Tectonic controls on alluvial architecture in the upper Cretaceous John Henry Member, Straight Cliffs Formation, Southern Utah

A detailed understanding of spatial and temporal changes in alluvial architecture is essential in deciphering the mechanisms controlling stratigraphic trends, and their relative influence throughout a sedimentary basin. In this study, nonmarine exposures of the John Henry Member (Upper Cretaceous Straight Cliffs Formation) are characterized by measured stratigraphic sections, paleocurrent measurements, lateral facies mapping, and channel belt dimension measurements along a three km wide by ~200 m tall outcrop in Bull Canyon of the southwest Kaiparowits Plateau, Utah. Seven depositional units (DU) are interpreted, which define intervals with distinct alluvial architecture. The DUs are grouped into three stratigraphic intervals, each of which represents a large-scale trend in alluvial architecture. The lower stratigraphic interval includes a trend of upward decreasing grain size and channel belt width, thickness, and amalgamation, with isolated tidal influence. The middle stratigraphic interval is composed of consistently thick floodplain and coal mire deposits with laterally restricted channel belts. The upper stratigraphic interval includes vertical trends of increasing grain size, channel belt widths and amalgamation. These data are combined with previously published sections to produce a ~60 km long, dip-oriented correlation of the stratigraphic intervals, which is used to investigate variability within the depositional system. Paleomorphodynamic parameters are calculated from measurements made in fluvial strata at Bull Canyon and Rock House Cove, located ~20 km to the west (paleo-landward). The regional stratigraphy and paleomorphodynamic parameters indicate that the impact of tectonic activity is the primary driver of preserved alluvial architecture. Although some effect of sea level change, climate, and fluvial autogenic processes may be identifiable, they exert lower order control on the preserved system. In the proposed depositional model, changes in alluvial architecture represent the dynamic responses of fluvial fan systems (or distributive fluvial systems) to changes in accommodation and sediment supply resulting from episodic tectonic activity. This study contributes to the evolving paradigm that is transiting from an emphasis on downstream controls on fluvial systems (i.e., accommodation due to relative sea level changes) to more accurately account for hinterland controls (i.e., the interaction of accommodation and sediment supply due factors such as tectonic activity). These concepts are more broadly applicable to the understanding of depositional cyclicity in terrestrial (fluvial) deposits and are particularly relevant to foreland basin systems, which are common in the Cretaceous strata of western North America. iv

TECTONIC CONTROLS ON ALLUVIAL ARCHITECTURE IN THE UPPER CRETACEOUS JOHN HENRY MEMBER, STRAIGHT CLIFFS FORMATION, SOUTHERN UTAH by Luke Andrew Pettinga A thesis submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Master of Science in Geology Department of Geology and Geophysics The University of Utah May 2013 Copyright © Luke Andrew Pettinga 2013 All Rights Reserved The Uni v e r s i t y of Utah Graduat e School STATEMENT OF THESIS APPROVAL The thesis of _____________________ Luke Andrew Pettinga__________________ has been approved by the following supervisory committee members: Cari L. Johnson , Chair 11/15/2012 Date Approved Lauren P. Birgenheier , Member 11/19/2012 Date Approved Lisa E. Stright , Member 11/16/2012 Date Approved and by _____________________D. Kip Solomon_____________________ , Chair of the Department of __________________ Geology and Geophysics_______________ and by Donna M. White, Interim Dean of The Graduate School. ABSTRACT A detailed understanding of spatial and temporal changes in alluvial architecture is essential in deciphering the mechanisms controlling stratigraphic trends, and their relative influence throughout a sedimentary basin. In this study, nonmarine exposures of the John Henry Member (Upper Cretaceous Straight Cliffs Formation) are characterized by measured stratigraphic sections, paleocurrent measurements, lateral facies mapping, and channel belt dimension measurements along a three km wide by ~200 m tall outcrop in Bull Canyon of the southwest Kaiparowits Plateau, Utah. Seven depositional units (DU) are interpreted, which define intervals with distinct alluvial architecture. The DUs are grouped into three stratigraphic intervals, each of which represents a large-scale trend in alluvial architecture. The lower stratigraphic interval includes a trend of upward decreasing grain size and channel belt width, thickness, and amalgamation, with isolated tidal influence. The middle stratigraphic interval is composed of consistently thick floodplain and coal mire deposits with laterally restricted channel belts. The upper stratigraphic interval includes vertical trends of increasing grain size, channel belt widths and amalgamation. These data are combined with previously published sections to produce a ~60 km long, dip-oriented correlation of the stratigraphic intervals, which is used to investigate variability within the depositional system. Paleomorphodynamic parameters are calculated from measurements made in fluvial strata at Bull Canyon and Rock House Cove, located ~20 km to the west (paleo-landward). The regional stratigraphy and paleomorphodynamic parameters indicate that the impact of tectonic activity is the primary driver of preserved alluvial architecture. Although some effect of sea level change, climate, and fluvial autogenic processes may be identifiable, they exert lower order control on the preserved system. In the proposed depositional model, changes in alluvial architecture represent the dynamic responses of fluvial fan systems (or distributive fluvial systems) to changes in accommodation and sediment supply resulting from episodic tectonic activity. This study contributes to the evolving paradigm that is transiting from an emphasis on downstream controls on fluvial systems (i.e., accommodation due to relative sea level changes) to more accurately account for hinterland controls (i.e., the interaction of accommodation and sediment supply due factors such as tectonic activity). These concepts are more broadly applicable to the understanding of depositional cyclicity in terrestrial (fluvial) deposits and are particularly relevant to foreland basin systems, which are common in the Cretaceous strata of western North America. iv ABSTRACT........................................................................................................................... Ill ACKNOWLEDGEMENTS..................................................................................................vll INTRODUCTION...................................................................................................................1 REGIONAL GEOLOGY........................................................................................................ 4 Tectonic and Paleogeographic Setting...................................................................... 4 Stratigraphy.................................................................................................................5 Genetic Stratigraphy...................................................................................................6 RESEARCH GOALS........................................................................................................... 22 METHODS...........................................................................................................................24 DESCRIPTIVE CLASSIFICATION SCHEME................................................................ 45 Fluvial Terminology.................................................................................................45 Channel Classification..............................................................................................46 Descriptive Hierarchy...............................................................................................46 RESULTS............................................................................................................................. 54 Llthofacles..................................................................................................................54 Facles Associations...................................................................................................55 Channel Belt Dimensions and Stacking Patterns................................................... 61 Depositional Units and Bounding Members...........................................................63 Paleomorphodynamics..............................................................................................76 Sedimentary Petrology..............................................................................................81 INTERPRETATION AND DISCUSSION...................................................................... 130 Stratigraphic Intervals............................................................................................ 130 Correlation of Fluvial to Parallc Facles ...............................................................132 Relationship Between Fluvial Architecture and Shoreline Successions............ 134 Controls on Fluvial Sequences...............................................................................136 TABLE OF CONTENTS Depositional Model.................................................................................................139 CONCLUSIONS ............................................................................................................... 160 APPENDICES A. MEASURED STRATIGRAPHIC SECTIONS................................................... 163 B. INTERPRETED PHOTOMOSAICS................................................................... 188 C. PALEOCURRENT AND ACCRETION DATA................................................ 216 D. PALEOMORPHODYNAMIC SPREADSHEET................................................ 242 E. SAMPLES LIST.....................................................................................................243 F. PETROGRAPHIC DATA..................................................................................... 246 G. GPS DATASET ...................................................................................................... 248 REFERENCES....................................................................................................................278 vi ACKNOWLEDGEMENTS I would like to express gratitude to my advisor, Dr. Cari Johnson, for guidance throughout this project. Thanks to Dr. Lisa Stright and Dr. Lauren Birgenheier for agreeing to participate as members of my thesis committee. Thanks to everyone who assisted me in the field: Patrick Dooling, Ian Semple, Alex Turner, Tyler Szwarc, and Brenton Chentnik. This project was made possible by funding through donors to the American Chemical Society - Petroleum Research Fund (grant to Cari Johnson), and support through University of Utah Rocks to Models research group from Chevron, ConocoPhillips, Hess, Shell, and Statoil. Thanks to ConocoPhillips for providing a research fellowship for my studies. Thanks to the Bureau of Land Management and Grand Staircase-Escalante National Monument for granting me permission to conduct research on federal land. I am extremely grateful to all of my family, friends, and colleagues who have helped support and guide me through this process. INTRODUCTION Knowledge of the spatial and temporal changes in terrigenous-clastic deposits within foreland basin systems can yield insight into the mechanisms controlling deposition (Paola et al., 1992; Dreyer et al., 1993; Blum and Tornqvist, 2000; Holbrook et al., 2006) and has implications for subsurface reservoir characterization (Tye, 2004; Pranter and Sommer, 2011). Alluvial architecture can be influenced by autogenic controls, like avulsion, progradation, and scouring (Mackey and Bridge, 1995; Sheets et al., 2007; Hajek et al., 2010), and allogenic controls, such as relative sea level, climate, hinterland tectonics, and basin subsidence (Fisk, 1944; Ouchi, 1985; Shanley and McCabe, 1994; Holbrook, 2006; Hampson etal., 2012). Large-scale alluvial architecture is the product of some combination of autogenic and allogenic controls, and their complex interaction. Many of the current predictive models of alluvial architecture and geometry focus on variations in downstream base level (commonly sea level) as the primary allogeneic control (Shanley and McCabe, 1993; Wright and Marriott, 1993; Richards, 1996; Currie, 1997; van Heijst and Postma, 2001). These models likely have relevance in the lower reaches of fluvial systems, but do not account for variations in allogenic controls that may exist within a basin (Slingerland and Snow, 1988; Mackey and Bridge, 1995; Blum and Tornqvist, 2000; Bridge, 2003; Holbrook et al., 2006). Sedimentary cycles present in the upper reaches of fluvial systems are more likely to be influenced by factors such as tectonics, climate, and autogenic processes than downstream base level (Olsen et al., 1995, Legarreta and Uliana, 1998; Kjemperud et al., 2008; Hampson et al., 2012). Quantitative modeling and studies of modern and recent (Pleistocene) fluvial deposits have emphasized some of the complexity within the interaction of upstream and downstream controls (Blum and Aslan, 2006; Martin et al., 2010; Gibling et al., 2011). The datasets from numerous large-scale, outcrop-based studies are required to account for the spatial variability of these controls on fluvial systems preserved in the sedimentary record; however, relatively few studies of this scope exist (Hirst, 1991; Dreyer et al., 1993; Shanley & McCabe, 1993; Burns et al., 1997; Currie, 1997; Holbrook et al., 2006; Corbett et al., 2011; Hampson et al., 2012). This study incorporates what we are beginning to understand about the modern depositional systems in order to re-evaluate the ancient; specifically, how fluvial depositional models should be revised. The John Henry Member of the Straight Cliffs Formation contains one of several cyclical stratigraphic sequences present within the Upper Cretaceous strata of the Kaiparowits Plateau, southern Utah (Peterson, 1969a; Ryer, 1993). The John Henry Member in this area is particularly well suited for a detailed study of alluvial architecture and basin scale controlling mechanisms because of nearly continuous exposures from fluvial strata in the west to marginal marine strata in the east. A commonly cited predictive model of alluvial architecture was based on sequence stratigraphic interpretations of outcrops of the John Henry Member (Shanley, 1991; Shanley and McCabe, 1991, 1993). This model focuses on downstream relative sea level controls on deposition. Some aspects of these interpretations have been called into question by several more recent studies (Little, 1995; Gallin et al., 2010; Gooley, 2010; Allen and 2 Johnson, 2011). In order to resolve the primary controls on alluvial architecture, the spatial and temporal variability present within the John Henry Member must be better understood. This study builds on previous research in advancing toward a comprehensive understanding of the relationship between coeval fluvial and marginal marine strata, and the mechanisms that controlled deposition within the Kaiparowits Basin. The specific goals of this study are to investigate primary controls in deposition by documenting the alluvial architecture and paleomorphodynamic characteristics in a previously unstudied exposure of the John Henry Member, creating a dip oriented correlation from fluvial to marine strata, and refining the depositional model. The results of this study have implications for a number of terrigenous-clastic depositional systems including analogous foreland basin and passive margin systems. 