Geomorphological Evolution of Propagating Fault Tips in Extensional and Compressional Settings

This thesis is submitted in fulfilment of the requirements of the University of London for the degree of Doctor of Philosophy and the Diploma of Imperial College

Alexander M. Davis Department of Earth Science and Engineering, Imperial College of Science, Technology and Medicine London. June, 2005 Research Declaration

The research presented in this thesis is the work of the candidate named on the front page.

Signed: AOw A. M. Davis of Contents

Table of Contents 2 List of Figures 6 Abstract 14 Acknowledgements 15 Chapter 1: Introduction 16 1.1 Research Rationale 17 1.2 Methodology 18 1.3 Thesis Structure 19 Chapter 2: Tectonic 20 2.1 Introduction 20 2.1.1 Landscape Development 20 2.2 Fault Segment Evolution 23 2.2.1 Extension 23 2.2.2 Compression 26 2.3 Tectonic Geomorphology 27 2.3.1 Extension 27 2.3.2 Compression 29 2.4 Landscape Modelling 31 2.4.1 Extension 31 2.4.2 Compression 32 2.5 Geomorphic Indicators 34 2.5.1 Extension 34 2.5.2 Compression 34 2.6 Rationale 35 Chapter 3: Methodology and Data Processing 36 3.1 Introduction 36 3.2 Digital Processing Methodology 37 3.3 Data Acquisition 39 3.4 DEM generation 40 3.4.1 DEM Accuracy 42 3.5 Geo-spatial Database Compilation 42 3.6 Spatial analysis 43 3.7 Image Processing 47 2

3.8 Field Mapping 48 3.9 Interpretation 50 Chapter 4: Tectonic Geomorphology of Star , Wyoming, USA: Implications for Normal Fault Segment Growth at Extensional Relays. 51 4.1 Research Rationale 51 4.1.1 Location 52 4.1.2 Previous Work 52 4.2 Regional Background 54 4.2.1 Star Valley Geological Setting 57 4.2.2 Fault Segment Geometry 59 4.2.3 Latest Quaternary Displacements 61 4.2.4 Stratigraphy 62 4.2.5 Geomorphology 62 4.3 Tectonic Geomorphology Investigation of Star Valley 64 4.3.1 Structural Geology 64 4.3.1.1 Rear Fault Segment 64 4.3.1.2 Front Fault Segment 66 4.3.1.3 Grover Relay Structure 66 4.3.1.4 Front Segment Tip 68 4.3.2 Relay Zone Stratigraphy 70 4.2.3 Front Segment Stratigraphy 72 4.4 Star Valley Geomorphology 78 4.4.1 Footwall Topography 78 4.4.2 Drainage 81 4.5 Observations of Topography and Morphology of the Grover Relay 83 4.6 Front Segment Footwall Catchments 88 4.6.1 Longitudinal Profiles 88 4.6.2 Footwall Catchment Topography 90 4.6.3 Front Segment Topography 94 4.6.4 Catchment Comparisons 96 4.6.5 Rear Segment Catchments 99 4.6.6 Catchment Capture 100 4.6.7 Front Segment Topography Transition 103 4.7 Grover Relay Geomorphology 104 4.7.1 Front Segment Footwall Drainage 108

3

4.7.2 Grover Relay Drainage 108 4.7.2 Drainage evolution 110 4.8 Grover Relay Fault 113 4.8.1 Grover Relay Fault Topography 116 4.8.2 Hangingwall Footwall Topography Profiles 120 4.8.3 Transverse Fault Topography Profiles 120 4.8.4 Transverse Stream Profiles 124 4.8.5 Quaternary Grover Relay Fault Evolution 125 4.9 Front Segment Tip Topography 126 4.9.1 Willow Creek Drainage 130 4.9.2 Willow Creek Terraces 132 4.9.3 Front Fault Tip Fold Development 132 4.10 Results Summary 136 4.11 Discussion: Star Valley Landscape Evolution 136 4.12 Implications for Normal Fault Growth 141 4.13 Conclusion 143 Chapter 5: Tectonic Geomorphology of the Turpan Blind Thrust Fold Belt, Turpan Basin, Xinjiang, NW China. 144 5.1 Research Rationale 144 5.1.1 Location 145 5.1.2 Previous Work 147 5.2 Regional Background 148 5.2.1 Geological Setting 148 5.2.2 Geomorphology Setting 153 5.3 Flaming Mountain Geomorphology 154 5.4 Investigation of the Flaming Mountain Western Tip 165 5.4.1 Blind Thrust Fold Geometry 165 5.4.2 Stratigraphy 177 5.5 Water Gaps and Wind Gaps 182 5.5.1 1 187 5.5.2 Water Gap 2 187 5.5.3 Water Gap 3 190 5.5.4 Water Gap 4 190 5.5.5 Water Gap 5 193 5.5.6 Wind Gaps 195

