenced channel deepening and deposit coarsening from Blossburg

South to Blossburg west. However, evidence for incised valley filling from rapid sea level drop during this time, documented in western Pennsylvania successions, is not observed at Blossburg South or West. Future work enhancing relative age correlations between latest Devonian glacigenic strata in eastern Pennsylvania and fluvial strata in north-central Pennsylvania is necessary to confirm glacial influences on landscape/channel processes.

100

REFERENCES

Allen, J. R., 1965, A review of the origin and characteristics of recent alluvial sediments:

Sedimentology, v. 5, no. 2, p. 89-191.

Allen, J. R., 1983, Studies in fluviatile sedimentation: bars, bar-complexes and

sandstone sheets (low-sinuosity braided streams) in the Brownstones (L.

Devonian), Welsh Borders: Sedimentary Geology, v. 33, no. 4, p. 237-293.

Batezelli, A., Ladeira, F. S. B., Nascimento, D. L. D., and Da Silva, M. L., 2019, Facies

and palaeosol analysis in a progradational distributive fluvial system from the

Campanian–Maastrichtian Bauru Group, Brazil: Sedimentology, v. 66, no. 2, p.

699-735.

Berg, T. M., 1999, Chapter 8: Devonian- Mississippian Transition, in Shultz C.H. ed.,

The Geology of Pennsylvania: Pennsylvania Bureau of Topographic and

Geologic Survey Special Publication 1, p. 128-137.

Berg, T. M., and Edmunds, W. E., 1979, The Huntley Mountain Formation: Catskill-to-

Burgoon transition in north-central Pennsylvania: Pennsylvania Geological

Survey, 4th ser., Information Circular 83, 80 p.

Bjerstedt, T. W., and Kammer, T. W., 1988, Genetic stratigraphy and depositional

systems of the Upper Devonian-Lower Mississippian Price-Rockwell deltaic

complex in the Central Appalachians, USA: Sedimentary Geology, v. 54, no. 4,

p. 265-301.

101

Blum, M. D., and Tornqvist, T. E., 2000, Fluvial response to climate and sea level: a

review and look forward: Sedimentology, v. 47, p. 1-48.

Bradley, D. C., and O'Sullivan, P., 2017, Detrital zircon geochronology of pre‐and

syncollisional strata, Acadian orogen, Maine Appalachians: Basin Research, v.

29, no. 5, p. 571-590.

Brezinski, D. K., Cecil, C. B., Skema, V. W., and Stamm, R., 2008, Late Devonian

glacial deposits from the eastern United States signal an end of the mid-Paleozoic

warm period: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 268, no. 3-

4, p. 143-151.

Brezinski, D. K., Cecil, C. B., and Skema, V. W., 2010, Late Devonian glacigenic and

associated facies from the central Appalachian Basin, eastern United States: GSA

Bulletin, v. 122, no. 1-2, p. 265-281.

Bridge, J. S., 2000, The geometry, flow patterns and sedimentary processes of Devonian

rivers and coasts, New York and Pennsylvania, USA: Geological Society of

London, Special Publication 180, no. 1, p. 85-108.

Bridge, J. S., 2006, Fluvial facies models: recent developments: SEPM Special

Publication 84, 85 p.

Bridge, J. S., and Leeder, M. R., 1979, A simulation model of alluvial stratigraphy:

Sedimentology, v. 26, no. 5, p. 617-644.

Broussard, D. R., Trop, J. M., Benowitz, J. A., Daeschler, E. B., Chamberlain Jr, J. A.,

and Chamberlain, R. B., 2018, Depositional setting, taphonomy and

geochronology of new fossil sites in the Catskill Formation (Upper Devonian) of

102

north-central Pennsylvania, USA, including a new early tetrapod fossil:

Palaeogeography, Palaeoclimatology, Palaeoecology, v. 511, p. 168-187.

Cecil, C. B., Brezinski, D. K., and Dulong, F., 2004, The Paleozoic record of changes in

global climate and sea level: Central Appalachian Basin: Geology of the National

Capital Region, in U.S. Geological Survey Circular 1264, Field Trip Guidebook,

p. 77-133.

Chamberlin, E. P., and Hajek, E. A., 2015, Interpreting paleo-avulsion dynamics from

multistory sand bodies: Journal of Sedimentary Research, v. 85, no. 2, p. 82-94.

Chamberlin E. P., Hajek, E. A., and Trampush, S. M., 2016, Measuring scales of

autogenic organization in fluvial stratigraphy: an example from the Cretaceous

Lower Williams Fork Formation, Colorado: in Budd, D. A., Hajek, E. A., and

Purkis, S. J, eds., Autogenic Dynamics and Self-Organization in Sedimentary

Systems: SEPM Special Publication No. 15, 13 p.

Chamberlin, E. P., and Hajek, E. A., 2019, Using bar preservation to constrain

reworking in channel-dominated fluvial stratigraphy: Geology, v. 47, no. 6, p.

531-534.

Chestnut, D. R., 1991, Timing of the Alleghanian tectonics determined by Central

Appalachian foreland basin analysis: Southeastern Geology, v. 31, no. 4, p. 203-

221.

Colombera, L., Mountney, N. P., and McCaffrey, W. D., 2015, A meta-study of

relationships between fluvial channel-body stacking pattern and aggradation rate:

implications for sequence stratigraphy: Geology, v. 43, no. 4, p. 283-286.

103

Colton, G. W., 1968, Bedrock geology of the Waterville quadrangle, Lycoming County,

Pennsylvania: Pennsylvania Geological Survey, 4th ser., Progress Report 174, 1

pl.

Daeschler, E. B., Clack, J. A., and Shubin, N. H., 2009, Late Devonian tetrapod remains

from Red Hill, Pennsylvania, USA: how much diversity?: Acta Zoologica, 90, p.

306-317. doi:10.1111/j.1463-6395.2008.00361.x

Davidson, S. K., Hartley, A. J., Weissmann, G. S., Nichols, G. J., and Scuderi, L. A.,

2013, Geomorphic elements on modern distributive fluvial systems:

Geomorphology, v. 180, p. 82-95.