3 REGIONAL GEOLOGY Tectonic and Paleogeographic Setting The Kaiparowits Plateau of southern Utah contains expansive exposures of the Upper Cretaceous Straight Cliffs Formation (Figure 1). Located near the western margin of the Colorado Plateau physiographic province, the plateau is bounded by the Escalante River valley to the east, erosional features of the Colorado River to the south, and the Cockscomb (the physiographic expression of the East Kaibab Monocline) to the west. The strata of the plateau dip gently to the north and are locally deformed by a series of north-south trending anticlines, synclines, and normal faults (Peterson, 1969a; Doelling et al., 2000). In the Late Cretaceous, isostatic adjustment due to the east-west shortening along the Sevier orogenic belt and dynamic subsidence associated with subduction of the Farallan plate led to the development of a foreland basin system along the western edge of Western Interior Seaway (Figure 2) (Kauffman, 1977; Kauffman and Caldwell, 1993; DeCelles and Coogan, 2006; Liu et al., 2011). The Upper Cretaceous strata of the Kaiparowits Plateau are composed of siliciclastic sediments that were transported eastward from the Sevier orogenic belt and northeastward from the Mogollon Highlands (Peterson, 1969a, 1969b; Eaton, 1991; Hettinger et al., 1993; Lawton et al., 2003). The paleoshoreline in this area was generally oriented northwest-southeast, and paleocurrent indicators and facies transitions indicate that the fluvial systems mainly flowed to the east-northeast (Peterson, 1969a; Ryer, 1977, 1993; Bobb, 1991; Shanley and McCabe, 1991, 1995; Roberts and Kirschbaum, 1995; Christensen, 2002; Lawton et al., 2003). Stratigraphy Gregory and Moore (1931) conducted an initial geologic investigation of the Kaiparowits Plateau. These authors identified a thick, sandstone dominated section overlying the Tropic Shale, which they named the Straight Cliffs Sandstone after the large marine sandstone exposures along the Straight Cliffs escarpment (i.e., Fifty Mile Mountain) south of Escalante, Utah. Peterson and Waldrop (1965) redefined it as the Straight Cliffs Formation after finding that away from the type area it was composed of interbedded sandstone, smectitic mudstone, carbonaceous mudstone, and coal. They also proposed three informal members (Lower, Middle, and Upper). Peterson (1969a, 1969b) conducted the first detailed stratigraphic correlation of the Straight Cliffs Formation on the Kaiparowits Plateau. This work along the southern margin of the plateau led to the division of the formation into four members that were formally named, in ascending order: Tibbet Canyon, Smoky Hollow, John Henry, and Drip Tank (Peterson, 1969b). Additionally, seven shoreface units, named A through G, and four coal zones, name Lower, Christenson, Rees, and Alvey, were identified within the John Henry Member (Figure 3) (Peterson, 1969a). The John Henry Member is the thickest member of the Straight Cliffs Formation and preserves a fluvial-marine transition; with continental deposits in the southwest that grade into paralic and marine deposits to the northeast (Peterson, 1969a, 1969b). Preservation of these transitional facies makes the plateau an ideal location to investigate 5 the relationship between marine and terrestrial sequence stratigraphy. The John Henry Member is believed to have been deposited over ~6.2 my from the Early Coniacian to Late Santonian (89.8 to 83.6 Ma) (Eaton 1987, 1991; Eaton and Nations, 1991; Kauffman et al., 1993; Gradstein et al., 2012); however, the dates are primarily based on biozones identified within the marine strata, which are poorly constrained. The presence of the inoceramid Volviceramus involutus at the base of the John Henry Member indicates that deposition began in the Early Coniacian (Eaton, 1987, 1991). The upper part of the John Henry Member is assumed to be no younger than Late Santonian based on the assemblage of inoceramid bivalves and the presence of the ammonite Desmoscaphites (Eaton 1987, 1991; Peterson, 1969a; Kauffman et al., 1993). A number of early studies focused on the distribution and nature of coal-bearing facies (Robison 1963, 1964, 1966; Doelling, 1967, 1968; Doelling and Graham, 1972). More recently, studies by Vaninetti (1979), Hettinger (1995, 2000), and Hettinger et al. (1996) investigated the distribution and westward thinning of coal zones in the southwestern portion of the plateau. Vaninetti (1979) investigated the relationship between fluvial channel systems and peat-forming (swamp) conditions along the southern border of the plateau. Hettinger (1995, 2000) and Hettinger et al. (1996) used outcrop, cored wells, and well logs to document the distribution of coal throughout the plateau and its relationship with shoreline facies. Genetic Stratigraphy Shanley (1991) and Shanley and McCabe (1991, 1993) developed the first sequence stratigraphic framework for the Straight Cliffs Formation, which included four unconformity-bound sequences. The sequence boundaries were identified by a distinct 6 basinward shift in facies and included (Figure 3): (1) Tibbet Canyon sequence boundary, located at the top of the Tibbet Canyon Member; (2) Calico sequence boundary, located near the top of the Smoky Hollow Member at the base of the Calico bed; (3) the "A-sequence boundary," in the lower John Henry Member near the top of the A-shoreface; and (4) Drip Tank sequence Boundary, located at the base of the Drip Tank Member. Shanley and McCabe (1991) attempted to trace the sequence stratigraphic surfaces from marine into terrestrial strata. For example, the "A-sequence boundary" was traced from the top of the A-shoreface into fluvial strata of the lower John Henry Member, where the unconformity is represented by the transition to amalgamated sandstones of incised valley fill and overlying estuarine and tidally-influenced fluvial strata. Subsequent studies have also reported incised valley candidates, which may be associated with the "A-sequence boundary" (Hettinger et al., 1993; Stephen and Dalrymple, 2002). Correlation of sequence stratigraphic surfaces into terrestrial deposits resulted in development of a frequently-cited sequence stratigraphic model of alluvial architecture (Figure 4a) (Shanley and McCabe, 1991, 1993, 1994, 1995). This model identifies four stages based on distinctive patterns of alluvial architecture, each of which has been tied to changes in base level: (1) early lowstand systems tract, erosion and incised valleys development during base level fall; (2) late lowstand systems tract, coarse-grained, vertically and laterally amalgamated braided fluvial channels that backfilled the valleys during slow base level rise; (3) transgressive systems tract, heterolithic strata and tidally-influenced fluvial channel-fill deposited during marine transgression; and (4) highstand 7 systems tract, fine-grained, isolated meandering fluvial channels in thick floodplain alluvium deposited during high rates of base level rise. Little (1995, 1997) investigated Upper Cretaceous strata of the Kaiparowits Plateau, and speculated on the relative role of tectonic and eustatic controls in the development of the cyclical fluvial depositional patterns. Cyclicity within alluvial stratigraphy was interpreted using the concept of thrust propagation (Figure 4b), a modified version of the two-phase tectonic deposition proposed by Heller et al. (1988). In this interpretation, each cycle begins with the onset of active tectonism in the Sevier fold and thrust belt causing subsidence and highly aggradational fluvial systems with thick floodplain deposits. Subsequent decreasing tectonism, and a corresponding decrease in subsidence, allow the basin to fill, producing an upward-coarsening fluvial succession. Tectonic quiescence results in the formation of a coarse-grained sheet, which caps each cycle. These sheets develop due to the combination of denudation and isostatic rebound that cause the redistribution of the coarse sediments that had been deposited adjacent to the thrust belts during initial rapid subsidence. Little (1997) concluded that hinterland tectonics was the main factor controlling the depositional architecture with the basin, but acknowledged that eustatic sea level may have contributed to depositional cycles during the early stages of basin development, including deposition of the Straight Cliffs Formation. Conditions typical of the early stages of foreland basin development, including low-lying sediment source area and a relatively wide and shallow basin, may have allowed eustatic sea level to influence alluvial architecture during deposition of the John Henry Member. However, in later stages of the foredeep the eustatic signal was completely overpowered by tectonic control due to continued eastward movement of the Sevier 8 thrust belt. Similar cyclicity in fluvial deposits has been interpreted as tectonically driven (Dreyer et al., 1993; Legarreta and Uliana, 1998; Horton and DeCelles, 2001; Weissmann et al., 2010). Several studies have noted that the coarse clastic sheets within the Straight Cliffs Formation (Calico bed and Drip Tank Member) are conformable and interfinger with underlying strata (Peterson, 1969b; Vaninetti, 1979; Eaton, 1991; Tilton, 1991; Little, 1995; Lawton et al. 2003). However, Shanley and McCabe (1995) described the bases of these clastic sheets as sharp erosional surfaces, which were used as evidence that they represent sequence boundaries. Lawton et al. (2003) described the base of the Drip Tank Member to be a gradational contact, but still concluded that a drop in sea level caused the formation of a sequence boundary. Allen (2009) and Allen and Johnson (2010) investigated the stratigraphy of marine deposits in the John Henry Member in Rogers Canyon. These studies advocate the use of transgressive-regressive cycles (T-R cycles) to describe the marine stratigraphic architecture. Seven T-R cycles, named 0 through 6, were identified to form three large-scale trends: net progradation in cycles 0 through 2; net retrogradation in cycles 3 through 5; and net aggradation in cycles 5 to 6 (Figure 3). Allen (2009) did not identify any sequence boundaries in this part of the plateau, but noted that the transition between the B and C sandstones records the greatest basinward shift in facies and is the best candidate for a sequence boundary. Gallin (2010) and Gallin et al. (2010) investigated paralic facies within the John Henry Member of the south-central Kaiparowits Plateau and their correlation with marine strata to the east. Gallin (2010) identified three facies associations within the John Henry 9 Member, in ascending order (Figure 5b): Facies association 1 includes tidally influenced, laterally restricted fluvial channel belts, coastal mires and shoreface sandstones, which correlate with the lower, net progradational marine packages. Facies association 2 includes laterally restricted, highly sinuous fluvial channel belts, lagoonal and estuarine coastal plain mires, bay-head deltas, isolated distributary and tidal channels, which correlate with the middle, net retrogradational marine packages. Facies association 3 includes laterally extensive, low sinuosity fluvial channel belts, and vertically amalgamated fluvial channel belt complexes, and floodplain deposits, which correlate with the upper, net progradational marine packages. Gallin et al. (2010) noted the absence of an unconformable surface that would correlate with the A-sequence boundary. Rather, the section was characterized by fairly consistent vertical trends of increasing grain size, channel belt amalgamation, and downstream accretion, and a vertical trend of decreasing tidal influence. In the southwestern corner of the Kaiparowits Plateau, Gooley (2010) divided the proximal fluvial strata of John Henry Member into seven depositional units, named DU-0 through DU-6 (Figure 5a), in ascending order: DU-0 is composed of tidally-influenced channel belts. DU-1 overlies an unconformable surface and is composed of highly amalgamated, laterally accreting channel belts with tidally-influenced channel belts in the upper section. DU-2 is composed of laterally extensive, laterally accreting channel belts. DU-3 is composed of isolated channel belts embedded within thick floodplain muds. DU- 4 through DU-6 grade from laterally restricted laterally accreting channel belts with low vertical amalgamation, to highly vertically amalgamating downstream accreting channels. It is important to note that the use of the term depositional units by Gooley (2010) and the 10 current study is equivalent to the use of facies associations in Gallin (2010). Gooley (2010) documented two characteristic, large-scale trends within the John Henry Member: First, there is an upward decrease in grain size, decrease in channel belt frequency and lateral extent, and episodic increase in tidal influence in DU-1 through DU-3. Second, there is an upward increase in grain size, increased channel belt frequency and lateral extent, and a shift from lateral to downstream accretion in DU-3 through DU-6 and continues into the Drip Tank Member. Additionally, Gooley (2010) correlated Rock House Cove depositional units and stratigraphic trends across the plateau, incorporating the measured sections and interpretations of previous studies (Vaninetti, 1979; Allen, 2009, Gallen, 2010). One conclusion from this work was that additional inquiry was required to adequately document the stratigraphic changes between the Rock House Cove and Kelly Grade areas. Bull Canyon, the focus of this study, is located approximately half way between these two locations. The Straight Cliffs Formation has been used in outcrop analogue studies for subsurface reservoirs. Two studies used exposures of the lower John Henry Member in the Blue Cove area as a reservoir outcrop analog to calculate statistics and construct reservoir models (Dalrymple, 2001; Stephen and Dalrymple, 2002). Gooley (2010) evaluated the reservoir potential of nonmarine outcrops in Rock House Cove. 11 12 Figure 1 - Map of the Kaiparowits Plateau showing the locations of measured stratigraphic sections and logged cores from this study and those of relevant previous studies. The Kaiparowits Plateau is shaded in light gray. Outcrops of the members of the Straight Cliffs Formation are shown (modified from Doelling and Willis, 1999, 2006). The transect A-A'-A" marks the locations used to construct stratigraphic correlations through the plateau (Figures 37 and 38). The inset map of Utah shows the location of the Kaiparowits Plateau and the approximate locations of the Sevier orogenic belt and the Western Interior Seaway during the middle Santonian, ~85 Ma (DeCelles, 2004). 13 14 Figure 2 - Paleogeography of southwestern North America in Coniacian-Santonian time (modified from Roberts and Kirschbaum, 1995). The location of Kaiparowits Plateau (KP) is outlined in black. 