4

5.5.7 Water Gap / Wind Gap Comparisons 199 5.5.8 Flaming Mountain Wind Gap Development 202 5.5.9 Water Gap / Wind Gap Transition 203 5.6 Terrain Morphology 204 5.6.1 Terrain 1: Western fold tip 204 5.6.2 Terrain 2: Minor wind gaps 206 5.6.3 Terrain: Water Gap 3-4 210 5.6.4 Terrain: Water Gap 4-5 210 5.6.5 Terrain Morphology Comparison 213 5.6.6 N-S Terrain Swath Profiles 215 5.7 Results Summary 216 5.8 Tectonic Geomorphology Model 219 5.8.1 Stage 1: Pre-Growth Bajada Fan Strata 222 5.8.2 Stage 2: Initial Fold Growth, Water Gap Development 222 5.8.3 Stage 3:Syn-Fold Growth 223 5.8.4 Stage 4: Wind Gap Development 223 5.8.5 Stage 5: Lateral Propagation and Minor Wind Gap Development 224 5.8.6 Stage 6: New Fold Tip Position 224 5.9 Discussion 225 5.9.1 Implications for Fault Segment Evolution 225 5.9.2 Comparison with Numerical Landscape Model Simulations 226 5.9.3 Fold Growth by Segmentation 227 5.10 Conclusion and Summary 229 Chapter 6: Conclusions 230 6.1 Summary of research conclusions 230 6.1.1 Major contributions of this research 230 Star Valley 230 Flaming Mountain 231 6.2 Implications for research in academia and industry 232 Chapter 7: Further Work 234 7.1 Star Valley 234 7.2 Flaming Mountain 238 Chapter 8: Bibliography 242 Appendix 1: Star Valley Data 251 Appendix 2: Turpan Fold Data 278