DeCelles, P. G., Carrapa, B., Horton, B. K., and Gehrels, G. E., 2011, Cenozoic foreland

basin system in the central Andes of northwestern Argentina: Implications for

Andean geodynamics and modes of deformation: Tectonics, v. 30, no. 6.

Dennison, J. M., 1985, Catskill Delta shallow marine strata, in Woodrow, D. L, and

Sevon, W. D, eds., The Catskill delta: Geological Society of America Special

Paper 201, p. 91–106. doi:10.1130/spe201-p91

Dennison, J. M., Beuthin, J. D., and Hasson, K. O., 1986, Latest Devonian-Earliest

Carboniferous marine transgressions: Central and Southern Appalachians, USA:

Annales de la Société géologique de Belgique, v. 109, p. 123-129.

Ettensohn, F. R., 1985, Controls on development of Catskill Delta complex basin-facies,

in Woodrow, D. L, and Sevon, W. D, eds., The Catskill delta: Geological Society

of America Special Paper 201, 65-77.

104

Ettensohn, F. R., 2009, The Mississippian of the Appalachian Basin, in Greb, S. F., and

Chestnut, D. R., eds., Carboniferous Geology and Biostratigraphy of the

Appalachian Basin: Kentucky Geologic Survey, p. 22-31.

Ettensohn, F. R., 2008, The Appalachian foreland basin in Eastern United States, in

Miall, A. D., eds., The Sedimentary Basins of the United States and Canada:

Sedimentary Basins of the World, v. 5, p. 105–179. doi:10.1016/s1874-

5997(08)00004-x

Faill, R. T., 1985, The Acadian orogeny and the Catskill Delta, in Woodrow, D. L, and

Sevon, W. D, eds., The Catskill delta: Geological Society of America Special

Paper 201, p. 15–38. doi:10.1130/spe201-p15

Flint, R. F., 1971, Glacial and Quaternary Geology: New York, John Wiley and Sons,

892 p.

Friend, P. F., 1983, Towards the ﬁeld classiﬁcation of alluvial architecture or sequence,

in Collinson, J. D., and Lewin, J., eds., Modern and ancient fluvial systems, p.

345-354.

Friend, P. F., Slater, M. J., and Williams, R. C., 1979, Vertical and lateral building of

river sandstone bodies, Ebro Basin, Spain: Journal of the Geological Society, v.

136, no. 1, p. 39–46. doi:10.1144/gsjgs.136.1.0039

Gardner, C. T., 2019, Late Devonian sedimentary record of Appalachian tectonics and

erosion: geochronology and geochemistry of detrital muscovite and zircon from

central Pennsylvania strata [Undergraduate Honors Thesis]: Bucknell University,

101 p.

105

Gaurav, K., Métivier, F., Devauchelle, D., Sinha, R., Chauvet, H., Houssais, M., and

Bouquerel, H., 2015, Morphology of the Kosi megafan channels: Earth Surface

Dynamics, European Geosciences Union, v. 3, no. 3, p. 321-331.

Gasparini, N. M., Tucker, G. E., and Bras, R. L., 2004, Network‐scale dynamics of

grain‐size sorting: Implications for downstream fining, stream‐profile concavity,

and drainage basin morphology: Earth Surface Processes and Landforms, v. 29,

no. 4, p. 401-421.

Gordon, E. A., and Bridge, J. S., 1987, Evolution of Catskill (Upper Devonian) river

systems; intra-and extrabasinal controls: Journal of Sedimentary Research, v. 57,

no. 2, p. 234-249.

Hajek, E. A., and Heller, P. L., 2012, Flow-depth scaling in alluvial architecture and

nonmarine sequence stratigraphy: example from the Castlegate Sandstone,

central Utah, USA: Journal of Sedimentary Research, v. 82, no. 2, p. 121-130.

Hajek, E. A., and Edmonds, D. A., 2014, Is river avulsion style controlled by floodplain

morphodynamics?: Geology, v. 42, no. 3, p. 199-202.

Hallet, B., Hunter, L., and Bogen, J., 1996, Rates of erosion and sediment evacuation by

glaciers: A review of field data and their implications: Global and Planetary

Change, v. 12, p. 213–235.

Harper, J. A., 1999, Chapter 7: Devonian, in Shultz C.H. ed., The Geology of

Pennsylvania: Pennsylvania Bureau of Topographic and Geologic Survey Special

Publication 1, p. 108-127.

106

Hartley, A. J., Weissmann, G. S., Nichols, G. J., and Warwick, G. L., 2010, Large

distributive fluvial systems: characteristics, distribution, and controls on

development: Journal of Sedimentary Research, v. 80, no. 2, p. 167-183.

Hartley, A. J., Weissmann, G. S., Bhattacharayya, P., Nichols, G. J., Scuderi, L. A.,

Davidson, S. K., Leleu, S., Chakraborty, T., Ghosh, P., and Mather, A. E., 2013,

Soil development on modern distributive fluvial systems: preliminary

observations with implications for interpretation of paleosols in the rock record,

in Driese, S. G., and Nordt, L. C., eds., New frontiers in paleopedology and

terrestrial paleoclimatology: Paleosols and soil surface analog systems: SEPM

Special Publication 104, p. 149-158.

Heller, P. L., and Paola, C., 1996, Downstream changes in alluvial architecture; an

exploration of controls on channel-stacking patterns: Journal of Sedimentary

Research, v. 66, no. 2, p. 297-306.

Holbrook, J., 2001, Origin, genetic interrelationships, and stratigraphy over the

continuum of fluvial channel-form bounding surfaces: an illustration from middle

Cretaceous strata, southeastern Colorado: Sedimentary Geology, v. 144, no. 3-4,

p. 179-222.

Huerta, P., Armenteros, I., and Silva, P. G., 2011, Large‐scale architecture in non‐marine

basins: the response to the interplay between accommodation space and sediment

supply: Sedimentology, v. 58, no. 7, p. 1716-1736.