15 Highlands Mires (coal beds) Alluvial plain Offshore marine Coastal plain Fold and thrust front 16 Figure 3 - Upper Cretaceous stratigraphy of south-central Utah. A) Regional stratigraphy. B) Lithostratigraphy of the Straight Cliffs Formation showing the fluvial to marginal marine strata. The coastal onlap curve of Allen and Johnson (2010) and sequence stratigraphic interpretation of Shanley and McCabe (1991) are shown alongside the lithostratigraphic divisions for locations along a SW-NE transect of the Kaiparowits Plateau: Rock House Cove (Gooley, 2010), Bull Canyon (this study), Kelly Grade (Gallin, 2010), and Rogers Canyon (Peterson, 1969b; Allen, 2009). The labels on the left side of the Rogers canyon section are the shoreface sandstone names from Peterson (1969b) and on the right are the T-R cycle labels from Allen (2009). A. Regional Stratigraphy B. Straight Cliffs Formation, Kaiparowits Plateau Age Kaiparowits Plateau Campanian Kaiparowits Fm Wahweap Fm Straight Cliffs Fm. Drip Tank Mbr 1 1 Cenomanian TuronianConiacian Santonian John Henry Mbr Smoky Hollow Tibbet Canyon Tropic Shale Dakota Fm 18 Figure 4 - Predictive models for alluvial stratigraphy. A) Eustatic model of alluvial stratigraphy (Shanley and McCabe, 1993). B) Tectonic model of alluvial stratigraphy (Little, 1997). Both models are based on observations made in Kaiparowits basin strata. A. Eustatic Model (Shanley and McCabe, 1993) Hiahstand I Isolated, high (sinuosity fluvial Ichannels Transqressive , " ■.Tidally-influenced fluvial deposits Amalgamated fluvial channel deposits Valley incision and terrace formation Low-sinuosity high gradient rivers B. Tectonic Model (Little, 1997) k - ^ Redistributed Clastic Sheet r&ScrA Gravelly braided river X i^ W w X ^ W ^ X \\V n S W - g g g y »>^gai^5»arjr& tfispM Slow Subsidence Sandy braided river Slowing Subsidence Meandering streams Rapid Subsidence Anastomosing and meandering streams Early Slow Subsidence Meandering rivers Redistributed Clastic Sheet Gravelly braided river Isolated Channels and Channel Belts Channel Belts Vertically Amalgamated Channel Belts Inclined Heterolithic Strata / Estuarine Floodplain Gravelly Downstream Accreting Channels 20 Figure 5 - Simplified stratigraphic diagrams of the John Henry Member at (A) Kelly Grade (Gallin, 2010) and (B) Rock House Cove (Gooley, 2010). The locations of these study areas are shown on Figure 1. A) Rock House Cove B) Kelly Grade DU-6 DU-1 DU-0 DTM <D -Q FA-3 (D ^ --------- C (D x FA-2 - FA-1 SHM Floodplain Coarse-grained braided nver Coal / coal mire Isolated channels Channel belts Vertically amalgamated channel belts Inclined heterolithic strata / estuarine O Paleocurrent direction ------ Member contact ------ Unconformity RESEARCH GOALS Preservation of the fluvial-marine transition within the John Henry Member in the Kaiparowits Plateau makes it an ideal case study to investigate an ancient sediment-routing system (i.e., the transfer zone) (Castelltort and Van Den Driessche, 2003, Goodbred, 2003; Armitage et al., 2011). In this system sediment shed from the Sevier thrust belt was transported through the coastal plain, with some degree of storage occurring enroute, and ultimately deposited along the shore of the Western Interior Seaway. Unlike depositional systems connecting to ocean basins, the shorelines may be the ultimate sink for much of the coarser sediment (i.e., sand sized sediment) in this system, particularly due to the limited presence of this material in age-equivalent deepwater turbidite systems to the east. Within the current body of research involving the interpretation, stratigraphic correlation, and depositional models of the John Henry Member, there is a relative lack of detailed inquiry into the nature of the transition from purely fluvial deposition on the west side of the plateau to paralic deposition in the central portion of the plateau and marine successions to the east. This study attempts to provide insight into many of the questions associated with this gap in the current literature by conducting a detailed stratigraphic study of the Bull Canyon region, located approximately halfway between the previously studied areas of proximal fluvial strata at Rock House Cove and fluvial/paralic strata at Kelly Grade (Figure 1). The outstanding questions include: What is the nature of the fluvial architecture in this portion of the system, and how do the trends in fluvial architecture correlate with previously studied areas throughout the plateau? Do the trends in fluvial architecture fit the existing depositional and sequence stratigraphic models, and if not, what modifications should be made to these models? What implications does this area have on changes in geographic and temporal shift in the depositional environment (e.g., changes in provenance)? What are the broader implications of this study for depositional cycles present in the terrestrial (fluvial) deposits? 23 METHODS The stratigraphy of Bull Canyon was initially characterized using a series of logged stratigraphic sections measured along the east side of the canyon (Figures 6 and 7; Appendix A). Two of the measured sections (SXN1 and SXN4) cumulatively covered the entire section exposed at Bull Canyon and were used to delineate the Tibbet Canyon, Smoky Hollow, John Henry, and Drip Tank Members. A total of three sections (SXN1, SXN2, and SXN3) were measured through the John Henry Member along the eastern side of the canyon at ~0.5 km spacing (Figure 6). Each section began at the lowest exposure of the John Henry Member and extended the top of the canyon. Stratigraphic sections were described using the methods outlined in Miall (2000), which include documenting lithology, grain size, sedimentary structures, and body and trace fossils. While measuring sections, straight ‘vertical' paths were maintained whenever possible; however, offsets were occasionally necessary due to topography and outcrop accessibility. The use of trenching along sloped portions of the outcrop allowed for nearly continuous section documentation. The measured sections were used to determine the percent distribution of lithofacies within each depositional unit. These distributions were used to determine the ratio of net sandstone to gross sediment within each depositional unit (i.e., net-to-gross ratio). Two high-resolution composite photographs (Gigapans: Appendix B) of the east side of Bull Canyon, which cover approximately 2.5 km of lateral outcrop, were used to map facies distributions and channel stacking patterns in the John Henry Member (Figure 8a,b). A composite photo of the west side of Bull Canyon was used to confirm observations made on the east side of the canyon (Figure 8c). An additional composite photograph was used to document lower portion of the canyon, which was poorly represented in the previously mentioned composite photographs (Figure 9). Paleocurrent and barform accretion measurements were taken along sandstone beds from the base of the John Henry Member through the Drip Tank Member in order to document the paleotransport direction and mode of accretion (Appendix C). Paleocurrent measurements were taken on trough cross-stratification, ripple cross-lamination, and tabular cross sets (Allen, 1966; Selley, 1968; Miall, 1974). The orientation of barform accretion sets were measured and evaluated in relation to the paleocurrent direction to identify the dominant mode of accretion (i.e., lateral or downstream accretion) for the seven depositional units using an approach similar to McLaurin and Steel (2007). Rose diagrams of accretion set directions were each overlain with the corresponding mean paleocurrent direction, a blue colored wedge marking orientations within 45° of the mean paleocurrent direction, and two yellow wedges marking orientations within 45° of being perpendicular to paleocurrent (Figure 10). Accretion sets measurements that fall within the yellow wedges are considered to be generally lateral accreting and those with the blue wedges are considered to be generally downstream accreting. The dimensions and distribution of sandstone bodies were documented using a differential GPS unit paired with a laser rangefinder. A series of base stations were established on the west side of Bull Canyon, and their locations were recorded with sub­meter accuracy (99% of points with 5-15 cm accuracy) using a Trimble® GeoExplorer® 25 6000 GPS unit equipped with a Tempest™ antenna. The locations of sandstone bodies were documented using a series of points that were recorded as offsets from the base station using a TruPulse 306b rangefinder. The most effective method for documenting the sandstone bodies was to collect points at their furthest lateral extents to the left and right and a series of paired points to record their thickness, which consisted of vertically aligned measurements of the base and top of the bed. The paired measurements were generally evenly spaced at ~10 m or less, but additional points were collected where changes in thickness or significant topographic relief along the cliff face. Differential corrections were made with GPS Pathfinder® Office software using the reference station in Page, Arizona. Petrel™ software (by Schlumberger) was used to display the data in 3D, which allowed for the thickness, apparent width and orientation of the sandstone bodies to be measured (Figure 11; Appendix G). The apparent channel belt widths were corrected to true channel widths using the methods of Fabuel-Rerez et al. (2009). In this method, the paleocurrent, outcrop orientation, and apparent channel width are used to trigonometrically solve for true (cross channel) widths (Figure 12). Fluvial morphodynamics relate to the coevolution of flow hydrodynamics and surface morphology (Paola et al., 2009; Kleinhans, 2010). Various methods can be used to calculate paleomorphodynamic parameters using measurements of bankfull depths and the d50 (median grain size) of bedload material (Paola and Mohrig, 1996; Bridge and Tye, 2000; Leclair and Bridge, 2001; Hajek and Wolinsky, 2012). These calculations were developed for, and are commonly used in, Quaternary research. In this study, a workflow was developed that incorporates a number of paleomorphodynamic parameters, 26 which was used to investigate Upper Cretaceous fluvial strata in the John Henry Member (Appendix D). The workflow was applied to each depositional unit within the John Henry Member in Bull Canyon and those identified by Gooley (2010) in Rock House Cove. The parameters that were investigated include: bankfull depth, slope, Froude number (and Froude squared), and backwater length. These parameters are explained more fully in the Results section. Thin sections were prepared from 20 medium-grained sandstone samples, which were collected from channels at roughly even spacing vertically through the section (Figure 7; Appendices E and F). A modified version of the Gazzi-Dickinson method was used to point count the thin sections (Ingersoll et al., 1984; Dickinson, 1985; Zuffa, 1985). The results were examined for trends throughout the section in Bull Canyon and compared to similar work conducted in Rock House Cove and Kelly Grade. 27 28 Figure 6 - Map of Bull Canyon with the locations of the four measured sections (SXN 1­4) and the camera locations and areas included in the three panoramic photographs (lower east Bull Canyon, upper east Bull Canyon, and west Bull Canyon). 29 ° 42' 00'' 111° 44' 00'' 1111°43'00 1111° 42' 00'' Measured Sections ------- SXN1 ------- SXN2 ------- SXN3 ------- SXN4 Areas in Panoramic Photos/Camera Locations | | • Lower East Bull Canyon Upper East Bull Canyon | | • West Bull Canyon 30 Figure 7 - Measured stratigraphic sections. A) Stratigraphic sections legend. B) Measured stratigraphic sections of the Straight Cliffs Formation from the east side of Bull Canyon. Section locations are shown on Figure 5. Blue lines mark the contacts between the Smoky Hollow Member (SHM), John Henry Member (JHM) and Drip Tank Member (DTM). Boundaries between the seven depositional units (DU-0 through DU-6) are marked with grey dashed lines. A black line marks the location of a fault with ~15 m of normal displacement. 31 A) Stratigraphic Sections Legend Flaser cross lamination Conglomerate, cast-supported Current ripple cross lamination c>o0o°o°o0o0 Conglomerate, matrix-supported Planar cross stratification Mud rip-up clast Trough cross stratification (lamination) O Lithic clast Horizontal bedding / lamination & Coal clast Hummocky / swaley cross stratification Organics Convolute beds or laminations Sharp contact J ) L Water escape structures Gradational contact G£> Concretions Erosional contact CCD Siderite □ Sandstone a Wood fragment □ Siltstone / claystone Roots ■ Carbonaceous shale Plant material ■ Coal 0a Skolithos / Planolites «-U06 Sandstone samples (abbreviated names) Depositional Unit contact Member contact 32 33 Figure 8 - Three photomosaics and interpretations of the John Henry Member in lower east Bull Canyon (A), upper east Bull Canyon (B) and west Bull Canyon (C). The field of view and camera locations for each of the photomosaics are shown in Figure 3. A) Photomosaic and Interpretation of Lower East Bull Canyon Siltstone/mudstone - Depositional unit boundary Sandstone Member boundary - Fault (approximate location) Measured sections: SXN 1 (blue), SXN 2 (purple) £ B) Photomosaic and Interpretation of Upper East Bull Canyon J Siltstone/mudstone - Depositional unit boundary ^ Sandstone - Fault . . . . . Measured - Member boundary - _VK, _ . sectioonsv:. .S .X N. 1 (v blue, ), 3 SXN 2 (purp e), SXN 3 (orange) Figure 8 continued U> C) Photomosaic and Interpretation of West Bull Canyon Siltstone/mudstone - Depositional unit boundary Sandstone - Member boundary Figure 8 continued G\ 37 Figure 9 - Photomosaic and interpretation of the Smoky Hollow Member (Calico bed) and DU-0 of the John Henry Member near the mouth of Bull Canyon facing northwest. The interpretations show the distribution of sandstone (yellow) and floodplain fines (gray). Blue lines mark formation member boundaries. Green lines mark the boundaries of depositional units, which are stratigraphic intervals that contain distinct fluvial architecture. ] Siltstone/mudstone - Depositional unit boundary ] Sandstone - Member boundary 39 Figure 10 - Documentation of paleocurrent and accretion measurements. A) A key to the paired paleocurrent and accretion diagrams that were constructed for each depositional unit (Figure 18). The rose diagrams are oriented so that north is up. The blue colored zone includes 45° on either side of the mean paleocurrent direction. Accretion measurements within the blue zone are considered to be downstream accreting. The yellow zones each included 90° areas with their centers oriented perpendicular to the mean paleocurrent. Accretion measurements within the yellow zones are considered to be laterally accreting. B) An example of dominantly downstream accreting DU-0. C) An example of dominantly lateral accretion in DU-4. 40 A. Paleocurrent and accretion diagram key Mean paleocurrent direction B. Dominant downstream accretion (ex. DU-0) PC C. Dominant lateral accretion (ex. DU-4) 41 Figure 11 - GPS point cloud and digital elevation model (DEM) of the east side of Bull Canyon displayed using Petrel™ (by Schlumberger). Northing and easting are shown in UTM coordinates (zone 12N) and elevation is shown in meters above sea level. The DEM is marked with 20 m contour lines. Groups of colored spheres represent the channel belt boundaries (left, right, and paired points from the base and top). Channel centers are displayed as white boxes. 437600 437200 Easting 436800 1700 1600 Elevation (m) 1500 1400 4120600 4120200 4119800 4119400 Northing 4119000 4118600 to 43 Figure 12 - True channel belt width correction. The diagram shows how paleocurrent direction was used to correct apparent channel belt widths (measured on outcrops) to true channel belt widths (i.e., actual width of the channel belt) (modified from Fabuel-Perez et al., 2009). 