5 List of Figures Figure 2.1. Diagram showing morphology, including typical drainage boundaries, in the Bogda Shan mountain range NW China. The boundaries and stream networks have been generated using an ASTER derived DEM (Davis et al., 2001). 22 Figure 2.2. Sketch graphs showing typical stream slope shapes (after Burbank and Pinter, 2000) 23 Figure 2.3. Diagram showing the geometry of an en echelon normal fault (modified after Huggins et al., 1995; Cowie et al., 2000; Crider et al., 2001). 24 Figure 2.4. a) Diagram of drainage and catchment development at extensional relays, b) drainage indicating fault propagation on the Pearce fault segment, Nevada, and c) sketch diagram of drainage indicating past position of fault tip (all diagrams modified after Jackson and Leeder, 1994). 28 Figure 2.5. Diagram of water gap and wind gap generation during blind thrust fold growth (modified after Burbank et al, 1999). 30 Figure 3.1. Diagram outlining research methodology involving data preparation, analysis and interpretation. 38 Figure 3.2. Diagram showing calculation of flow direction across a DEM based on a 3 x 3 kernel window 44 Figure 3.3. Flow diagram showing DEM hydrological spatial analysis process methodology. 45 Figure 3.4. Schematic diagram showing the relationship of longitudinal stream profile to a linear regression line fitted to catchment relief 47 Figure 3.5. Diagram showing the levelling surveying method and calculation of height along a survey traverse using a laser ranging instrument. 49 Figure 4.1. Sun-shaded regional topography image showing location of the Grand Valley Fault and Star Valley study area (Piety et al., 1992) 53 Figure 4.2. a) Overthrust Belt structure map showing location of the Absaroka Thrust, b) west — east cross section based on regional seismic line (after Piety et al., 1986) showing relationship between Absaroka Thrust and the Star Valley fault. 55 Figure 4.3. a) Cenozoic fault map of the Snake Plain region in the Eastern Basin and Range Province, b) regional structural geology map of the Grand Valley Fault (modified after Piety et al., 1992). 56 Figure 4.4. Summary geology map of Star Valley (modified after Piety et al., 1992) 58 6 Figure 4.5. Perspective view of Star Valley with side panel showing west — east orientated cross section (modified after Ruby, 1957) latitude 42°47'30", showing half graben geometry. 59 Figure 4.6. Star Valley Fault map overlain on sun-shaded topography. Red tones indicate elevations over 2500m and blue tones around 1000m (modified after Piety et al., 1992). 60 Figure 4.7. Perspective terrain view of Star Valley, orientated north-east, showing range front, basin and mountain morphology 63 Figure 4.8. a) Plan view fault map of the Grover study area with fault segment overlain on a sun-shaded DEM image, b) perspective view of the geology draped over the DEM showing the relationship of the rock types to elevation 65 Figure 4.9. a) Annotated photomontage of the rear segment orientated N-S facing E, b) annotated photomontage of the front fault 67 Figure 4.10. a) Geological map of fault tip propagation fold. 69 Figure 4.10. b) formline cross section of Strawberry Creek and the area to the north showing bedding asymmetry and stereonet of poles to bedding indicating fold axial plunge 70 Figure 4.11. Geological map of the Grover Relay (modified after Ruby, 1957) 71 Figure 4.12. Plan view geological map of fault footwall basement and Long Formation conglomerates centred on Phillips Creek. 73 Figure 4.13. Photograph of Long Spring Formation polymict conglomerate in Phillips Creek. 74 Figure 4.14. a) perspective view of the geological map overlain on digital terrain model, orientated facing north, b) Cross section of Phillips Creek. 76 Figure 4.15. Photomontage of the north side of Phillips Creek showing the unconformity and stratigraphic relationships between the Long Spring Formation and the footwall basement. 77 Figure 4.16. Transverse range topography profiles showing fault segment relief. 79 Figure 4.17. Swath profiles generated using the 30 m DEM orientated parallel to range front of the eastern side of Star Valley ranging from Dry Creek in the north to southern end of the Star Valley Fault south of Smoot. 80 Figure 4.18. Star Valley geomorphology; upstream range front catchments extracted from DEM with overlain drainage. 82 Figure 4.19. Star Valley longitudinal stream profiles measured using the DEM. 84

7 Figure 4.20. Grover relay fault footwall catchment plan view map extracted from the 10m DEM using Rivertools software. The main catchment names are marked 86 Figure 4.21. Grover relay drainage map extracted from the 10m DEM. The yellow circle identifies the change in plan view drainage patterns at Phillips Creek. 87 Figure 4.22. a) Longitudinal stream profiles measured using the 10m DEM, b) composite graph show the change in profile shape south from the fault tip. 89 Figure 4.23. a) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography. 92 Figure 4.23. b) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography. 93 Figure 4.23. c) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography. 94 Figure 4.24. Front segment maximum topographic profile and drainage divide profile measured using the 10m DEM 95 Figure 4.25. a) Catchment characteristic comparison along front footwall strike. Calculations have been made using the 10m DEM using Rivertools software. 97 Figure 4.25. b) Composite graphs of swath calculations of maximum / minimum difference and minimum slope angles. 98 Figure 4.25. c) Front footwall catchment concavity (see chapter 3 for calculation method). 98 Figure 4.26. a) Longitudinal stream profiles of rear footwall, b) swath profile of 30. 100 Figure 4.27. a) Addition of Astle Creek and canyon 30 swath and difference profiles. 102 Figure 4.28. Plan view map of the Grover relay with drainage overlain on a sun-shaded DEM 104 Figure 4.29. Aspect and slope plan view maps generated using the DEM 106 Figure 4.30. Topography and stream profiles in relay measured using the DEM. 107 Figure 4.31. Relay drainage map with SW-NE transverse topographic profiles measured using the DEM showing cross sectional fluvial morphology. 109 Figure 4.32. Model of Grover relay terrain development, 1) early extensional fault development and 2) drainage migration. 111 Figure 4.32. Model of Grover relay terrain development (continued), 3) latest Holocene ruptures cross cut the drainage 112