Ielpi, A., and Lapôtre, M. G., 2020, A tenfold slowdown in river meander migration

driven by plant life: Nature Geoscience, v. 13, no. 1, p. 82-86.

107

Isaacson, P. E., Hladil, J., Jian-Wei, S., Kalvoda, J., and Grader, G., 1999, Late

Devonian (Famennian) glaciation in South America and marine offlap on other

continents: Gabhandlungen der Geologischen Bundesanstalt, v. 54, p. 239–257.

Jervey, M. T., 1988, Quantitative geological modeling of siliciclastic rock sequences and

their seismic expression, in Wilgus, C. K., Hastings, B. S., Kendall, C. G. S. C.,

Posamentier, H. W., Ross, C. A., and Van Wagoner, J. C., eds., Sea-level

change: An integrated approach: SEPM Special Publication 42, p. 47–70.

Kaiser, S. I., Steuber, T., Becker, R. T., and Joachimski, M. M., 2006, Geochemical

evidence for major environmental change at the Devonian–Carboniferous

boundary in the Carnic Alps and the Rhenish Massif: Palaeogeography,

Palaeoclimatology, Palaeoecology, v. 240, no. 1-2, p. 146-160.

Kaiser, S. I., Becker, R. T., Steuber, T., and Aboussalam, S. Z., 2011, Climate-controlled

mass extinctions, facies, and sea-level changes around the Devonian–

Carboniferous boundary in the eastern Anti-Atlas (SE Morocco):

Palaeogeography, Palaeoclimatology, Palaeoecology, v. 310, no. 3-4, p. 340-364.

Keppie, J. D., and Dallmeyer, R. D., 1995, Late Paleozoic collision, delamination,

shortlived magmatism, and rapid denudation in the Meguma Terrane (Nova

Scotia, Canada): constraints from 40Ar/39Ar isotopic data: Canadian Journal of

Earth Sciences, v. 32, no. 5, p. 644-659.

Kraus, M. J., 2002, Basin-scale changes in floodplain paleosols: implications for

interpreting alluvial architecture: Journal of Sedimentary Research, v. 72, no. 4,

p. 500-509.

108

Krumbein, W. C., 1934, Size frequency distributions of sediments: Journal of

Sedimentary Research, v. 4, no. 2, p. 65-77. doi:10.1306/d4268eb9-2b26-11d7-

8648000102c1865d

Kunk, M. J., Wintsch, R. P., Naeser, C. W., Naeser, N. D., Southworth, C. S., Drake, A.

A., and Becker, J. L., 2005, Contrasting tectonothermal domains and faulting in

the Potomac terrane, Virginia–Maryland—discrimination by 40Ar/39Ar and 82

fission-track thermochronology: Geological Society of America Bulletin, v. 117,

no. 9-10, p. 1347-1366.

Lane, E. W., 1957, Study of the shape of channels formed by natural streams flowing in

erodible material: U.S. Army Corps of Engineers Sediment Studies Program for

Missouri River Basin, Special Publication No. 9, 140 p.

Lankin, J. A., Marshall, J. E. A., Troth, I., and Harding, I. C., 2016, Greenhouse to

icehouse: a biostratigraphic review of latest Devonian–Mississippian glaciations

and their global effects: Geological Society of London, Special Publications, v.

423, no. 1, p. 439-464.

Laursen, E. M., 1958, The total sediment load of streams: Journal of the Hydraulics

Division, v. 84, no. 1, p. 1-36.

Leopold, L. B., and Wolman, M. G., 1957, River channel patterns: braided, meandering,

and straight: US Geologic Survey Professional Paper 282-B, 85 p.

Lynds, R., and Hajek, E., 2006, Conceptual model for predicting mudstone dimensions

in sandy braided-river reservoirs: AAPG Bulletin, v. 90, no. 8, p. 1273-1288.

109

Lynds, R. M., Mohrig, D., Hajek, E. A., and Heller, P. L., 2014, Paleoslope

reconstruction in sandy suspended-load-dominant rivers: Journal of Sedimentary

Research, v. 84, no. 10, p. 825-836.

Martin, J. H., 1993, A review of braided fluvial hydrocarbon reservoirs: the petroleum

engineer’s perspective: Geological Society of London, Special Publications, v.

75, no. 1, p. 333-367.

McCave, I. N., Bryant, R. J., Cook, H. F., and Coughanowr, C. A., 1986, Evaluation of a

laser-diffraction-size analyzer for use with natural sediments: Journal of

Sedimentary Research, v. 56, no. 4, p. 561-564. doi:10.1306/212f89cc-2b24-

11d7-8648000102c1865d

Miall, A. D., 1977, A review of the braided-river depositional environment: Earth-

Science Reviews, v. 13, no. 1, p. 1-62.

Miall, A. D., 1978, Lithofacies types and vertical profile models in braided river

deposits: a summary: Fluvial Sedimentology, Memoir 5, p. 597-604.

Miall, A. D., 1985, Architectural-element analysis: a new method of facies analysis

applied to fluvial deposits: Earth-Science Reviews, v. 22, no. 4, p. 261-308.

Miall, A. D., 1996, The Geology of Fluvial Deposits: Sedimentary Facies, Basin

Analysis and Petroleum Geology: Springer, New York.

Miller, J. M. G., 1989, Glacial advance and retreat sequences in the Permo-

Carboniferous section, central Transantarctic Mountains: Sedimentology, v. 36,

p. 419–430, doi: 10.1111/j.1365-3091.1989.tb00617.x.

110

Mohrig, D., Heller, P. L., Paola, C., and Lyons, W. J., 2000, Interpreting avulsion

process from ancient alluvial sequences: Guadalope-Matarranya system (northern

Spain) and Wasatch Formation (western Colorado): Geological Society of

America Bulletin, v. 112, no. 12, p. 1787-1803.

Murphy, J. B., and Keppie, J. D., 2005, The Acadian orogeny in the northern

Appalachians: International Geology Review, v. 47, no. 7, p. 663-687.