44 + -Z. DESCRIPTIVE CLASSIFICATION SCHEME Fluvial Terminology This study addresses the descriptive hierarchy of alluvial architecture and architectural elements using an approach similar to previous studies in the John Henry Member (Gallin, 2010; Gooley, 2010). Alluvial architecture, the largest scale of fluvial description, is the distribution, proportion, and geometry of alluvial deposits with a sedimentary basin (Allen, 1978). Thus, alluvial architecture relates to the long-term, large-scale aspects of alluvial deposition and erosion. Fluvial architecture, though sometimes used synonymously with alluvial architecture, more appropriately refers to smaller scale three-dimensional arrangement of architectural elements (Mackey and Bridge, 1995; Heller and Paola, 1996; Bridge, 2003; Ethridge, 2011). Architectural elements range from channel scale down to lamination scale features, which include: channels, gravel bars and bedforms, sandy bedforms, foreset macroforms, lateral accretion deposits, sediment gravity flow deposits, laminated sand sheets and overbank fines (Miall, 1985). The hierarchy of channel descriptions includes, in increasing scale, channel stories, channel belts, and channel belt complexes. Channel stories are composed of amalgamated barforms that are bound by scour surfaces and represent a single channel migration. Channel belts consist of a series of related channel stories that represent continued migration of a channel across the floodplain (Bridge, 2003). Channel belt complexes (or amalgamated complexes) are composed of a series of amalgamated channel belts that were deposited at separate points in time on the same position on the floodplain (Hirst, 1991). Channel Classification There are numerous problems associated with relating the fluvial sedimentary record to modern streams, which have been the topic of much debate (see Bridge, 2003 and authors therein). As a result, the application of the traditional classification, which uses planform geometry to identify paleo-rivers as straight, meandering, anastomosing, or braided, in outcrop studies is particularly controversial (Ethridge, 2011). In this study, the criteria for classifying channels was based on the sediment type, architectural elements, sedimentary structures, and cross-section form typical of these channel types (detailed descriptions are included in the facies association section below). Descriptive Hierarchy This study implemented a descriptive classification hierarchy that includes, in increasing scale, lithofacies, facies associations, and depositional units. This descriptive hierarchy is based on standard methods of fluvial description outlined by Miall (1996) and was modified in a similar fashion to those used by other researchers on outcrop studies in the Straight Cliffs Formation (Gallen, 2010; Gooley, 2010; Semple, 2011). Lithofacies are defined using bedding, grain size, texture, and sedimentary structures. This study employs a modified version of the lithofacies classification scheme proposed and refined by Miall (1977, 1978, 1985, 1996) that uses two or three letter facies codes in which the first capital letter indicates the dominant grain size (G = gravel, 46 S = sand, F = fine grained facies, including very fine sand, silt and mud) and the accompanying lower case letters indicate bedding, sedimentary structures, and texture. The lithofacies used in this study are listed and defined in Table 1. Facies associations are groups of genetically related facies that have an environmental significance (Collinson, 1969). The facies associations identified in this study include: coal mire, floodplain, meandering and anastomosing fluvial, braided fluvial, and tidally-influenced fluvial (Table 2). Depositional units (DU) are defined as stratigraphic intervals that possess a distinct alluvial architecture (i.e. the nature and arrangement of various facies associations) (Gooley, 2010). The depositional units identified in this study are summarized in Table 3. Transitions between depositional units are commonly gradational, and as a result, it is not always possible to place their boundaries along discrete surfaces. The depositional units are defined in Bull Canyon, but they are more broadly applicable; some can be correlated more than 40 km across the surrounding region of the southern Kaiparowits Plateau. 47 Table 1 Descriptions of Lithofacies Facies Code Facies Description Interpretation Gem, Get Gmm, Gmt Smrc Sm St Clast-supported conglomerate, massive and trough cross-stratified Matrix-supported conglomerate, massive and trough cross-stratified Mud-rip up clast conglomerate Massive sandstone Trough cross­bedded sandstone Structureless and trough cross-stratified matrix supported conglomerate; the grain size of the sandstone matrix was fine to very course and clasts varied from 0.2 to >4 cm; sometimes graded; typically 10 to 50 cm thick. Gravel bedforms, lag deposits Subangular to well rounded pebbles and gravel supported in a sandstone Barforms and channel matrix; grain size of sandstone matrix grain size varied from fine to very bedforms course and clasts varied from 0.2 to >4 cm; typically 10 to 30 cm thick. Angular to subrounded, granule to pebble size clasts of mud and silt in a sandstone matrix; commonly located at the base of channels and barforms; coal and wood fragments are often present; typically 15 to 30 cm thick. Very fine to course sand that lacks distinguishable sedimentary structures, convolute bedding, and bioturbation. Trough sets ranging from 0.05 to 0.5 m tick and cosets up to 4 m thick; grain size varied from fine to very course. Initial deposition during a flooding event High sedimentation rate Migration of sinuous-crested and linguoid dunes (3D bedforms) OO Table 1 Continued Facies Code Facies Descriptions Interpretation Sp Sc Sb Sr Sh Sf Fm, FI Planar cross-bedded Planar sets are generally 0.05 to 0.15 m thick; grain size varied from sandstone fine to very course. Convoluted sandstone Bioturbated sandstone Distorted bedding within fine to medium grain sandstones; features included dewatering, oversteepened, and erratic structures. Fine- to medium- grain sandstone bodies in which bioturbation has modified the primary depositional structures; ichnogenera include Skolithos, Rhizocorallium, Teredolites, Thalassinoides, and Planolites. Ripple cross- Current ripples, including climbing ripples, in very fine to medium laminated sandstone sandstone; less than 5 cm in height. Horizontally Horizontally laminated fine to medium grain sandstone dipping <5C laminated sandstone Migration of transverse and liguoid dunes (2D bedforms) High sedimentation rate or slumping Biogenic activity Flaser bedded sandstone Massive and laminated siltstone/mudstone Fine to medium grain sandstone crossbeds draped with carbonaceous mud or micas. Deposition during lower flow regime flow Deposition during critical flow Deposits representing fluctuations in flow Structureless or horizontally laminated clay and silt; often Floodplain deposits interbedded with coals and very fine to medium grain sandstone; may from suspension, contain plant material, wood fragments and leaf impressions; weak traction currents, occasionally siderite nodules and lenses are present. and floodplain ponds. Table 1 Continued Facies Code Facies Descriptions Interpretation Fr Rooted mud and silt Layers of incipient soils are found within sections of floodplain mud and Vegetated floodplain silt; roots and plant material are abundant; bioturbation and siderite nodules are occasionally present. Fc Carbonaceous mud Black to dark grey clay and silt; contains abundant organic material, plant and coal fragments; often laminated, but occasionally massive; commonly found grading up into coal or below channel scours. High organic content floodplain or silt diluted mire C Coal Lignite and sub-bituminous coal; black, occasionally displaying vitrinite reflectance; beds range from a few cm to 0.5 m thick; commonly underlain by carbonaceous mud that grades vertically into coal and overlain gradationally by carbonaceous shale or, more commonly, by a sharp contact with sandstone. Mire peat development Table 2 Descriptions of Facies Associations Facies Associations Architectural Elements Lithofacies Avg. Grain Size Biogenic Sed. Structures Macrofossils Coal Mire Floodplain Braided Fluvial Horizontal-tabular beds C, Fc Horizontal-tabular beds Downstream accretion sets, gravel beds, and gravel barforms Meandering Fluvial Lenticular and tabular (Braided Fluvial) sandstone bodies with lateral accretion sets Tidally-Influenced Fluvial Inclined heterolithic strata, and lateral accretion sets C, Fc, Fr, FI, Fm, Sr, Sb, Sm Sr, Sh, St, Sp, Sc, Smrc, Gmm, Gmt, Gem, Get Sf, Sr, Sh, St, Sp, Sm, Smrc FI, Fm, Sf, Sr, Sh, St, Sp, Sb, Sc silt-coal Planolites and Rooting mud-silt Planolites and Rooting cs-vcs ms fs Teredolites, Planolites, and Thalassinoides Skolithos, Teredolites, Rhizocorallium, and Planolites Teredolites, Planolites, Thalassinoides, and Skolithos Leaf impressions and wood fragments Leaf impressions and wood fragments Wood fragments Wood fragments, logs, leaf impressions Shells and wood fragments Table 3 Descriptions of Depositional Units and Bounding Members DU Facies Accociation Avg. Grain Channel Description Size Avg. Thickness Net to Gross D TM 1. B rai ded fluvi al DU-6 l.Braided fluvi al 2. Meandering fluvial vcs; granule- Highly amalgamated; laterally extensive channel belt pebble complex NA ms-vcs; Multi-story, highly amalgamated channel belt granule lenses complexes; moderately extensive; laterally accreting; 11m low horizontal offset 1.00 1.00 DU-5 1. Meandering Fluvial DU-4 1. Meandering Fluvial DU-3 1. Floodplain 2. Meandering to anastomosing fluvial ms-vcs; Multi-story, thick and laterally extensive, sheet-like granule lenses channel belts; laterally accreting; moderate horizontal 43 m 0.51 offset; channel belt complexes fs-ms Multi-story, thick, laterally accreting, and laterally 71m 0.42 extensive channel belts; large lateral offset; locally amalgamated silt-fs Single story; thin; laterally accreting; laterally isolated 46 m 0.30 channels in thick floodplain deposits; very low amalgamation N> Table 3 Continued DU Facies Accociation Avg. Grain Channel Description Size Avg. Thickness Net to Gross DU-2 1. Meandeing fluvial 2. Floodplain fs-ms Single and multi-story channels; lateral accretion; isolated to moderately laterally extensive; low to moderate amalgamation 17 m 0.58 DU-1 1. Tidally-influenced fluvial 2. floodplain 3. Coal mire ms Single story; downstream and laterally accreting; laterally isolated; high lateral offset; little to no amalgamation 14 m 0.42 DU-0 1. Braided fluvial 2. Tidally-influenced fluvial cs-ms Highly amalgamated, downstream accreting, channels that compose a laterally extensive channel belt complex 13 m 0.89 SHM 1. Floodplain 2. Braided fluvial vfs-fs; cs Single story, isolated channels and multi-story channel (Calico Bed) belt (highly amalgamated fluvial sheet) 50 m 0.56 LU>ti RESULTS Lithofacies Distribution of lithofacies (Figure 13; Table 1) varies vertically through the section (Figure 14). The most common lithofacies present in the measured sections were trough cross-stratified sandstone (St) and massive and laminated mudstone (Fm, Fl) (Figure 13h). Siltstone and mudstone facies (Fm, Fl, Fr) are nearly to completely absent from the lowermost and uppermost portions of the John Henry Member and the Drip Tank Member, but are the dominant lithofacies in the middle of the John Henry Member. Carbonaceous shale (Fc) and organic-rich siltstone/mudstone are most abundant in the lower portion of the John Henry Member, in an interval that begins directly above the amalgamated sand at the base of the section and extends about half way to the top of the member (Figure 7). This interval also contains the greatest occurrence of coal beds, which are most abundant lower in the interval and decrease vertically in abundance and thickness (Figure 14). The abundance and lateral distribution of sandstone facies varies throughout the section. The base and top of the John Henry Member, as well as the Drip Tank Member, are primarily composed of sandstone channels that are amalgamated into extensive sheets, whereas the middle of the John Henry Member is composed of isolated channel sandstone and laterally restricted sandstone channel belts. Trough cross-stratified sandstone (St) and massive sandstone (Sm) are the dominant lithofacies within channel deposits (Figure 13c,e). Convolute bedded sandstone (Sc) is also commonly located in channel sandstones, and is most abundant in DU-2 and the large channel sandstones located near the top of the John Henry Member (Figure 13d). Planar and ripple cross-laminated sandstone (Sp and Sr) are commonly located at the top of channel deposits, but at times compose entire beds (Figure 13a). Bioturbated sandstone (Sb) is commonly located in fine and medium grained sandstone at the bases and tops of channels (Figure 13b). Flaser bedded sandstone (Sf) is abundant in the lower John Henry Member and is absent higher in the section. Gravel lithofacies (Gcm, Gct, Gmm, and Gmt) are primarily limited to the Drip Tank Member and the uppermost channels of the John Henry Member (Figure 13g,f). These lithofacies are only sparsely present within the uppermost channel deposits in the John Henry Member and occur in greater abundance (Figure 13g), at times forming an entire bed, within Drip Tank Member. Facies Associations Five facies associations, which link assemblages of lithofacies with specific depositional environments, are identified within Bull Canyon. They include coal mire, floodplain, meandering and anastomosing fluvial, braided fluvial, and tidally-influenced fluvial (Figure 15; Table 2). Coal Mire Facies Association The coal mire facies association (Figure 15a) is composed of coal (C), carbonaceous shale (Fc), and organic rich siltstone/mudstone with abundant coal spar, 55 rooting, wood, and leaf impressions (Figure 16a,e). Occurrences of mire facies range in thickness from less than 1 m to greater than 8 m and extend laterally for tens of meters. Individual coal beds vary in thickness from 5 to 120 cm, with an average thickness of 35 cm. Mire facies laterally interfinger with floodplain facies, or are truncated by post depositional scouring, and are commonly capped by a sandstone fluvial channel. The coal mire facies association is present in the lower half of the John Henry Member, above the highly amalgamated sandstone at its base (DU-0), and almost completely absent in the upper half (Figure 14). The term mire generally applies to any of a number of non-saline peat-forming environments (McCabe and Parrish, 1992). Coal mires can develop in floodplain, coastal plain, delta plain, and back-barrier environments (Horne et al., 1978; Thornton, 1979; Thomas, 2002). The coals at Bull Canyon are interpreted as coastal plain mires that developed in swamps, small ponds, or lakes adjacent to fluvial channels. The distribution of coastal plain mires can be used as an indicator of changes in base level. Variations in the ratio of accommodation/peat accumulation, which dictates peat preservation, are mainly attributed to oscillations in base level (Deissel et al., 2000; Wadsworth et al., 2002; Davies et al., 2006). The groundwater table, which is one of the factors controlling peat development, can be impacted by sea-level (base-level) changes up to 150 km landward (Koster and Suter, 1993; Tornqvist, 1993). The lateral shifts in paralic coal distribution resulting from base-level influenced changes in the groundwater table correspond to progradational and retrogradational parasequences in sequence stratigraphy (Deissel et al., 2000). 