8 Figure 4.33. a) Plan view map of the breaching fault, generated using the DEM, showing the hangingwall and footwall trace, b) annotated aerial photograph of relay fault. 114 Figure 4.34. Photographs of fault scarp in the field near to triangular facet three (TF3) (see Fig. 4.35 for location). 115 Figure 4.35. a) Plan view map of relay fault showing locations of measured fault traverses, b) annotated photograph showing relay morphology and preferred orientation of fault for presentation. 117 Figure 4.36. (a) Graph of hangingwall and footwall profiles orientated NE-SW measured using the DEM locations shown in Figure 3.35, (b) graph showing the subtraction of footwall profile from hangingwall profile resulting in a displacement profile along strike. 118 Figure 4.37. Schematic perspective sketch of the breaching fault showing the footwall and hangingwall geomorphology based on field observations and DEM profiles. 119 Figure 4.38. Footwall and hanginwall profiles measured using the EDM verifying the change in shape along fault strike 121 Figure 4.39. EDM levelled traverses measured normal to the fault scarp, b) composite graph of vertical scarp displacements along strike calculated from transverse scarp profiles. 123 Figure 4.40. Longitudinal stream profiles, measured using the EDM, of two cutting across the fault scarp, locations shown in Figure 4.35(a). 125 Figure 4.41. Composite graph of measured scarp heights with published Holocene rupture scarps heights (after Piety et al., 1992; Warren 1992). 126 Figure 4.42. Plan view map of drainage overlain on sun-shaded DEM showing fault tip topography. Blue arrows indicate drainage diversion. 127 Figure 4.43. Aspect and slope maps generated using the 10m DEM. 128 Figure 4.44. Fault tip transverse topographic profiles showing the change in topographic morphology. 129 Figure 4.45. a) Plan view orthophotograph map of Willow Creek showing change in drainage pattern to the north of the front fault and locations of differential GPS survey, b) longitudinal stream and topographic DGPS survey 131 Figure 4.46. Perspective view of the terrace features on the east side of the intrabasin high/fault tip fold structure. 132

9 Figure 4.47. a) Diagram showing early drainage development related to the monocline evolution. Yellow tone indicates deformed and uplifted Long Spring Formation conglomerates. 134 Figure 4.47. b) Diagram showing formation of terraces in monocline topography. 135 Figure 4.48. a) Conceptual sketch model of coupled catchment and drainage development in response to fault evolution in Star Valley 139 Figure 4.48. b), Continuation of conceptual sketch model. 140 Figure 5.1. Sun-shaded topography image showing location of Turpan Basin study area. 146 Figure 5.2. a) Summary of Turpan Basin geology overlain on a Landsat / ASTER mosaic image (modified after Allen et al., 1994). 149 Figure 5.2. b) North — south orientated cross — section of the Turpan Basin based on seismic section (modified after Nishidai and Berry, 1991). 150 Figure 5.3. Geological map of the Flaming Mountain (modified after Deng et al., 2000). 151 Figure 5.4. Schematic perspective cross sections across the Flaming Mountain showing the structural geometry at the a) fold tip and b) the fold centre (see Figure 5.3 for location of cross sections). 152 Figure 5.5. Sun-shaded pseudocolour DEM image indicating Turpan foreland basin topography. 155 Figure 5.6. a) Aspect image generated using spatial neighbourhood analysis of the SRTM DEM 156 Figure 5.6. b) Slope classified image generated using spatial neighbourhood analysis of the SRTM DEM 157 Figure 5.7. West — east profiles generated using the SRTM DEM showing cross sectional Bajada fan morphology. 158 Figure 5.8. North-south swath profiles generated using the SRTM DEM orientated normal to the Flaming Mountain showing fold topography. 160 Figure 5.9. Flaming Mountain composite west-east swath profiles showing maximum topography and range front profile. 161 Figure 5.10. Drainage overlain on sun-shaded SRTM DEM. 163 Figure 5.11. a) Graph showing relief of water gaps, b) graph of water gap spacing along fold axial strike 164 Figure 5.12. Geological map of the study area interpreted from ASTER and corona imagery (modified after Deng et al., 2000) 166