Myers, K. J., and Milton, N. J., 1996, Concepts and Principles of Sequence Stratigraphy,

in Emery, D. and Myers, K. J., eds., Sequence Stratigraphy: Blackwell Science,

Oxford, p. 9–41. doi:10.1002/9781444313710.ch2

Oest, C., 2015, Paleopedology and fluvial sedimentology of the Upper Devonian

Catskill Formation, central Pennsylvania: A test of the distributive fluvial system

model [Masters Thesis]: Temple University, 115 p.

Owen, A., Nichols, G. J., Hartley, A. J., Weissmann, G. S., and Scuderi, L. A., 2015b,

Quantification of a distributive fluvial system: the Salt Wash DFS of the

Morrison Formation, SW USA: Journal of Sedimentary Research, v. 85, no. 5, p.

544-56.

Owen, A., Nichols, G. J., Hartley, A. J., and Weissmann, G. S., 2015a, Vertical trends

within the prograding Salt Wash distributive fluvial system, SW United States:

Basin Research, v. 29, no. 1, p. 64–80. doi:10.1111/bre.12165

Phipps, A., 2020, Detrital muscovite geochemistry and detrital zircon geochronology of

the Late Devonian Spechty Kopf Formation: Improved constraints on sediment

provenance and depositional age [Undergraduate Thesis], Bucknell University.

111

Plint, A. G., 1986, Slump blocks, intraformational conglomerates and associated

erosional structures in Pennsylvanian fluvial strata of eastern Canada:

Sedimentology, v. 33, no. 3, p. 387-399.

Posamentier, H. W., and Vail, P. R., 1988, Eustatic controls on clastic deposition II -

sequence and systems tract models, in Wilgus, C. K., Hastings, B. S., Kendall, C.

G. S. C., Posamentier, H. W., Ross, C. A., and Van Wagoner, J. C., eds., Sea-

level change: An integrated approach: SEPM Special Publication 42, p. 125-154.

Posamentier, H. W., and Allen, G. P., 1999, Siliciclastic sequence stratigraphy; sequence

stratigraphy: concepts and applications: SEPM Concepts in Sedimentology and

Paleontology, no. 7, 210 p.

Proctor, B. P., McAleer, R., Kunk, M. J., and Wintsch, R. P., 2013, Post-Taconic tilting

and Acadian structural overprint of the classic Barrovian metamorphic gradient

in Dutchess County, New York: American Journal of Science, v. 313, no. 7, p.

649-682.

Retallack, G. J., 2006, Deep-calcic paleosols at times of oceanic anoxic events in Late

Devonian rocks of Pennsylvania: Geologic Society of America Abstracts with

Programs, v. 38, no. 7, p. 471.

Reusser, L. J., Bierman, P. R., Pavich, M. J., Zen, E., Larsen, J., and Finkel, R., 2004,

Rapid late Pleistocene incision of Atlantic passive-margin river gorges: Science,

v. 305, p. 499–502, doi: 10.1126/science.1097780.

Risser, D. W., Williams, J. H., Hand, K. L., and Behr, R., 2013, Geohydrologic and

water-quality characterization of a fractured-bedrock test hole in an area of

112

Marcellus shale gas development, Bradford County, Pennsylvania: Pennsylvania

Geological Survey, 4th ser., Open-File Report OFMI 13–01.1, 49 p.

Sandberg, C.A., Morrow, J.R., and Ziegler, W., 2002, Late Devonian sea-level changes,

catastrophic events, and mass extinctions, in Koeberl, C., and MacLeod, K.G.,

eds., Catastrophic Events and Mass Extinctions: Impacts and Beyond: Geological

Society of America Special Paper 356, p. 473–487.

Schumm, S. A., 1960, The shape of alluvial channels in relation to sediment type: U.S.

Geological Survey Professional Paper 352-B, 30 p.

Slane, D. C., and Rygel, M. C., 2009, Marginal-marine facies of the Catskill Formation

(Upper Devonian), Tioga County, Pennsylvania [abs.]: Geological Society of

America Abstracts with Programs, v. 41, no. 7, p. 145.

Slingerland, R., and Smith, N. D., 2004, River avulsions and their deposits: Annual

Review of Earth and Planetary Sciences, v. 32, p. 257–285.

doi:10.1146/annurev.earth.32.101802. 120201.

Slingerland, R., Patzkowski, M., and Peterson, D., 2009, Facies and sedimentary

environments of the Catskill systems tract in central Pennsylvania, in Crazy

about the Catskill, Field Trip Guide: Pittsburgh Association of Petroleum

Geologists, 46 p.

Straub, K. M., and Esposito, C. R., 2013, Influence of water and sediment supply on the

stratigraphic record of alluvial fans and deltas: Process controls on stratigraphic

completeness: Journal of Geophysical Research: Earth Surface, v. 118, no. 2, p.

625-637.

113

Streel, M., 1986, Miospore contribution to the upper Famennian-Strunian event

stratigraphy: Annales de la Société géologique de Belgique, v. 109, no. 1, p. 75-

92.

Van de Lageweg, W. I., Schuurman, F., Cohen, K. M., Van Dijk, W. M., Shimizu, Y.,

and Kleinhans, M. G., 2016, Preservation of meandering river channels in

uniformly aggrading channel belts: Sedimentology, v. 63, no. 3, p. 586-608.

Ver Straeten, C. A., 2013, Beneath it all: bedrock geology of the Catskill Mountains and

implications of its weathering: Annals of the New York Academy of Sciences

1298, p. 1-29. doi:10.1111/nyas.12221

Viseras, C., Calvache, M. L., Soria, J. M., and Fernández, J., 2003, Differential features

of alluvial fans controlled by tectonic or eustatic accommodation space.

Examples from the Betic Cordillera, Spain: Geomorphology, v. 50, no. 1-3, p.

181–202. doi:10.1016/s0169-555x(02)00214-3

Weissmann, G. S., Hartley, A. J., Nichols, G. J., Scuderi, L. A., Olson, M., Buehler, H.,

and Banteah, R., 2010, Fluvial form in modern continental sedimentary basins:

distributive fluvial systems: Geology, v. 38, no. 1, p. 39-42.