56 The floodplain facies association (Figure 13b) is predominantly composed of massive and laminated siltstone and mudstone (Fm, Fl). Layers of abundant rooting (Fr) and organic material are common within the floodplain fine-grained deposits, and poorly developed soil horizons are occasionally present. Siderite nodules, either isolated or concentrated along distinct horizons, commonly occur within the floodplain fines. Siderite nodules develop in reducing environments below the water table during the early stages of diagenesis (Driese et al., 2010; Surez et al., 2010). Carbonaceous shales (Fc) are frequently present at the base of channels and at times grade vertically into coal (C). Tabular beds of very fine- to fine-grained massive, planar, and rippled sandstones (Sm, Sp, Sr), interpreted as crevasse splays, occur throughout this facies association, but are most abundant below and adjacent to channel sandstones bodies. Series of inclined (dipping ~10°), tapering downward medium-grained sandstone (Sm, Sl) beds that interfinger with the floodplain siltstones and mudstone are occasionally present at the edges of channel sandstones. These deposits are interpreted as levee deposits or floodplain deposits that were deposited in close proximity to a channel. The architectural elements (Table 2; Figure 15b) that comprise this facies association include: levee deposits, crevasse-splay deposits, floodplain fines, and incipient soil horizons (Miall, 1996; Bridge, 2003). Meandering and Anastomosing Fluvial Facies Association The meandering and anastomosing fluvial facies association is dominated by medium grain-size sandstone facies (Sm, St, Sp, Sc, Sr, Sh) that form lenticular or tabular bodies with a large width to depth ratio (Figure 15c). These sandstone bodies consist of a 57 Floodplain Facies Association single or multiple channel stories, which are composed of amalgamated barforms and fine upward. At their base is a sharp, erosive contact with the underlying floodplain and coal mire facies associations or an older channel sandstone. The channel facies laterally interfinger with crevasse splay and levee sandstone bodies of the floodplain facies association. Lag deposits composed of mud rip-up clasts (Smrc) (Figure 13e), gravel (Gmm, Gcm) (Figure 13g), wood, coal, bone fragments, and shell material (Figure 16h) are occasionally present at the bases of channel stories and along omission surfaces. Wood and coal are sporadically distributed throughout this facies association (Figure 16b). Trace fossils are sparse, but when present, they are often in dense concentrations of a single ichnospecies. Planolities, Rhizocorallium, and Teredolites are often located near the bases of beds and Skolithos and root traces are present in the fine grain sandstones near the tops of some sandstone bodies (Figure 17c,d,g,f). Accretion-set measurements indicate a wide range of accretion directions, a high proportion of which are oriented sub­lateral and lateral to the paleocurrent directions (DU-3 through DU-5 in Figure 18). This facies association is interpreted as channel deposits with characteristics that are indicative of low gradient, low discharge fluvial systems with moderate sediment load and infrequent channel avulsions. The arrangement of facies and the dominant lateral accretion in most of this facies association indicates a meandering fluvial system (Bridge, 2003). However, intervals dominated by coal mire and overbank deposits (interpreted as crevasse-splay, swamp, pond and lake deposits) with infrequent and laterally isolated channel sandstone bodies (DU-3) are consistent with the facies model for anastomosing fluvial deposits (Nadon, 1994). It is important to note that it is not possible to identify whether multiple channels were active at the same time, which is necessary to have 58 anastomosing channels (Schumm, 1968; Bridge, 2003). The preservation of coal is more likely in an anastomosing than meandering river environment, because the floodplain is not reworked through channel migration and frequent avulsions and abundant vegetation can develop in interfluvial wetlands (Smith and Smith, 1980). Anastomosing fluvial deposits are common in foreland basin systems and develop due to the combination of high-suspended load, seasonal water flow, and low depositional gradient (Nadon, 1994). Braided Fluvial Facies Association The braided fluvial facies association (Figure 15d) is primarily composed of coarse-grained sandstone (Sm, St, Sp, Sc, Sr, Sh) and gravel facies (Gcm, Gct, Gmm, Gmc). The median grain size is coarse to very coarse sand, but gravel beds (Gcm) ranging from 5 to 100 cm thick are common and fine-grain sediments are almost completely absent. Large-scale dune foresets (up to 2 m tall) composed of coarse-grained sand and gravel, are pervasive within this facies association. Undulating scour surfaces with up to several meters of relief are abundant and often intersect one another. Paleocurrent and accretion set measurements indicate that downstream accretion is more dominant than in the meandering and anastomosing facies association (DU-0 and DU-6 in Figure 18). This facies association is indicative of a higher gradient, higher discharge, bedload dominated fluvial system that experiences frequent avulsions and thalweg shifts. These characteristics fit the depositional model for a braided fluvial system (Schumm and Kahn, 1972; Chitale, 1973). 59 Tidally-influenced fluvial facies association (Figure 15e) primarily consists of very fine to lower-medium grain sandstone (Sr, Sp, Sb) and siltstone (Fl, Fr). This association contains flaser bedding, wavy bedding, lenticular bedding, and inclined heterolithic strata (IHS), which is rhythmically interbedded sandstone and siltstone facies that are dipping at angles of 5° to 20°. Interlaminated sandstone/mudstone within a few channel bodies were identified as possible double mud drapes, which are a good indicator of tidal influence (Visser, 1980). This facies association is commonly adjacent to coal mire facies and capped by the meandering fluvial facies association. In addition to Skolithos and Planolites (Figure 17g,c), the traces fossil assemblage of this association includes Thalassinoides (Figure 17a,b), most commonly present in marine to brackish salinities (Kern and Warme, 1974), and Teredolites (Figure 17f), a brackish water indicator (Bromley et al., 1984). Paleocurrent and accretion set measurements indicate a slight tendency towards lateral barform accretion (DU-1 in Figure 18). The IHS and rhythmic mudstone/organic drapes present in this facies association are formed under conditions of fluctuating current energy (Thomas et al., 1987; Shanley et al., 1992). These types of bedding have been documented in distinctly non-marine influenced fluvial environments (Jackson, 1981; Page et al., 2003); however, the presence of brackish/marine trace fossils at this location makes an exclusive fluvial influence unlikely, and supports a mixed fluvial and tidal depositional environment. Therefore, the IHS and rhythmic drapes are interpreted as representing periodic fine-grained sediment fallout that occurred when fluvial currents slowed due to influence of the tidal cycle (Thomas et al., 1987; Kvale and Mastalerz, 1998; Dalrymple and Choi, 2007). In other 60 Tidally-Influenced Fluvial Facies Association parts of the Kaiparowits Plateau, previous studies of the John Henry Member have also interpreted tidal influence within a portion of the fluvial deposits (Gallen, 2010; Gooley, 2010) or the entire fluvial section (Shanley et al., 1992; Shanley and McCabe, 1995). Channel Belt Dimensions and Stacking Patterns The true channel belt dimensions (Figure 12) for 136 channel belts on the east side of Bull Canyon were calculated using the GPS point cloud and paleocurrent measurements (Figure 18; Table 4). Due to the large variance in channel dimensions and the range of stacking patterns within Bull Canyon, a number of measurements were used to quantitatively document the variability. Channel belt widths are highly variable (from 15 to 733 m), and show trends of reducing widths from the base of the member to the middle and increasing widths from the middle to the top of the member (Figure 19a). The patterns in average channel belt thicknesses range from ~1 to 7 m and match the trends in channel belt widths (Figure 19b). The cross plots of channel belt width to thickness ratio (W/T) vs. stratigraphic order show an increase in the maximum W/T ratio higher in the section (Figure 20a). The W/T ratios for meandering channel belt are commonly 30-100 and braided systems are commonly 50-1000 (Gibling, 2006). In Bull Canyon, W/T ratios generally fit within the typical range of meandering channel belts. However, some studies would consider the fully amalgamated channel belts at the base and top of the section as a single unit (i.e., a fluvial sheet) and would document the dimensions of the entire amalgamated complex (Cotter, 1978). If this approach were used these intervals would have widths of over a km and much higher W/T ratios. The plot of channel belt width vs. thickness shows a possible linear relationship between these two variables (Figure 20b). 61 Vertical and lateral offsets describe the spatial relationship between channel belts. Channel belt offset measurements were made by projecting the channel belt midpoints (documented in the GPS point cloud) onto a vertical plane perpendicular to the mean paleocurrent and recording their relative positions. Figure 21 contains a graph and statistics of the lateral and vertical offset of channel belts based on the stratigraphic order assigned within each depositional unit. The moving average (with a range of five) for the lateral channel belt offsets is also plotted, which shows the larger trends in channel belt locations. The average magnitude of lateral offsets (dx) increases vertically from DU-1 (dx = 134 m) through DU-3 (dx = 531 m) and remains in fairly constant (dy = 523-556 m) from DU-3 through DU-6. Groups of channel belts separated from one another by large lateral offsets (as seen in DU-4) may indicate clustering and/or compensational stacking (Straub et al., 2009). The average magnitude of the vertical offset decreases vertically from DU-1 (dy = 10.1 m) through DU-4 (dy = 1.8 m) and then remains fairly consistent from DU-4 through DU-6 (dy = 1.4-1.8 m). DU-0 and DU-1 have small sample numbers due to lack of exposure, and thus, offset data for from these depositional units are less reliable than those higher in the section. Channel belt complex geometries are highly variable and range from groups of partially amalgamated channel belts, local clusters of high amalgamation, and extensive fully amalgamated intervals (or fluvial sheets). The degree of vertical and lateral amalgamation was quantified using an amalgamation index, in which, each channel belt was assigned a number for both vertical and lateral amalgamation: 0 = not amalgamated, 1 = amalgamated on one side, or on either the top or base, 2 = amalgamated on both sides, or the top and the base. These numbers were averaged for each depositional unit 62 and are displayed in Figure 22. Depositional units with very little amalgamation have index numbers ~0, and those with very high amalgamation have index numbers ~2. The degree of amalgamation does not appear to be related to changes in lateral channel belt offset in this outcrop, though there is a slight correlation between reducing vertical offset and increasing amalgamation. The degree of amalgamation appears to be related to channel belt dimensions, where larger channel belts are coincident with higher amalgamation index (Figures 19 and 22). Depositional Units and Bounding Members This section includes detailed descriptions of the Smoky Hollow Member, seven depositional units within the John Henry Member, and the Drip Tank Member, including: stratigraphic thickness, distribution of facies associations, channel belt dimensions, paleocurrent and accretion directions, and trace fossil assemblages. Smoky Hollow Member The Smoky Hollow Member underlies the John Henry Member, and is 33.5 m thick where it crops out near the mouth of Bull Canyon. A measured section and photos showing the extent of the Smoky Hollow Member are in Figure 23a,b. The Smoky Hollow Member was informally divided into three units (Peterson, 1969b), in ascending order: the coal zone, the barren zone, and the Calico bed. The coal zone consists of the lowermost 3.5 m of the Member and is composed of carbonaceous mudstone and coal beds (Figure 23g,h). Overlying the coal zone, the barren zone is 12 m thick and composed of mudstones and fine- to medium-grained horizontally and cross-stratified sandstones (Sh, St). The barren zone contains plant fossils, but no coal or carbonaceous 63 shale. The uppermost portion of the member, the Calico bed, is up to 16 m thick and consists of white and very light grey, fine- to very coarse-grained cross-stratified sandstone (St). The sandstones are poorly sorted and contain conglomerate lenses and isolated quartz and chert pebbles (Figure 23c,d). In this location, the Calico bed is composed of a series of sandstone bodies that are either separated by mudstones or are amalgamated into a large channel belt complex (Figure 23b). Teredolites was identified near the base of the Calico bed, which indicates a tidal influence (Figure 23e). The base of the Calico bed interfingers with the barren zone. Interpretation - Fluvial sandstone bodies at the base of the Smoky Hollow have incised into the underlying shorefaces of the Tibbet Canyon Member. The coal and carbonaceous mudstone of the coal zone were deposited in a coastal plain environment. The overlying isolated channel sandstones and floodplain deposits of the barren zone (which lack coal) represent an increase in sedimentation and fluvial influence (McCabe and Parish, 1992). At the top of the Smoky Hollow Member, the Calico bed is composed of tabular channelized deposits of the braided fluvial facies association with minor amounts of floodplain facies. The Calico bed has been interpreted as a braided fluvial system that formed during high sediment supply and low accommodation (Bobb, 1991; Semple, 2011), and associated with an unconformity and possible sequence boundary (Shanley and McCabe, 1991, 1995). The abrupt increased grain size in the Calico bed reflects a distinct change in the system at this time, which caused an increase in its carrying capacity. 64 Depositional Unit 0 (DU-0): Laterally Extensive, Braided Channel Belts Depositional Unit 0, the lowermost depositional unit in the John Henry Member, has an average thickness of 13 m and forms a prominent cliff (Figure 24a,b). The base of DU-0 is erosive, with over 2 m of relief, and directly overlies coarse-grained sandstones or a few meters of floodplain fines of the Calico bed of the Smoky Hollow Member. There is a recessive bench at the top of DU-0 (Figure 24a,c), which likely developed due to the presence of fine-grained floodplain and coal mire facies associations at the base of DU-1. Near the mouth of the canyon, a fault has caused a portion of the DU-0 section to be repeated in the measured section (Figure 6). The thick (~6 m) and laterally extensive (ave. 316 m) channel belts that comprise this unit contain many internal erosive surfaces and macroform accretion sets showing both downstream (Figure 24c,d) and lateral accretion (Figure 18f,g). These beds are composed of upper-medium to very-coarse grain sandstones of the braided fluvial facies association (Figure 24h). Overbank and floodplain deposits are generally absent from this depositional unit. Small, isolated zones composed of IHS and flaser bedding within the upper half of DU-0 weakly fit the tidally influenced fluvial facies association (Figure 24e), but this is a minor component of the unit. DU-0 has a high net-to-gross ratio (0.89), is highly laterally and vertically amalgamated (Figure 22), and has low lateral offset. The paleocurrent direction is to the northeast and barform accretion is dominantly downstream, with some sublateral accretion (Figure 18). Large-scale, up to several meters tall, high-angle dune foresets are distinctive features of DU-0 (Figure 24d). No trace fossils were identified within this depositional unit. 65 Interpretation - DU-0 is interpreted as a sand-dominated, low-sinuosity braided fluvial deposit composed of multiple, superimposed channel belts. Channel belts contain both bank attached bars and in channel, downstream accreting bars. The large dune foresets are oriented in the downstream direction and are interpreted as part of mid­channel, downstream accreting bars. Internal scours represent frequent avulsions and thalweg shifts, and larger scours may represent channel confluences. Depositional Unit 1 (DU-1): Tidally-Influenced Channel Belts Depositional Unit 1 has an average thickness of 14 m and is composed of floodplain and coal mire facies surrounding isolated, single story fluvial channels. A measured section and photographs of distinctive features are in Figure 25. Channel