10 Figure 5.13. Photomontage of the eastern side of water gap 1 showing apparent dips of gentle asymmetric anticline. 168 Figure 5.14. Photomontage of the eastern side of water gap 2 showing apparent dips of asymmetric anticline. 169 Figure 5.15. a) Structural traverse map showing location of bedding measurements made along a north-south traverse of water gap 3. 170 Figure 5.15. b) stereonet plot of poles to bedding, and c) cross section of water gap 3 based on imagery and traverse. 171 Figure 5.16. Photograph of back thrusting south of water gap 3. 172 Figure 5.17. Photograph mosaic of folded lower Miocene to Upper Miocene rocks and the unconformable relationships of the Q3 and Q4 conglomerates. 173 Figure 5.18. a) Structural traverse map, b) stereonet plot of poles to bedding, and c) cross section of water gap 5 based on imagery and traverse. 174 Figure 5.18. b) stereonet plot of poles to bedding, and c) cross section of water gap 5 based on imagery and traverse. 175 Figure 5.19. Photograph of normal faulting in Lower Miocene sandstones 176 Figure 5.20. West — east profile of maximum — minimum topography difference. 177 Figure 5.21. Photograph of angular unconformity between Q1 and Q3 deposits in water gap 5 178 Figure 5.22. Photograph of Q4 (recent) fans within water gap 5 fluvial valley floor. 178 Figure 5.23. Photograph showing fill conglomerates and fan deposits 179 Figure 5.24. Levelled traverse orientated normal to water gap showing position of terraces. 180 Figure 5.25. Schematic diagram showing water gap depositional history. 181 Figure 5.26. a) ASTER 321 BCET DDS (Liu, 1991; Liu and McM Moore, 1996) image geomorphology map of the western 20 km showing locations of water gaps. 183 Figure 5.26. b) Map of water gap upstream watershed boundaries automatically generated using the SRTM DEM. The process calculates the area of the upstream watershed based on the outlet of the water gaps at the range front. The watershed calculation area is restricted to the fold topography and the Bajada fan area to the north at the range front of the Bogda Shan mountain 184 Figure 5.27. West east swath profiles generated using the SRTM DEM. 185 Figure 5.28. Greyscale Corona image of the western tip of the Flaming Mountain with overlain drainage and water channel morphology. 186

11 Figure 5.29. Geomorphology of Water Gap 1 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross fluvial valley morphology measured using the DEM. 188 Figure 5.30. Geomorphology of Water Gap 2 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross fluvial valley morphology measured using the DEM. 189 Figure 5.31. Geomorphology of Water Gap 3 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross-fluvial valley morphology measured using the DEM. 191 Figure 5.32. Geomorphology of Water Gap 4 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross-fluvial valley morphology measured using the DEM. 192 Figure 5.33. Geomorphology of Water Gap 5 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal channel profile measured using the DEM, c) transverse profiles showing cross fluvial channel morphology measured using the DEM. 194 Figure 5.34. Geomorphology of Wind Gaps a) greyscale Corona image showing geomorphology in plan view, b) swath profile of longitudinal channel profile measured using the DEM c) transverse profiles showing cross fluvial channel morphology measured using the DEM. 196 Figure 5.35. a) Longitudinal valley profiles b) west-east EDM levelled traverse, c) field photograph montage looking NE to SE,. 197 Figure 5.35. d) aspect and e) slope maps generated using the SRTM DEM. Watershed boundaries have been added to show the location of the drainage divides. 198 Figure 5.36. Composite graph showing N-S orientated water gap and wind gap longitudinal valley profiles. 200 Figure 5.37. a) Plot of water gap channel width at the southern range front, mid channel and north of channel, b) graph plot of maximum and mean difference N-S water / wind gap relief. 200