Weissmann, G. S., Hartley, A. J., Scuderi, L. A., Nichols, G. J., Davidson, S. K., Owen,

A., Atchley, S. C., Bhattacharyya, P., Chakraborty, T., Ghosh, P., Nordt, L. C.,

Michel, L., and Tabor, N. J., 2013, Prograding distributive fluvial systems:

geomorphic models and ancient examples, in Driese, S. G., and Nordt, L. C.,

eds., New frontiers in paleopedology and terrestrial paleoclimatology: Paleosols

and soil surface analog systems: SEPM Special Publication 104, p. 131-147.

114

Willis, B. J., and Bridge, J. S., 1988, Evolution of Catskill river systems, New York

state, in Devonian of the World: Proceedings of the 2nd International

Symposium on the Devonian System, Memoir 14: Sedimentation, v. 2, p. 85-

106.

Willis, B. J., and Tang, H., 2010, Three-dimensional connectivity of point-bar deposits:

Journal of Sedimentary Research, v. 80, no. 5, p. 440-454.

Wintsch, R. P., Kunk, M. J., Mulvey, B. K., and Southworth, C. S., 2010, 40Ar/39Ar

dating of Silurian and Late Devonian cleavages in lower greenschist-facies rocks

in the Westminster terrane, Maryland, USA: Geologic Society of America

Bulletin, v. 122, no. 5-6, p. 658-677.

Zaklicki, M. B., 2020, Sedimentological, fluvial architecture, and paleoenvironment

analysis of the fluvial Upper Devonian Catskill Formation, north-central

Pennsylvania, USA [Undergraduate Honors Thesis]: Bucknell University.

115

APPENDICES

116

Appendix A: Comparison to Catskill Magnafacies in Southern New York

Studies of Middle and Late Devonian strata from the Catskill clastic wedge in southern New York (Gordon and Bridge, 1987; Willis and Bridge, 1988; Bridge, 2000) show similar sedimentological and architectural measurements to those from the

Blossburg outcrops. Outcrops exposing fluvial sediments of the Oneonta Formation

(Gordon and Bridge, 1987), an upper Givetian and lower Frasnian deposit (Ver Straeten,

2013), qualitatively show similar stacking patterns of single-story sand bodies and floodplain sediments to that of the upper Catskill Formation (Zaklicki, 2020) and

Blossburg South. Similarly, average slope estimates (0.5-1.0 x10-4), grain size

(120-180 μm), and paleochannel depth measurements are also consistent with upper

Catskill Formation (Blossburg Middle) measurements (Tables 6 and C5). Outcrops of the

Upper Walton Formation (Gordon and Bridge, 1987), a higher, middle Frasnian unit of

New York stratigraphy contemporaneous with marine deposits in north-central

Pennsylvania (Harper, 1999; Ver Straeten, 2013), qualitatively show similar amalgamated multistory sand bodies to those of Blossburg West. Average slope estimates

(0.72 x10-4) and paleochannel depth measurements are similar to two of the five

Blossburg West values, while average grain size (210 μm) is slightly coarser (Tables 6 and C5).

117

Appendix B: Palynology of the Blossburg Outcrops

Samples of red, green, and gray silty shales from the Blossburg outcrops were sent by Dr. Dave Broussard, Assistant Professor of Biology at Lycoming College,

Williamsport, Pennsylvania, to Biostratigraphy.com, LLC for palynological analysis.

Unpublished data from a report prepared by Dr. Pierre Zippi indicates miospore ages for the Blossburg South and West outcrops are within upper late Famennian time (Fig. B1).

Two green-gray silty shale samples from Blossburg South (at 382 and 405 m in the section) suggest sediments in the upper half of the outcrop were deposited within the upper VH biozone near the beginning of Strunian time (~362.6 Ma). The report suggests that the recovered spore assemblages in the Blossburg South samples could represent a downstream swamp margin paleoenvironment. Two black shale samples from Blossburg

West (between 410 and 450 m in the section) suggest outcrop sediments were deposited within the LN biozone during latest Famennian time (361-359 Ma). The report suggests that Blossburg West spores are also indicative of a downstream swamp margin paleoenvironment. Samples from the Blossburg South and West outcrops were collected by Dr. Jeff Trop, Professor of Geology at Bucknell University, Lewisburg, Pennsylvania.

Data from this report were provided generously through personal communications with

Dr. Broussard and Dr. Trop.

118

Figure B1: Composite palynozonation chart for late Famennian time with estimated biozone age boundaries (reproduced from Biostratigraphy.com, LLC report prepared by Dr. Pierre Zippi). Ages interpreted from recovered miospore assemblages are estimated with brackets on the right.

119

Appendix C: Supplementary Data Tables

Table C1: Facies area measurements and calculations for Blossburg South. Sum of Retraced Polygon Areas for Each Facies in Each Box Box # Facies Group 1 2 3 4 5 6 Channel 4.553 24.115 12.848 26.172 101.705 59.138 Proximal Floodplain 0.857 0.247 0 0.917 18.182 7.035 Distal Floodplain 4.757 14.353 4.493 9.635 19.278 5.893 (Unknown) 0 0.129 1.348 0 0 0 Total 10.167 38.844 18.689 36.724 139.165 72.066

Total - (Unknown) 10.167 38.715 17.341 36.724 139.165 72.066 Box Polygon Area 10.243 38.913 18.696 36.775 139.436 72.197 Area Accuracy of Polygons 99.3% 99.5% 92.8% 99.9% 99.8% 99.8%

Vertical Planar Area Represented by Box 500 1000 500 460 930 460

Box-Size Weighting Factor 49.179 25.744 26.754 12.526 6.683 6.383

Area Percentage for Each Facies in Each Box Box # Facies Group 1 2 3 4 5 6 Channel 44.78% 62.29% 74.09% 71.27% 73.08% 82.06% Proximal Floodplain 8.43% 0.64% 0.00% 2.50% 13.07% 9.76% Distal Floodplain 46.79% 37.07% 25.91% 26.24% 13.85% 8.18%

Area Values Weighted by Box Size for Each Facies Box # Facies Group 1 2 3 4 5 6 Total Total % Channel 223.91 620.82 343.73 327.83 679.67 377.48 2573.43 67.5% Proximal Floodplain 42.15 6.36 0.00 11.49 121.51 44.90 226.40 5.9% Distal Floodplain 233.94 369.50 120.20 120.69 128.83 37.62 1010.78 26.5% Total 500 997 464 460 930 460 3810.62

Box 1 10.243 Box 2 38.913 Box 3 18.696 Green = measured 500 m2 1000 m2 500 m2 polygon area Red = actual measured area in Box 5 Box 6 Box 4 36.775 139.436 72.197 vertical plane of 460 m2 930 m2 460 m2 outcrop

120

Table C2: Facies area measurements and calculations for Blossburg West.