belts are thin (ave. <2 m), tabular, laterally isolated (ave. 75 m), have a low degree of amalgamation (amalgamation index = 0.40; Figure 22), and moderate lateral offset (Figure 19). There is a dramatic reduction in the net-to-gross ratio from 0.89 in DU-0 to 0.42 in DU-1. The average paleocurrent is to the east, with both downstream and lateral barform accretion (Figure 18). Floodplain and coal mire facies associations are commonly interbedded. Coal beds comprise ~2% of the unit and are located in clusters distributed irregularly throughout the unit (Figure 25c). Coal spar is commonly present within sandstones. Rooting and siderite nodules, indicative of poorly drained soils, are common in floodplain deposits. Rhythmically laminated flaser, wavy, and lenticular bedding, which are indicative of fluctuations in flow, occur within channel deposits (Figure 25d,e). IHS and sigmoidal heterolithic deposits occur both in and around channel deposits (Figure 25b,f). Due to the 66 limited number and isolated nature of the sigmoidal structures, they are not definitively identified as tidally derived (Kreisa and Moiola, 1986; Shanley et al., 1992). This unit contains the most diverse suite of trace fossils, which includes Thalassinoides, Teredolites, Skolithos, and Planolities, and the greatest presence of shell material (Figure 25g,h). The top of DU-1 was placed above the last layer showing prominent "tidal" indicators, commonly the stratigraphically highest IHS deposit. Interpretation - The depositional environment is interpreted as distal coastal plain located downstream from the bayline, where tidal cycles cause fluctuations in flow and brackish water may be present. This interpretation is supported by the presence of IHS, rhythmically laminated beds (possible double mud drapes) and marine/brackish trace fossils in channels that have a unidirectional flow; which indicate a dominant fluvial current that experienced fluctuation, likely due to tidal influence. The wide range of paleocurrent directions indicate that channels were highly sinuous, a trait common to the distal coastal plain (Dalrymple and Choi, 2007). The multiple adjacent channels could be part of a distributary system, though it is not possible to identify from outcrop whether channels are coeval and part of the same branching fluvial system. The presence of coal mires fits with this interpretation, as they are common to interdistributary locations in the lower coastal plain (Thomas, 2002). Depositional Unit 2 (DU-2): Laterally Restricted Channel Belts Depositional Unit 2 has an average thickness of 17 m, and is composed meandering fluvial and floodplain facies associations. Sandstone bodies form a series of back stepping ledges, up to 6 m tall, separated by slope forming floodplain deposits. The 67 net-to-gross ratio (0.58) for DU-2 is somewhat higher than DU-1. Figure 26 contains a typical measured section and photographs of the typical sedimentary features from DU-2. Fluvial channels are an average of 2.4 m thick, and isolated to laterally restricted (ave. 92 m). Some channels are multistoried, but the amalgamation index for DU-2 is low (Figure 20). Paleoflow is to the east with moderate sinuosity and sub-lateral barform accretion (Figure 26g). Planar, trough, and ripple cross-laminated sandstone (Fp, Ft, Fr) are common within channels (Figure 26b,c). In comparison with DU-1, this depositional unit contains very little evidence of any tidal influence; some flaser bedding is present (Figure 26f), but IHS is completely absent. Trace fossils are limited to those typical to fluvial setting, Skolithos and Planolities. Averages from the measured sections indicate that floodplain fines comprise 42% of this depositional unit, while the rest is composed of sandstone (Figure 14). Floodplain deposits are locally rooted and contain some carbonaceous content, but no coal was identified in this unit (Figure 26d,e). Near the mouth of the canyon, DU-2 is composed of a series of tabular sandstone beds that are successively thicker and show evidence of rapid sedimentation, including climbing ripples and convolute bedding. These sandstone bodies are separated by fine­grained floodplain deposits, thicker higher in the depositional unit, and eroded into and are capped by a series of meandering fluvial channel belts (Figure 26a). Interpretation - This unit is interpreted to contain meandering fluvial channels that developed in the distal portion of the coastal plain, though slightly less distal than DU-1. This environment could be near or just upstream of where distributary channels develop, which could explain why the channel belts are slightly larger than those in DU-1 (ave. 68 width: DU-1 = 75 m and DU-2 = 91 m). Being deposited further from the coeval shoreline than DU-1 could also explain the relative lack of tidal indicators. The succession near the mouth of the canyon is interpreted as a crevasse delta (Figure 26a) that was developed as a series of crevasse splay deposits that prograded into a body of water (Flores, 1983; Fielding, 1984). These deposits represent repeated overbank flows that eventually caused the main channel to avulse over these deposits, which produced the channel belts that scour into the splay deposits. Depositional Unit 3 (DU-3): Isolated Channel Belts in Thick Floodplain Deposits Depositional Unit 3 has an average thickness of 46 m and is composed of floodplain, coal mire, and meandering or anastomosing fluvial facies associations. DU-3 has the lowest net-to-gross ratio (0.30), least amount of amalgamation, and greatest abundance of coal of any depositional unit. The dominant feature of this DU-3 is the abundance of floodplain and coal mire facies associations (Figure 27a,b), which cumulatively compose ~70% of the deposit (Figure 14). Coal mire facies are present throughout DU-3, but occur in much greater abundance in the lower portions of the unit. The thickest coal beds (up to 1.2 m thick) within the John Henry Member are located in DU-3 and are encased in thick carbonaceous shale deposits. The presence of cover made it difficult to measure the lateral extent of the coals, but some extend laterally for 10s of meter. The top of this unit generally marks the vertical limit of the distribution of the coal mire facies association; a limited amount of coal was identified near the base of DU-4. 69 The floodplain deposits are primarily composed of siltstones (Fm, Fl) containing abundant rooting and siderite nodules (Figure 27c,d), which indicate poorly drained environments. Root haloes are common in both channel and floodplain material. Incipient soils were identified based on evidence of the early stages of horizon development and presence of blocky pedogenic structures. Floodplain fines are commonly red-brown, and the sandstones adjacent to them show signs of diagenetic red (hematite?) staining, which was most prominent along the tops of sandstone beds. Channel belts are typically single storied, thin (ave. thickness of 1.5 m), laterally isolated (ave. width 60 m), and have the lowest degree of amalgamation of all the depositional units (Figure 22). Well-defined channels are uncommon, and when present, composed of fine- to medium-grained sandstone with trough-cross stratification (St) (Figure 27f). Paleotransport was to the east with dominant lateral barform accretion (Figure 18). Many of the sandstone bodes of this DU-3 are poorly channelized, fine grained, and dominated by soft sediment deformation, planar bedding, and ripple cross laminations, commonly climbing ripples (Sc, Sp, Sr) (Figure 27g,h). Interpretation - The channels in DU-3 are interpreted as low sinuosity, slightly laterally accreting fluvial system with stabilized banks, which resulted in infrequent channel avulsion. The limitations of outcrop studies preclude the ability to determine if groups of these channels were coeval and interconnected to form an anastomosing fluvial system; however, this depositional unit generally fits the criteria established by Nadon (1994) for ancient anastamosed systems in the sedimentary record. Frequent flooding events could have caused deposition of thin tabular sandstone bodies, which are interpreted as avulsion deposits and crevasse splays/channels (Gibling, 2006). The 70 increased clastic input higher in the depositional unit likely caused the coincident reduction in the abundance of coal mire facies (McCabe and Parrish, 1992). Depositional Unit 4 (DU-4): Laterally Extensive, Meandering Channel Belts Depositional Unit 4 has an average thickness of 71 m, and is composed of meandering fluvial and floodplain facies associations. Figure 28 contains a typical measured section and photographs of sedimentary features of DU-4. The base of this depositional unit was placed where there is a marked increases in channel belt size, abundance, and degree of amalgamation. The percentage of coarse-grained sediment in DU-4 (net-to-gross ratio of 0.42) is greater that the underlying depositional unit. Floodplain facies are generally light yellow to light olive gray, massive siltstones (Fm) that are devoid of organics and siderite nodules. These deposits can be relatively thick (up to 10 m) and contain occasional crevasse splay deposits. Channel belts are composed of medium grained sandstone, typically thick (ave. 3.3 m) and laterally restricted to laterally extensive (ave. 180 m). However, a number of channel belts are laterally extensive (exceeding 300 m in width, Figure 19). There is moderate amalgamation among channels (Figure 22); some channel belts are locally amalgamated into groups, forming channel belt complexes. Paleotransport is to the east with dominant lateral barform accretion (Figure 18). Channels contain trough-cross stratification, planar lamination, and some ripple lamination (Figure 28b,c,d). Skolithos burrows were common at the tops of channel deposits (Figure 25 d,e). Some lag deposits contain shell and bone fragments (Figure 28g). 71 Interpretation - DU-4 is interpreted as a meandering fluvial system that is laterally accreting and moderately sinuous. Frequent avulsions likely resulted from less stabilized channel banks and caused the depositional unit to have a higher net-to-gross ratio than DU-3. Multiple avulsions that occurred in close proximity to one another produced the channel belt complexes. The lack of siderite nodules indicates that there was a transition from poorly drained floodplain in DU-3 to a well-drained floodplain in DU-4. Depositional Unit 5 (DU-5): Moderately Amalgamated, Meandering Channel Belts Depositional Unit 5 has an average thickness of 43 m, and is composed of meandering fluvial and floodplain facies associations. Figure 29 contains a typical measured section and photographs of sedimentary features from DU-5. The base of DU-5 is located below the first channel that contains granule and gravel lag deposits (Figure 27g). Fine grained sediments compose less than 50% of this depositional unit (Figure 14). The percentage of fine-grained sediment is less than the underlying depositional unit, continuing the trend that is present in DU-4. Floodplain deposits are similar to those of DU-3 and consist of light yellow to light olive gray, massive siltstones (Fm) that are devoid of organics and siderite nodules. The occurrences of overbank deposits are relatively limited. Channels belts are composed of medium to lower coarse sandstones, with a steady vertical increase in the presence of gravel lags and average grain-size. The average channel belts thickness is 3.9 m and average width is 225 m (Figure 19). Channel belts are commonly amalgamated into channel belt complexes (moderate amalgamation index, 72 Figure 22). Channel belt complexes are much larger in size and are comprised of a larger number of channel belts than those of DU-4. Paleotransport is to the east, with dominant lateral accretion (Figure 18). The top of DU-5 is located along the stratigraphic level, above which the channel belts are completely amalgamated. In DU-5, trace fossils are more abundant and the assemblage is more diverse than DU-2, DU-3, and DU4, and includes Rhizocorallium, Thalassinoides, and Skolithos. Rhizocorallium is typically considered a marine trace produced in firmgrounds (Figure 27e); however, there is one documented occurrence of non-marine Rhizocorallium (Fursich and Mayr, 1981). Interpretation - DU-5 is interpreted as a meandering fluvial system due to its channel characteristics and facies associations. The presence of gravel within channel deposits demonstrated an increase in the carrying capacity of the fluvial system. The trace fossil assemblage may indicate an increase in marine influence; a similar trend was documented in Rock House Cove (Gooley, 2010). Depositional Unit 6 (DU-6): Highly Amalgamated Channel Belts Depositional Unit 6 has an average thickness of 11 m and is composed of the meandering and braided fluvial facies associations. Figure 30 contains a typical measured section and photographs of sedimentary features from DU-6. This depositional unit is cliff-forming, composed of medium sandstone through gravel clasts, with no silt or fine­grained sandstone (net-to-gross ratio = 1.0), and contained occasional mud rip-up clasts (Figure 30b,c). Sandstone bodies typically contain trough cross-stratification and low angle bedding (Figure 30d,f). The mean grain size increases stratigraphically higher 73 within the DU. Teredolites traces were the only trace fossils present with DU-6 (Figure 30e). Channel belts have an average thickness of 3.5 m and average width of 190 m (Figure 18). The smaller channel belt widths compared to DU-5 are likely due to the increase in channel amalgamation (amalgamation index: vertical = 1.93 and lateral = 1.83), which caused truncation of channel belts. Low lateral offset and high amalgamation of channel belts resulted in a sheet-like unit with abundant internal scour surfaces, up to several meters deep, which frequently crosscut one another (Figure 30a). Paleoflow is to the east, with intermediate lateral accretion (Figure 18). Interpretation - The channel belts of DU-6 are transitional from more meandering fluvial at the base to sandy braided fluvial at the top. The paleocurrent and accretion measurements were taken at a limited number of locations due to the inaccessibility of the cliff forming exposures, and as a result, it was not possible to document any vertical changes in paleocurrent or accretion through the depositional unit. However, the fluvial system experienced frequent avulsions and no floodplain or overbank deposits are preserved, which are more characteristic of a braided fluvial system. The presence of Teredolites in DU-6, in combination the trace fossil assemblage in DU-5, indicates a possible marine influence near the top of the John Henry Member. Drip Tank Member A few meters of the Drip Tank Member are preserved along the rim of Bull Canyon, which is composed of coarse-grained sandstone (Sm, St, Sl) through gravels (Gmm, Gmt, Gct, Gcm) of the braided fluvial facies association (Figure 31). No floodplain deposits are preserved among the channel deposits (Figure 14). In addition to 74 forming beds (up to ~1 m thick) and lag deposits, pebbles and granules are located in isolated areas throughout many of the sandstone beds (Figure 31d,e). No trace fossils, except Teredolites in float material, were identified within the Drip Tank Member. As documented in other locations, the top of the John Henry Member (DU-6) interfingers with the Drip Tank Member (Peterson, 1969a,b; Bobb, 1991; Gooley, 2010), and because of this, the member contact is considered to be gradational. The contact between these members was placed along a series of scour surfaces, above which there is a distinct increase in mean grain-size, from upper medium sandstone to very coarse sandstone or gravel (Figure 31a,f,g). Large dune foresets, up to several meters tall, and abundant scour surfaces with gravel lags are prominent sedimentary features (Figure 31b). The paleocurrent direction was highly variable, with an average orientation to the east-southeast (Figure 18). Barform accretion was also highly variable, with both downstream and sub-lateral accretion; however, the most common accretion direction is downstream. Due to the limited exposure of the Drip Tank Member at Bull Canyon, the dimensions of channel belts were not documented. Interpretation - The Drip Tank Member is interpreted as mixed sand and gravel, downstream accreting braided fluvial system; a continuation of the trend of increasing braided fluvial characteristics that began in DU-6. The base of the Drip Tank Member, where there is an increase in grain size, marks the onset of an increase in stream competence. This fluvial system experienced frequent avulsions and thalweg shifts that removed any of the fine-grain sediments that were present. 