12 Figure 5.37. c) Diagram showing upstream catchment area (km2) of outlet points positioned at the range font of the water and wind gaps. Catchments automatically generated using SRTM DEM. 201 Figure 5.38. Geomorphology of western fold tip terrain, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM. 205 Figure 5.39. Geomorphology of the terrain between water gap 1 and 2, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM 207 Figure 5.40. a) Photograph of preserved fluvial channel, b) EDM levelled longitudinal profiles of channels. 208 Figure 5.40. c) aspect and d) slope maps of terrain 2 generated using the SRTM DEM. Watershed boundaries have been added to show the location of the drainage divides. 209 Figure 5.41. Geomorphology of the terrain between water gap 3 and 4, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM 211 Figure 5.42. Geomorphology of the terrain between water gap 4 and 5, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM 212 Figure 5.43. Summary images showing the similarity between the plan view geometry of the minor wind gaps and drainage to the west. 214 Figure 5.44. Schematic diagram showing the change in topographic relief N-S profiles to the east. 215 Figure 5.45. Summary geomorphology map constructed using DEM derived watersheds and fieldwork observations. The watersheds and drainage network is displayed over a hillshaded SRTM DEM. 218 Figure 5.46. a) Conceptual sketch model of drainage and catchment reconstruction in the westernmost 20 km of the fold 220 Figure 5.46. b) Conceptual sketch model of drainage and catchment reconstruction (continued). 221 Figure 6.1. Perspective view of the Pearce and Tobin fault segments in Pleasant Valley using a 10m DEM. 237 Figure 6.2. a) Corona greyscale image showing fold morphology and diverted drainage , b) west — east laser ranged profiles of the north fold segment 239

13 Abstract Quaternary landscape evolution in active continental extensional and foreland basin fault segment arrays produce diverse geomorphic features resulting from the interplay between tectonics, erosional processes and climate. Numerical simulations of landscape development predict characteristic terrain features corresponding to varying tectonic and erosional model parameters. Therefore, comparison of numerical model results to modern day terrain enables an interpretation of the spatial and temporal terrain evolution of fault segments. The Grover Relay, a 4 km wide en-echelon overlap of the Star Valley normal fault segments, Wyoming, exhibits Holocene scarps. Geomorphological and stratigraphic observations has allowed the development of a catchment capture model indicating that the fault segments had initial rapidly grown into an overlap position at an early stage and continued to grow by a vertical displacement gradient which facilitated footwall catchment migration and capture of relay streams. Mapping of a new 2 km long fault segment indicates that the relay is in hard linkage. The western twenty kilometres of the Flaming Mountain, part of the Turpan foreland basin en echelon blind thrust anticline fold belt, Xinjiang China, shows evidence of Quaternary deformation. Mapping of water gaps, wind gaps and drainage diversions indicates that the blind thrust anticline tip has laterally propagated 5 km to the west since the Upper Pleistocene. The upstream steep regional slope is interpreted to have an effect on the temporal development of the anticline by maintaining water gaps across the developing fold. The change in rheology of the growing fold, by continued and removal of away from the fold topography by the water gaps, results in a change in deformation. Identification of key geomorphic features in these two contrasting tectonically active continental basin settings has allowed an interpretation of the spatial and temporal structural growth, which has implications for dispersal pathways into continental basins.

14 Acknowledgements I would like to thank Jian Guo Liu and Sanjeev Gupta for their guidance, enthusiasm and supervision during the course of the research. I am indebted to Nancye Dawers and Alex Densmore for discussions and providing invaluable insights to tectonic geomorphology in the field in Star Valley. Personal thanks goes to Liu for spending four weeks in the field and providing valuable translating services in Turpan. I acknowledge the following for awarding fieldwork funds for the Star Valley fieldwork: • Geological Society London fieldwork funds • University of London Central Research Funds • Imperial College Frank Merricks award The forestry office and County Planning office in Afton are acknowledged for providing advice and use of office. I would like to thank the people of Afton, notably Harry Keyfeuver for their friendly support and advice on the cougars, bears and deer near Grover. I acknowledge the following for providing support and funds for fieldwork in Turpan, NW China: • Imperial College Expedition Society • Royal Geographical Society Expedition funds • Seismological Bureau, Urumqi, Xinjinag, NW China The Seismological Bureau are thanked for providing essential logistic help and permits to carry out fieldwork. Personal thanks goes Mrs Wei in the Seismological Bureau, Urumqi.