Sum of Retraced Polygon Areas for Each Facies in Each Box Box # Facies Group 1 2 3 4 5 6 Channel 8.568 94.731 23.214 0.546 137.22 66.791 Proximal Floodplain 14.3 33.701 12.103 3.704 6.58 3.826 Distal Floodplain 0.919 0 0 0 0 0 (Unknown) 0.453 24.682 0 0 0 0 Total 24.24 153.114 35.317 4.25 143.8 70.617

Total - (Unknown) 23.787 128.432 35.317 4.25 143.8 70.617 Box Polygon Area 24.242 153.255 35.314 4.223 143.798 70.641 Area Accuracy of Polygons 98.1% 83.8% 100.0% 100.6% 100.0% 100.0%

Vertical Planar Area Represented by Box 315 945 315 45 608 353

Box-Size Weighting Factor 12.995 6.172 8.919 10.588 4.228 4.999

Area Percentage for Each Facies in Each Box Box # Facies Group 1 2 3 4 5 6 Channel 84.27% 244.69% 133.87% 1.49% 98.60% 92.68% Proximal Floodplain 140.65% 87.05% 69.79% 10.09% 4.73% 5.31% Distal Floodplain 9.04% 0.00% 0.00% 0.00% 0.00% 0.00%

Area Values Weighted by Box Size for Each Facies Box # Facies Group 1 2 3 4 5 6 Total Total % Channel 111.34 584.67 207.05 5.78 580.18 333.87 1822.89 75.2% Proximal Floodplain 185.83 208.00 107.95 39.22 27.82 19.13 587.94 24.3% Distal Floodplain 11.94 0.00 0.00 0.00 0.00 0.00 11.94 0.5% Total 309 793 315 45 608 353 2422.78

Box 1 24.242 Box 2 153.255 Box 3 35.314 Green = measured 315 m2 945 m2 315 m2 polygon area

Red = actual measured area in 4.223 Box 4 143.798 Box 5 2 Box 6 45 m2 608 m 70.641 vertical plane of 2 353 m outcrop

121

Table C3: Raw percentage of sample volume captured by each grain-size bin for five Blossburg West and four Blossburg South channel bar sandstone samples.

Raw Volume % for Grain Size Bins Samples Analyzed: 08/01/2019 by Lily Pfeifer

BLBW19- BLBW19- BLBW19- BLBW19- BLBW19- BLBS19- BLBS19- BLBS19- BLBS19- Sample ID 01-Bar1 07-Bar2 08-Bar3 04-Bar4 06-Bar5 03-Bar1 04-Bar2 06-Bar4 09-Bar5

Sample # 1 5 6 2 4 12 13 14 16 Size Volume Volume Volume Volume Volume Volume Volume Volume Volume Classes (%) (%) (%) (%) (%) (%) (%) (%) (%) (μm) 0.010 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.011 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.013 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.015 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.019 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.022 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.024 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.028 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.032 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.036 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.041 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.046 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.053 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.060 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.068 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.077 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.088 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.113 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.128 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.146 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.166 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.188 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.214 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.243 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.276 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.314 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.357 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.405 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.460 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.523 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.594 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.675 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.767 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.872 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.991 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.125 0.000 0.000 0.001 0.001 0.001 0.000 0.000 0.001 0.000 1.279 0.000 0.001 0.085 0.090 0.081 0.001 0.001 0.115 0.001 1.453 0.001 0.045 0.173 0.148 0.139 0.098 0.062 0.279 0.109 1.651 0.082 0.102 0.288 0.221 0.210 0.183 0.110 0.501 0.196 1.875 0.115 0.142 0.412 0.298 0.286 0.278 0.160 0.754 0.291