75 Paleomorphodynamics The steps included in the paleomorphodynamic workflow are discussed in detail within this section. The workflow was applied to the depositional units in Bull Canyon and Rock House Cove, using data collected by Gooley (2010), which permits observations of temporal and spatial variations in paleomorphodynamic parameters. Graphs of the key paleomorphodynamic parameters (bankfull depth, slope, Froude squared, and backwater length) are in Figures 32 and 33. The bankfull depth (H) is most reliably measured directly from non-truncated point bars, but various methods exist for calculating bankfull depth that can be used when direct measurements are not possible (Bridge and Tye, 2000; Leclair and Bridge, 2001). Due to the limited number and uneven distribution of nontruncated barforms identified within the field area, the bankfull depths were calculated using the method described by Leclair and Bridge (2001), which proved to be the most accurate when compared to direct barform measurements. In this method the bankfull depth is 6-10x the mean dune height (hm), which is calculated using the formula: hm = (2.9 ± 0.7) * Sm (1) where Sm is mean cross-set height in meters. Calculations of bankfull depth using 8x hm produced the most reliable estimates compared to story heights. Calculations using 6x and 10x hm were propagated through the rest of the workflow calculations and used to create the error bars, which document the uncertainty produced by initial uncertainty in bankfull depth (Figures 32 and 33). 76 The calculated bankfull depths at Bull Canyon show a relatively large range of flow depths, from 3 to 10 m, and a trend of decreasing flow depth vertically through the section (Figure 32a). The lower DUs have larger estimated flow depths (>4 m to 10 m), whereas, the upper DUs have smaller calculated flow depth (generally <4 m). Calculated bankfull depths for Rock House Cove vary from 2.4 to 5.1 m (Figure 33b). On average the Rock House Cove estimates are slightly smaller in magnitude than those of Bull Canyon (but generally within the error range), and they do not show a consistent decreasing trend through the section. The paleoslope (S) calculation provides insight into the slope of the streambed at the time of sediment deposition. The formula used to calculate paleoslope was (Paola and Mohrig, 1996): S = R"* (d50/H) (2) where R is the submerged specific gravity quartz (1.65), t * is the shield stress (1.0 for sand bedload and 0.4 for gravel bedload - 1.0 was used for these calculations), and d50 is the median grain size of bedload material. The d50 was estimated for each DU based on the grain size in sandstone cross-sets near the base of channels, lag deposits were excluded. This estimation is very likely an under-represented source of uncertainty in the present dataset; more thorough d50 measurements would improve the results. This calculation produces a dimensionless quantity, but was converted to degrees for the graphs in Figures 32b and 33b. The dimensionless quantity was used in the following equations. 77 The calculated paleoslopes for Bull Canyon indicate that the system had a low gradient, slopes of 6 x 10-5 (0.003°) to 3 x 10-4 (0.02°) (Figure 32b). Typical toe-to-apex slopes in modern distributive fluvial systems are between 6 x 10-5 (0.003°) and 2.6 x 10-2 (1.5°) (Hartley et al., 2010). A distinct change in the calculated slope occurs in the middle of the John Henry Member. On average, the calculated paleoslopes for Rock House Cove are slightly higher than those at Bull Canyon (Figure 33b), which is typical of a more proximal location on a graded stream profile. At both Rock House Cove and Bull Canyon, the paleoslopes in the lower half (DU-0 through DU-3) are relatively low and fairly consistent (perhaps showing a slight vertically decreasing trend at Bull Canyon), whereas, paleoslopes in the upper half (DU-4 through DU-6) are larger by ~ 0.01 degrees and increase vertically. The square of the Froude number, or Froude squared (Fr2), and backwater length (Lw) parameters provide insight into the interaction of upstream normal flow and downstream base level. Froude squared is calculated from the drag (Cd) and depth averaged velocity (U) using: Cd = (d50/H)116 (Hajek and Wolinsky, 2012) (3) U = Vt /Cd (Hajek and Wolinsky, 2012) (4) Fr2 = (U/-J g * H )2 (Parker, E-book) (5) where t is the kinematic shear stress (= stress / fluid density; (m/s)2) and g is gravitational acceleration (9.8 m/s). Far upstream Fr2~1 and the flow depth is determined by the discharge and topography. Far downstream Fr2~0 and the water surface is impacted by 78 the downstream base level (Sturm, 2001). In the transition zone, located between the two end members listed above, Fr2<<1 and flow is influenced by upstream discharge, topography, and downstream base level (i.e., within the backwater curve) (Figure 34). This backwater transitional zone is the length of a channelized flow that is included in the backwater curve, where flow "feels" the impact of the base level causing the slope of the water surface to be less than the channel bed slope (Chanson, 2004). The calculations of Fr2 are all quite low in Bull Canyon (0.03 to 0.07) and Rock House Cove (0.05 to 0.09) (Figure 33c and 34c). Fr2 numbers are much closer to 0 than 1, which indicates that these locations may be within the backwater transition zone. There is a trend of increasing Fr2 numbers vertically through the section at Bull Canyon; the most dramatic increase occurs above DU-3. The Fr2 numbers for Rock House Cove are relatively consistent with the exception of DU-3, which is slightly larger than the others (although error bars are large: Figure 33c). The backwater length (Lw) scales with the backwater transition zone (backwater transition zone ~ 3Lw) (Hajek and Wolinsky, 2012). The formula used to approximate the backwater length is (Paola and Mohrig, 1996): Lw = H / S (6) where H is upstream "normal flow" depth. In order for this calculation to be accurate the input data must be from a location upstream from the bayline (or landward of the transition zone) (Strum, 2001; Chanson, 2004). This formula will overestimate the backwater length if H is taken from within the transition zone. 79 Calculations of Lw for Rock House Cove range from 10-36 km and show a trend of vertically decreasing backwater length, with the exception of DU-0 (Figure 33d). Lw calculations at Bull Canyon show a greater degree of variability. Calculated Lw of DU-4 through DU-6 at Bull Canyon (4-14 km) are roughly equivalent to the upper DUs at Rock House Cove (Figure 32d). However, Lw lengths for DU-0 through DU-3 at Bull Canyon (75-97 km) are dramatically larger than those of the upper half of Bull Canyon and the entirety of Rock House Cove. Within the transition zone, bankfull channel depths should increase downstream as the surface gradient diverges from the channel bed slope (Strum, 2001; Petter, 2010), which causes the paleoslope calculation to predict increased Lw lengths downstream. This may explain the variations between calculations of Lw in the lower half of Rock House Cove and Bull Canyon. Assuming dominant eastward sediment transport, the distances from the John Henry Member at Rock House Cove and Bull Canyon to the landward pinchouts of correlative shorelines (i.e., the pinchout of the C sandstone) are ~55 km and ~35 km respectively (Allen, 2009; Dooling, in prep). The locations of shoreline pinchouts are variable within the John Henry Member, but the pinchouts are all within a ~15 km zone that is roughly parallel to Fiftymile Mountain (Allen, 2009). The distances from Rock House Cove and Bull Canyon to the shorelines are less than the transition zone predicted by Lw calculations from Rock House Cove; however, the measured distance to the shoreline is not necessarily equivalent to the actual distance to where the fluvial system interacted with the shoreline. The presence of tidal indicators in lower portions at both locations would support the interpretation that deposition occurred within the backwater transition zone, for at least some period of time (Gooley, 2010). 80 Sedimentary Petrology Petrographic analysis consisted of point counting 20 samples that were collected, at roughly even spacing, from just below the base of the John Henry Member up to the base of the Drip Tank Member in medium grained sandstones. Point counts for each sample included a minimum of 500 grains consisting of monocrystalline quartz (Qm), polycrystalline quartz (Qp), chert (Cht), potassium feldspar (K), plagioclase (P), and miscellaneous grains. Subcategories of the miscellaneous grains category include lithic carbonate (Lc), sedimentary lithic fragment (Ls), volcanic lithic clasts (Lv), and metamorphic lithic clasts (Lm) (these data are included in Appendix F). The data were plotted on standard ternary diagrams (Figure 35), QmFLt diagram with provenance fields (Dickinson, 1985) and QtFLu diagram with sandstone classification fields (Pettijohn, 1975). The end members of the QmFLt diagram include Qm, F (feldspar, K+P), and Lt (total lithic fragments, Lc+Ls+Lm+Lv+Cht+Qp) and the end members of the QtFLu diagram are Qt (total quartz, Qp+Qm+Cht), F, and Lu (unstable grains; Lc+Ls+Lm+Lv). The presence of detrital carbonates (Lc) is characteristic of many foreland basins (Dickinson, 1985); however, they are typically not recalculated with other lithic grains in QtFLu and QmFLt diagrams due to the difficulty in distinguishing between extrabasinal or intrabasinal origin and their different geochemical responses to weathering and diagenesis (Mack, 1984). In this study, lithic carbonates were included in lithic counts because intrabasinal carbonates are not a concern due to the location of the study area proximal to the orogenic belt in terrestrial (fluvial) strata. However, weathering and diagenesis may have an impacted their concentration. 81 The two samples from the Calico bed of the John Henry Member are quartz arenites (98-99% Qt) and plot within the craton interior field of the QtFLu diagram (Figure 35b). Samples from the John Henry Member are generally sublithic arenites or lithic arenites and fall within the recycled orogen field of the QtFLu diagram (Figure 35a). The compositions of John Henry Member samples are typical of fold-thrust systems (Dickinson and Suczek, 1979; Dickinson, 1985). The Calico samples are more Qt rich and poor in Lu and F than expected in this tectonic setting, although interstratified petrofacies with highly variable and contrasting Qm/Lt and Qt/Lu ratios have been documented in foreland basins sequences (Putnam, 1982). The differences in composition between the Calico bed and the John Henry Member may be attributed to either different proportions of recycled quartzose sand derived from the orogenic terranes or contrasting climatic/weathering influences. The percentage of lithic grains increases from the base to the middle of the John Henry Member (DU-0 to DU-3), and then gradually decreases vertically through the rest of the member. This trend is primarily due to the abundance of Lc (up to 28%) in the middle of the member (DU-3 and DU-4) compared to the upper and lower sections (0­14% Lc). A similar, though more pronounced, trend was documented in Rock House Cove (Gooley, 2010). The increase in lithic carbonates in the middle of the John Henry Member may have been caused by the exposure of Paleozoic rocks in the upper plate of the Blue Mountain thrust sheet of the Sevier orogenic belt (Fillmore, 1991; Goldstrand, 1992) or by differences in preservation potential in the middle depositional units. 82 83 Figure 13 - Lithofacies photos. A) Ripple cross-laminated sandstone (Sr) and massive siltstone/mudstone. B) Bioturbated sandstone (Sb), coal (C), and horizontally laminated sandstone (Sh). C) Trough cross-stratified sandstone (St). D) Convolute sandstone. E) Massive sandstone (Sm) and massive grain supported mud rip-up clast conglomerate. F) Grain and matrix supported cross-stratified conglomerate. G) Matrix and grain supported conglomerate (Gmm and Gcm). H) Massive siltstone/claystone (Fm), carbonaceous shale (Fc), and coal (C). 84 85 Figure 14 - Distribution of lithologies within depositional units (DU-0 through DU-6) and members (SHM = Smoky Hollow Member; DTM = Drip Tank Member). The lithologic information documented in the measured sections (Figure 7) was divided into five categories based on grain size: sandstone, interbedded sandstone and siltstone, siltstone/mudstone, coal, and covered (undefined lithology). The thicknesses of the columns are scaled to the average thickness of each depositional unit. 100 m 86 DTM DU-6 (11 m) DU-5 (43 m) DU-4 (71 m) DU-3 (46 m) DU-2 (17 m) DU-1 (14 m) DU-0 (13 m) SHM (33 m) CD s- .Q D° Etu Q ^ C%D 0 9> o <C,£ 10 20 30 40 50 60 70 80 90 100 Percent of Grainsize Range by Depositional Unit H Sandstone ■ Mixed sandstone and mudstone H Siltstone/mudstone ■ Coal ■ Cover 87 Figure 15 - Photos of the facies associations. A) Coal mire facies association. B) Floodplain facies association. C) Meandering (and anastomosing) fluvial facies association. D) Braided fluvial facies association. E) Tidally-influenced fluvial facies association. 88 89 Figure 16 - Body fossils and biologic structures. A) Gymnosperm needles and unidentified plant fragments. B) Petrified wood. C) Root traces in channel sandstone. D) Root traces in siltstone. E) Angiosperm leaf in carbonaceous shale. F) Angiosperm leaf in siltstone; G) Bivalve shell fragments. H) Shell and bone fragments in a sandstone channel lag deposit. 90 91 Figure 17 - Trace fossils. A) Large Thalassinoides. B) small Thalassinoides. C) Planolites. D) Rhizocorallium? E) highly bioturbated sandstone (bioturbation index 5). F) Teredolites. G) Lined Skolithos and Planolites. H) Skolithos. 92 M>«> •< 4 . •#" o«**> • ✓ . *• -• '-ac• -/---.-„ vA ~y_ M - ^ ' ' ' 'T & 'X j . s c . ' * * ‘ ^ .. - , • " ► -^KlP. 1 . ; . « ..■ • . . -n** '• ^ ■ 1 ■ ■ ^ , . * n h h n Mvife ^ - * r % o r I * - 1, ' ' » i*'?t T tI'^J IN i A a - J w i « » » t*. - ' !'?1.+- rV/ * -***. .« " 1 •- -■'J |h Society for Sedimentary Geology ■ vI ~ Ii ■?■ ■ 1 93 Figure 18 - Paleocurrent and barform accretion direction measurements from Bull Canyon grouped by depositional unit and member. Figure 10 includes a description of these graphs. DU-0 DU-1 DU-2 DU-3 DU-4 DU-5 DU-6 Drip Tank TOl 00 TlO cono TlO CO TlO TlO (35W TlO o cn >O >O CD >o -wtv >O VO 95 Figure 19 - Channel belt width and thickness. The average (ave), standard deviation (SD), and number of measurements are listed for each DU. Colored boxes represent the averages for each depositional unit (DU). The blue lines represent moving averages (with a range of five). A) True (corrected) channel belt widths in stratigraphic order (see Figure 12). B) Channel belt thicknesses in stratigraphic order. Channel Belts in Stratigraphic Order A) Channel Belt Widths 130 120 110 100 DU-6: Ave = 191 m SD = 87 m, n = 12 DU-3: Ave = 74 m, SD = 50 m, n = 19 DU-2: Ave = 92 m, SD = 67 m, n = 30 DU-1: Ave = 75 m, SD = 44 m, n = 5 DU-0: Ave = 316 m, SD = 123 m, n=3 Channel Belt Width (m) Channel Belts in Stratigraphic Order B) Channel Belt Thicknesses 800 130 120 110 100 DU-6: Ave = 3.5 m SD = 1.6 m, n = 12 DU-5: Ave = 3.9 m, SD = 1.6 m, n = 21 DU-4: Ave = 3.3 m SD = 1.3 m, n = 45 DU-3: Ave = 2.0 m, SD = 0.7 m, n = 19 DU-2: Ave = 2.4 m, SD = 1.0 m, n = 30 DU-1: Ave = 1.7 m, SD = 0.7 m, n = 5 DU-0: Ave = 6.0 m, SD = 1.0 m. n=3 10 Channel Belt Height (m) 12 VO Ov 97 Figure 20 - Cross plots of channel belt dimensions. A) Width-to-thickness ratios (W/T) of channel belts are displayed in stratigraphic order. The W/T ratios were calculated, using the measurements from the GPS dataset, and dividing the width by the thickness (channel belt width/channel belt thickness). B) Cross plot of channel belt widths and thicknesses. Measurements of channel belt width and thickness are from the GPS dataset. 98 A) Ratio of Channel Belt Width to Thickness (W/T) ■ DU6 ■ DU5 D DU4 ■ DU3 ■ DU2 ■ DU1 ■ DU0 B) sse nk c e Belnn ahC 8 7 6 5 4 3 2 1 0 c r. ■ □ ■ i ™* ■- b ■■ 4 ■ 4 V ' V l T t c ' r- . Z \ . ■ 0 100 200 300 400 500 600 700 800 Channel Belt Width (m) DU6 DU5 DU4 DU3 DU2 DU1 DU0 99 Figure 21 - Vertical and lateral channel belt offsets. A) Graph of the vertical and lateral offsets between channel belt centers in the stratigraphic order assigned within each depositional unit. A moving average of horizontal offsets with a range of five (MA Horizontal) is also included. B) List of the average (ave), median (med), and standard deviation (STDEV) for the change in vertical (dy) and lateral (dx) offsets between channel belts. A) Vertical and lateral offset of channel belts Channels belts in assigned order B) Average vertical (y) and lateral (x) offset of channel belts DU n= Ave dx Med dx SDEV dx Ave dy Med dy STDEV dy DU-6 12 555.7 603.9 200.0 1.4 1.3 1.3 DU-5 24 525.5 421.0 318.3 1.4 1.2 1.6 DU-4 45 522.6 439.0 330.2 1.8 1.2 1.8 DU-3 17 530.2 433.1 442.5 2.7 1.6 3.2 DU-2 30 344.8 283.5 247.3 2.1 0.7 3.8 DU-1 5 251.4 122.4 194.5 5.7 5.9 2.8 DU-0 3 133.8 133.8 50.0 10.1 10.1 0.4 Vertical offset of channel belts (m) 101 Figure 22 - Channel belt amalgamation index. Each channel belt was assigned a number for both vertical and lateral amalgamation: 0 = not amalgamated, 1 = amalgamated on one side, or on either the top or base, 2 = amalgamated on both sides, or the top and the base. The average vertical and lateral amalgamation index numbers for each depositional unit are displayed. 102 DU-6 n=12 DU-5 n=23 DU-4 n=48 DU-3 n=19 DU-2 n=30 DU-1 n=5 DU-0 n=3 Channel Belt Amalgamation Index 1.93 1.83 1.09 0.74 0.70 0.39 0.35 | 0.10 0.31 0.52 0.40 0.40 2.00 2.00 0.0 0.5 1.0 1.5 2.0 Average amalgamation index number Vertical amalgamation Lateral amalgamation 103 Figure 23 - Summary plate of the Smoky Hollow Member (SHM). On the left, an example section of the SHM that shows the base of the Calico bed in green. On the right, photos: A) The entire Smoky Hollow Member section is shown between the blue line and the green line, which marks the base of the Calico bed. B) The thickest exposure of the calico bed; located near the mouth of Bull Canyon. The blue line marks the top of the Smoky Hollow Member and the green line marks the base of the Calico bed. C) Bleached, very coarse- to medium-grained sandstone of the Calico bed with chert and quartzite gravel. D) Bleached, coarse-grained sandstone of the Calico bed. E) Teredolites from near the base of the Calico bed. The staff is marked with 10 cm increments. F) Ripple cross lamination (Sr) at the base of a sandstone bed near the base of the Calico bed. G) The top of the coal zone and transition to the barren zone at the base of the sandstone. The staff is marked with 10 cm increments. H) Angiosperm leaves in carbonaceous shale from the coal zone. M)I 105 Figure 24 - Summary plate of DU-0. On the left, an example section through DU-0. On the right, photos: A) Contact between Smoky Hollow (Calico bed) and John Henry Member (DU-0) at the transition from bleached to caramel color sandstone. White arrows point to the member contact. B) Erosive contact between the Calico bed of the Smoky Hollow and DU-0 of the John Henry Member indicated by white arrows. C) Large downstream accreting bar. D) Steeply inclined forests and scours from a downstream accreting bar. E) Inclined heterolithic strata from a muddy bar located in the middle of the DU-0. F) Laterally accreting point bar. G) Two stories of lateral accreting point bars. H) Very-coarse sand and gravel from DU-0. 107 Figure 25 - Summary plate of DU-1. On the left, an example section through DU-1. On the right, photos: A) Exposure of tidally influenced channel belts and laminated and carbonaceous shale, interpreted as lake/pond deposits. B) Inclined heterolithic strata from a muddy bar. C) Horizontally laminated and massive siltstone and carbonaceous shale capped by channelized sandstones. D) Carbonaceous shale drapes. E) Paired mud/sandstone couplets (double mud drapes?). F) Sigmoidal sandstone. G) Thalassinoides in a tidally influenced channel belt sandstone. H) Bivalve shells and small burrows in a channel sandstone. 109 Figure 26 - Summary plate of DU-2. On the left, an example section through DU-2. On the right, photos: A) Crevasse splay deposits capped by a channel belt complex, interpreted as a crevasse splay delta. B) Trough cross-stratified sandstone. C) Thick assemblage of ripple cross-stratification that includes climbing ripples. D) Carbonaceous shale under channelized sandstone. E) Carbonaceous shale. F) Flaser bedding. G) Lateral point-bar accretion. 110 CO CO CN CN 111 Figure 27 - Summary plate of DU-3. On the left, an example section through DU-3. On the right, photos: A) Thick accumulations of floodplain fines and coal mire facies, with few channelized deposits. The green line marks that top of DU-3. B) Variegated fine grain deposits with occasional thins crevasse play sandstones. C) Root haloes at the top of a crevasse splay deposit. D) Siderite nodule in siltstone/mudstone. E) Coal mire facies capped by channelized sandstone. F) Trough cross-stratified sandstone in a channel. G) Ripple and horizontally laminated sandstone. H) Asymmetric ripples (and climbing ripples) in a crevasse splay sandstone. 112 f ---------------------------------------------------------------------------------------------------------------------------------►c 113 Figure 28 - Summary plate of the DU-4. On the left, an example section through DU-4. On the right, photos: A) Panoramic photograph showing the channel belts of DU-4, located between the green lines. B) Abundant trough cross-stratification. C) Planar laminated sandstone. D) Lined Skolithos in a channel sandstone. E) Skolithos at the top of a channel fill. F) Bioturbation in rippled sandstone. G) Lag deposit with shell and bone fragments. 114 115 Figure 29 - Summary plate of the DU-5. On the left, an example section through DU-5. On the right, photos: A) Panoramic photograph showing the laterally extensive channel belts. B) Trough cross-stratified sandstone. C) Muddy toe sets of a bar, located at the base of a channel deposits. D) Convolute sandstone common at the bases of many of the channel belts. E) Rhizocorallium? located along the base of a channel sandstone. F) Climbing ripples in a splay deposit. G) Very coarse-grained sandstone and small pebbles in lags. 116 117 Figure 30 - Summary plate of the DU-6. On the left, an example section through DU-6. On the right, photos: A) Panoramic photograph of DU-6; the green line marks the based of DU-6 and the blue line marks the base of the Drip Tank Member. B) Amalgamated channel belts. One-meter increments marked in black and white. C) Matrix and grain supported conglomerates (Gmm and Gcm). D) Coarse-grained trough cross-stratification. E) Teredolites. F) Granules and mud rip-ups on large forests. 119 Figure 31 - Summary plate of the Drip Tank Member (DTM). On the left, an example section through the lower portion of the DTM. On the right, photos: A) Panoramic photo of the Drip Tank Member. White arrows point to the contact with the John Henry Member. B) Large dunes and downstream accreting barforms. C) Gravel bed about 0.5 m thick. D) Gravel lag (Gmm) at the base of the DTM. F) Coarse and very-coarse graded sandstone. G) ~2 m thick channel at the base of the DTM. G) The transition from lighter colored sandstones of DU-6 to caramel colored sandstones and gravels of the DTM (located above the notebook). 121 Figure 32 - Paleomorphodynamic parameters calculated for Bull Canyon (BC). Graphs show parameters calculated for each depositional unit and include: A) BC - bankfull depth, B) BC - Paleoslope, C) BC - Froude squared, and D) BC - backwater length. A) BC - Bankfull Depth (H) h 5 I-O-I 4 I- O Q 3 2 h 1 I------- ♦ -------1 0 I---------- ------------1 0 2 4 6 8 10 12 Depth (m) B) BC - Paleoslope (S) Paleoslope (degrees) C) BC - Froude Squared (Fr2) D) BC - Backwater Length (Lw) Backwater Length (km) 123 Figure 33 - Paleomorphodynamic parameters calculated for Rock House Cove (RHC), based on data collected by Gooley (2010). Graphs show parameters calculated for each depositional unit and include: A) RHC - bankfull depth, B) RHC - paleoslope, C) RHC - Froude squared, and D) RHC - backwater length. A) RHC - Bankfull Depth (H) 5 I----------O-------1 4 I-----------O----------1 g 3 K>-------- 1 2 I-------- O------- 1 1 I----------O--------- 1 0 I--------0------1 0 2 4 6 8 10 Depth (m) B) RHC - Paleoslope (S) 5 4 Q 3 2 1 0 0 I-------o - I- 0 ------------ 1 I-♦ -------1 I------ 0-I I-0 --------1 I-0 ------- 1 I------- 0 -----------------------1 0.01 0.02 0.03 Paleoslope (degrees) 12 0.04 C) RHC - Froude Squared (Fr2) D) RHC - Backwater Length (Lw) Backwater Length (km) 125 Figure 34 - Backwater and low-Froude flows (modified from Hajek and Wolinsky, 2012). A) The slope of the bed and water surface are typically equivalent in the upper reaches of a river, but in the lower reaches the slope of the water surface is often less than the bed slope. The area within the black box represents the backwater transition zone (i.e., the extent of the backwater curve). B) The backwater transition zone is where the influence of base level is felt upstream. This transition zone scales with the backwater length (Lw). 126 A) B) Lw Fr2 * 1 Fr2 * 0 127 Figure 35 - Sandstone petrographic analysis from the John Henry Member at Bull Canyon. A) QmFLt diagram with sandstone classification fields from Pettijohn (1985). B) QtFLu diagram with provenance fields from Dickinson (1975). 128 Bull Canyon Samples • DTM • DU-6 o DU-5 • DU-4 o DU-3 • DU-2 • DU-1 • DU-0 0 Calico John Henry Mbr Average Compositions and Standard Deviations 0 Bull Canyon 0 Kelly Grade (Gallin, 2010) + Rock House Cove (Gooley, 2010) A) B) Qt □ Continental Block □ Magmatic Arc □ Recycled Orogen Qt Total Quarts F Feldspar Lu Unstable Lithics Lu Table 4 Channel Belt Measurements from GPS Data and Stratigraphic Sections Depositional Unit Ave. Width* STDEV Width* Max Width* Ave. Thickness* STDEV Thickness* Max Thickness* Ave. W/T* Ave. d50 Bedload8 Ave. Cross-set Thickness8 Ave. Dune Height* DU6 190.98 86.94 360.90 3.5 1.6 5.9 67.3 0.63 0.17 0.51 DU5 225.37 205.20 733.48 3.7 1.3 5.5 60.8 0.51 0.12 0.36 DU4 180.10 154.92 637.31 3.3 1.7 10.0 62.6 0.34 0.15 0.46 DU3 73.75 52.64 216.74 2.0 0.7 3.3 38.0 0.21 0.24 0.72 DU2 91.96 67.13 273.27 2.4 1.0 4.7 40.8 0.26 0.28 0.84 DU1 75.09 43.62 141.37 1.7 0.7 2.5 46.1 0.30 0.26 0.79 DUO 316.63 122.96 455.90 6.0 1.0 6.8 54.0 0.53 0.36 1.08 All measurements are recorded in meters except average d50 bedload, which is in mm. * Measurements from GPS dataset 5 Estimated from grain size measurements in the stratigraphic sections § Measurements from stratigraphic sections and paleocurrent dataset + Estimated from average cross-set height using method from Leclair and Bridge (2001) 129 INTERPRETATION AND DISCUSSION Stratigraphic Intervals The depositional units were grouped into three stratigraphic intervals, which have implications for large-scale temporal changes in the depositional environment that resulted from changes in the relationship between sedimentation and accommodation. A summary of key attributes of the stratigraphic interval is included in Figures 36. Lower Stratigraphic Interval The lower stratigraphic interval includes DU-0 and DU-1, which show a general trend of decreasing net-to-gross, channel belt width, bankfull depth, and grain size (Figure 36). DU-0 is separated from the underlying Calico bed by an unconformity and potential sequence boundary. This boundary is irregular (with ~2 m of relief) and clasts of the Calico bed material are contained within lags near the base of DU-0. This unit contains medium- to coarse-grained sandstone and some of the thickest channels within the Bull Canyon section. The lower stratigraphic interval is interpreted as forming during low, but increasing, rates of accommodation relative to sediment supply. The distinct increase in fine-grained material (i.e., floodplain and coal mire facies) above the transition to DU-1 may have resulted from an abrupt increase in accommodation. The middle stratigraphic interval includes DU-2 and DU-3, which has the highest amount of coal mire and floodplain facies and the narrowest channel belts. This interval lacks the abundance of tidal indicators that are present in the lower stratigraphic interval. Locally, within DU-2 there is an upward coarsening pattern over approximately 17 meters, which show evidence of rapid sedimentation, limited presence of carbonaceous shale, and does not include any coals. This portion is interpreted as resulting from a local increase in sediment supply relative to accommodation that caused the formation of a crevasse delta. DU-3 contains thick accumulations of floodplain (overbank) and coal mire facies with relatively few channels, and tidal indicators are completely absent. Due to the isolated nature of channel deposits in thick floodplain accumulations, the middle stratigraphic interval is interpreted to represent a period of high accommodation formation relative to sediment supply in which the channels were relatively stable and experienced infrequent avulsions. In general, this is a continuation of the trend present in the lower stratigraphic interval, but it is much more pronounced. Upper Stratigraphic Interval The upper stratigraphic interval includes DU-4, DU-5, DU-6 and continues into the Drip Tank Member. This interval contains trends of vertically increasing average channel belt width and amalgamation, and a steady increase in median and maximum grain size. There is one exception, the average channel belt widths in DU-6 are slightly less than DU-5 (Figure 18); however, this is likely due to the truncation of channel belts by one another, and therefore is not representative of the original deposited widths. The changes in character of channel belts can be attributed to an increase in the rate of 131 Middle Stratigraphic Interval channel avulsion and/or a decrease in the sediment accommodation rate accompanied by progradation of a clastic wedge (cf., Ryer, 1993). The upper stratigraphic interval is interpreted as representing decreasing accommodation rate relative to sedimentation rate. Correlation of Fluvial and Paralic Facies A correlation along southern edge of the Kaiparowits Plateau was constructed (Figure 37), which incorporates the sections measured for this study with previously published stratigraphic sections and logged cores of Vaninetti (1979), Shanley (1991), Gallin (2010), and Gooley (2010) (A-A' in Figure 1). This correlation permits observations regarding how architectural trends in proximal fluvial facies are related to those present in more distal fluvial and paralic facies. The top of the Calico bed was used as the datum for correlation of sections. Correlations were driven by large-scale trends in outcrop architecture for published sections, and when this information was not available (i.e., logged cores), outcrop architecture was inferred