122

2.131 0.144 0.179 0.530 0.368 0.355 0.374 0.209 1.010 0.384 2.421 0.169 0.210 0.632 0.425 0.411 0.465 0.254 1.255 0.467 2.750 0.190 0.236 0.721 0.470 0.456 0.555 0.298 1.496 0.544 3.125 0.214 0.262 0.807 0.511 0.497 0.653 0.347 1.751 0.619 3.550 0.241 0.288 0.894 0.552 0.541 0.759 0.405 2.028 0.697 4.034 0.272 0.316 0.980 0.596 0.589 0.871 0.471 2.313 0.777 4.583 0.306 0.343 1.055 0.638 0.637 0.976 0.538 2.575 0.848 5.207 0.338 0.366 1.109 0.674 0.680 1.062 0.600 2.783 0.902 5.916 0.364 0.385 1.138 0.702 0.718 1.120 0.650 2.916 0.935 6.722 0.384 0.399 1.144 0.722 0.750 1.148 0.685 2.972 0.949 7.637 0.396 0.409 1.129 0.735 0.777 1.145 0.702 2.949 0.945 8.677 0.399 0.416 1.100 0.744 0.800 1.115 0.701 2.857 0.928 9.858 0.396 0.423 1.065 0.751 0.820 1.063 0.684 2.711 0.904 11.201 0.389 0.433 1.031 0.762 0.840 0.997 0.657 2.531 0.879 12.726 0.385 0.451 1.006 0.781 0.860 0.927 0.626 2.336 0.859 14.458 0.389 0.480 0.994 0.813 0.883 0.861 0.601 2.146 0.847 16.427 0.406 0.522 0.996 0.861 0.911 0.806 0.590 1.973 0.845 18.664 0.441 0.578 1.011 0.927 0.945 0.766 0.604 1.826 0.852 21.205 0.497 0.647 1.035 1.012 0.991 0.742 0.646 1.706 0.864 24.092 0.574 0.723 1.065 1.116 1.052 0.731 0.721 1.611 0.879 27.373 0.668 0.801 1.100 1.237 1.135 0.732 0.827 1.539 0.895 31.100 0.777 0.873 1.143 1.374 1.249 0.740 0.958 1.487 0.917 35.335 0.893 0.933 1.204 1.526 1.402 0.757 1.105 1.454 0.951 40.146 1.012 0.979 1.296 1.692 1.602 0.786 1.258 1.441 1.012 45.613 1.134 1.015 1.437 1.876 1.854 0.838 1.409 1.451 1.118 51.823 1.265 1.057 1.647 2.082 2.159 0.927 1.557 1.487 1.290 58.880 1.417 1.131 1.939 2.317 2.513 1.077 1.708 1.557 1.549 66.897 1.614 1.270 2.319 2.589 2.904 1.311 1.880 1.665 1.909 76.006 1.886 1.515 2.783 2.903 3.320 1.653 2.099 1.815 2.374 86.355 2.264 1.902 3.310 3.263 3.741 2.118 2.399 2.009 2.934 98.114 2.775 2.457 3.866 3.664 4.148 2.710 2.808 2.242 3.562 111.473 3.428 3.183 4.406 4.095 4.525 3.411 3.343 2.504 4.218 126.652 4.213 4.055 4.884 4.535 4.856 4.183 3.998 2.775 4.852 143.897 5.088 5.016 5.256 4.954 5.128 4.969 4.739 3.034 5.411 163.490 5.978 5.981 5.491 5.316 5.328 5.696 5.504 3.253 5.849 185.752 6.788 6.847 5.570 5.582 5.442 6.286 6.205 3.405 6.129 211.044 7.411 7.507 5.489 5.713 5.451 6.671 6.746 3.466 6.228 239.780 7.754 7.871 5.258 5.680 5.336 6.802 7.040 3.422 6.134 272.430 7.751 7.880 4.891 5.463 5.077 6.655 7.027 3.266 5.848 309.525 7.382 7.516 4.407 5.059 4.667 6.236 6.688 3.003 5.376 351.670 6.671 6.801 3.824 4.481 4.112 5.572 6.040 2.646 4.738 399.555 5.672 5.789 3.161 3.757 3.433 4.703 5.134 2.210 3.960 453.960 4.447 4.542 2.431 2.917 2.656 3.670 4.023 1.712 3.068 515.772 3.061 3.127 1.651 1.994 1.811 2.519 2.769 1.169 2.095 586.001 1.554 1.587 0.832 1.008 0.914 1.276 1.405 0.591 1.058 665.793 0.009 0.009 0.005 0.006 0.005 0.007 0.008 0.003 0.006 756.449 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 859.450 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 976.475 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1109.435 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1260.499 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1432.133 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1627.136 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1848.692 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

123

2100.416 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 2386.415 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 2711.357 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 3080.544 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 3500.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 MEDIAN = D50 217.67 221.1 150.76 151.5 141.63 187.4 159.4 40.61 165.4 (μm)

124

Table C4: Cumulative percentage of sample volume captured by each grain-size bin for five Blossburg West and four Blossburg South channel bar sandstone samples.

Cumulative Distribution of Grain Size Samples Analyzed: 08/01/2019 by Lily Pfeifer

BLBW19- BLBW19- BLBW19- BLBW19- BLBW19- BLBS19- BLBS19- BLBS19- BLBS19- Sample ID 01-Bar1 07-Bar2 08-Bar3 04-Bar4 06-Bar5 03-Bar1 04-Bar2 06-Bar4 09-Bar5

Sample # 1 5 6 2 4 12 13 14 16 Size Volume Volume Volume Volume Volume Volume Volume Volume Volume Classes (%) (%) (%) (%) (%) (%) (%) (%) (%) (μm) 0.010 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.011 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.013 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.015 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.019 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.022 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.024 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.028 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.032 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.036 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.041 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.046 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.053 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.060 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.068 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.077 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.088 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.113 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.128 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.146 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.166 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.188 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.214 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.243 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.276 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.314 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.357 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.405 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.460 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.523 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.594 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.675 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.767 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.872 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.991 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.125 0.000 0.000 0.001 0.001 0.001 0.000 0.000 0.001 0.000 1.279 0.000 0.001 0.086 0.091 0.082 0.001 0.001 0.117 0.001 1.453 0.001 0.045 0.259 0.239 0.221 0.099 0.063 0.396 0.111 1.651 0.083 0.147 0.547 0.460 0.431 0.281 0.173 0.897 0.307 1.875 0.197 0.289 0.959 0.758 0.718 0.559 0.333 1.651 0.598

125

2.131 0.341 0.468 1.489 1.126 1.073 0.933 0.542 2.661 0.982 2.421 0.510 0.678 2.121 1.551 1.483 1.398 0.796 3.915 1.449 2.750 0.701 0.914 2.843 2.021 1.939 1.953 1.094 5.411 1.993 3.125 0.914 1.176 3.649 2.532 2.436 2.606 1.441 7.162 2.611 3.550 1.155 1.464 4.543 3.084 2.978 3.365 1.846 9.190 3.308 4.034 1.427 1.780 5.523 3.681 3.567 4.236 2.317 11.503 4.085 4.583 1.733 2.123 6.578 4.319 4.204 5.212 2.856 14.078 4.933 5.207 2.071 2.489 7.688 4.993 4.884 6.274 3.456 16.860 5.835 5.916 2.435 2.874 8.826 5.695 5.602 7.394 4.106 19.777 6.770 6.722 2.819 3.273 9.969 6.417 6.352 8.542 4.791 22.748 7.719 7.637 3.215 3.682 11.098 7.151 7.129 9.687 5.493 25.698 8.664 8.677 3.613 4.098 12.199 7.895 7.929 10.801 6.194 28.555 9.592 9.858 4.009 4.521 13.263 8.646 8.750 11.864 6.879 31.266 10.496 11.201 4.398 4.954 14.295 9.409 9.589 12.861 7.536 33.797 11.375 12.726 4.783 5.406 15.301 10.190 10.450 13.788 8.162 36.133 12.234 14.458 5.172 5.886 16.294 11.003 11.333 14.649 8.762 38.279 13.082 16.427 5.577 6.408 17.290 11.864 12.243 15.455 9.353 40.252 13.927 18.664 6.018 6.986 18.301 12.790 13.189 16.221 9.956 42.078 14.778 21.205 6.515 7.633 19.336 13.803 14.180 16.963 10.602 43.784 15.642 24.092 7.089 8.356 20.401 14.919 15.231 17.694 11.324 45.395 16.521 27.373 7.757 9.158 21.501 16.156 16.366 18.426 12.151 46.934 17.416 31.100 8.534 10.031 22.644 17.530 17.615 19.166 13.109 48.422 18.332 35.335 9.427 10.964 23.848 19.056 19.016 19.923 14.213 49.876 19.283 40.146 10.439 11.942 25.144 20.748 20.618 20.710 15.471 51.317 20.295 45.613 11.573 12.957 26.581 22.624 22.473 21.547 16.880 52.767 21.413 51.823 12.838 14.015 28.227 24.707 24.632 22.475 18.437 54.254 22.703 58.880 14.254 15.145 30.166 27.024 27.145 23.552 20.145 55.811 24.251 66.897 15.869 16.415 32.485 29.612 30.049 24.863 22.025 57.476 26.160 76.006 17.755 17.930 35.269 32.515 33.369 26.515 24.124 59.291 28.534 86.355 20.019 19.832 38.579 35.778 37.109 28.634 26.523 61.300 31.468 98.114 22.793 22.289 42.445 39.442 41.257 31.344 29.332 63.543 35.031 111.473 26.221 25.472 46.851 43.537 45.783 34.754 32.674 66.046 39.249 126.652 30.435 29.527 51.734 48.071 50.639 38.938 36.672 68.821 44.101 143.897 35.523 34.543 56.990 53.025 55.767 43.907 41.411 71.855 49.512 163.490 41.501 40.524 62.481 58.340 61.096 49.603 46.915 75.108 55.361 185.752 48.288 47.371 68.051 63.922 66.538 55.889 53.120 78.513 61.490 211.044 55.700 54.878 73.540 69.636 71.989 62.560 59.865 81.979 67.718 239.780 63.454 62.749 78.798 75.316 77.325 69.362 66.906 85.401 73.852 272.430 71.204 70.629 83.690 80.779 82.402 76.017 73.933 88.667 79.700 309.525 78.586 78.145 88.097 85.837 87.069 82.254 80.621 91.670 85.076 351.670 85.257 84.946 91.921 90.318 91.181 87.826 86.661 94.315 89.814 399.555 90.929 90.735 95.082 94.075 94.614 92.528 91.795 96.525 93.774 453.960 95.376 95.277 97.512 96.992 97.270 96.198 95.818 98.237 96.842 515.772 98.437 98.404 99.164 98.986 99.081 98.717 98.587 99.406 98.936 586.001 99.991 99.991 99.995 99.994 99.995 99.993 99.992 99.997 99.994 665.793 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 756.449 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 859.450 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 976.475 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 1109.435 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 1260.499 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 1432.133 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 1627.136 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 1848.692 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000

126

2100.416 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 2386.415 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 2711.357 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 3080.544 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 3500.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 MEDIAN = D50 217.67 221.1 150.76 151.5 141.63 187.4 159.4 40.61 165.4 (μm)

127

), ),

Average Average

2.26E

1.49E

Geometric

Method 2

Slope(S2avg)

Average Average

2.13E

1.38E

Geometric

Method 1

Slope(S1avg)

(S2)

1.71E

1.53E

4.41E

7.36E

7.74E

1.37E

2.79E

3.41E

Paleoslope

8.55

1.04

8.10

6.78

7.51

7.45

10.32

13.23

12.92

dominant rivers recorded in in dominantrecorded rivers

Particle

Reynolds

load

Number(Rep)

1.67E

1.48E

1.58E

2.01E

1.31E

1.46E

1.45E

2.56E

2.50E

Settling Settling

Velocity Velocity

(ws)[m/s]

create Figure 28. Data from Zaklicki (2020) are not not are (2020) Data 28. from createFigure Zaklicki

2.87E

2.00E

2.43E

5.04E

1.40E

1.92E

1.88E

9.70E

1.04E+00

Method 2

Settling Settling

Velocity (W*) Velocity

Dimensionless

1.60E

5.42E

6.14E

2.98E

8.53E

7.16E

7.26E

1.32E

3.70E+00

(log W*) (log

Dimensionless

Settling Velocity Settling

paleoslope of sandy sandy of suspended paleoslope

1.08

73.18

65.57

45.99

56.33

55.46

106.46

174.95

166.94

(D*)

ParticleSize

Dimensionless

paleoslope values are used to to used are paleoslopevalues

(S1)

1.71E

1.64E

3.43E

1.02E

9.26E

1.66E

1.59E

2.00E

Paleoslope

Method 1

1.6

0.9

2.3

2.7

1.5

2.3

1.8

Bar

Height

(m) (Hbf)

2 calculations for for estimating calculations 2

est sandstones (based on Lynds et al., 2014). Table includes bar heights, median grain sizes sizes heights, includes (D grain Lynds on median bar Table 2014). al., et (based sandstones est

40.61

221.1

165.35

159.41

187.36

141.63

151.54

150.76

217.67

Raw Data Raw

Median

Grain Size

(μm) (D50)

351

339

332

329

443

442

436

434

441

(m)

Strat

Level Level

Method 1 and 1 Method

Bar

Deposit Deposit

SampleID

BLBS

BLBW

South

West

Blossburg Blossburg Blossburg

Table BlossburgSouthW and heights. 2 and 1 Method stratigraphic and included.