I was fortunate to have enthusiastic field assistants in both study areas. Thanks go to Nick, Joe and Matt. My time at Imperial has been helped socially with many beers and geological discussions with Philippa, Richard, Craig, Gary, Kerry, Cat, Paul and far too many others to mention. Thanks guys! Dick Selley is thanked for encouragement and spurring me on at the last.

Finally, Elleni is thanked for putting up with the highs and the many lows over the last five years. Her patience and support has been invaluable in making it though the thesis process in one piece.

15 Chapter 1: Introduction Tectonic geomorphology is the study of terrain evolution in response to active tectonic displacement (Burbank and Pinter, 1999; Burbank and Anderson, 2000; Gupta and Cowie, 2000; Keller and Pinter, 2002). Combination of the rates of structural deformation, erosion and climatic processes constantly modify landscape morphology. Therefore, identification of geomorphology in active tectonic regions allows a reconstruction of drainage and catchments along fault induced range topography enabling an interpretation of the fault growth history (Leeder and Jackson 1993; Burbank et al., 1999). Comparison of geomorphic features along strike of faulted terrain such as drainage patterns, catchment morphology, longitudinal stream profiles, and topographic profiles, allows an interpretation of the relative lateral propagation of fault segments (Leeder and Jackson, 1994; Burbank et al., 1996; Mueller and Tailing, 1997; Azor et al, 2002; Commins et al, 2005).

Erosional processes, most notably climatic and local hillslope processes, over time modify the landscape removing the key geomorphic features that indicate fault lateral propagation. Due to the temporal evolution and of landscapes it is important to select areas for study where fault generated geomorphology is not removed by climatic processes and long term local erosion. Active tectonic regions over Cenozoic geologic time are therefore more likely to preserve tectonic geomorphic features. It is therefore important to understand geomorphic processes of features generated by tectonic displacement, fault segment propagation evolution, and methods for classifying terrain.

This research focuses on two continental en echelon fault segment arrays in mountain basin and range topography. The Star Valley fault, Wyoming, U.S.A. is an extensional normal fault consisting of two fault segments with right en echelon relay linkage geometry (Piety et al., 1992). Recognition and reconstruction of catchments and drainages along fault strike indicates fault growth evolution. The approach and methodology applied to mapping catchments and drainage in this classic region, using digital elevation models (DEM), satellite remotely sensed imagery and fieldwork is applied to the second case study. The Turpan foreland basin, Xinjiang, NW China, is a less studied area for tectonic geomorphology research (Bihong et al., 2003). West-east en echelon blind thrust fold segments interact with Bajada fan and fluvial streams of this foreland basin. 16 The most efficient method of mapping intermediate to regional scale geomorphology is the processing, analyses and interpretation of remotely sensed imagery from airborne and spaceborne satellite sensors platforms (Davis et al., 2000; Goldsworthy and Jackson, 2000; Davis et al, 2001). Stereo images recorded on these platforms can be used to generate digital elevation models (DEM) or terrain models that are a digital grid representation of the landscape. DEM combined with multispectral imagery are an effective tool for visualising and quantifying geomorphology (Mason et al., 1998; Davis et al., 2000; Bihong et al., 2003). The interpreted imagery is also an excellent tool for planning targeted fieldwork and making efficient use of limited fieldwork time. Integrating fieldwork with imagery in a spatial database within a Geographical Information System (GIS) is an effective methodology for mapping geomorphology.

1.1 Research Rationale

The temporal evolution of terrain in tectonically active regions has relied on palaeoseismic studies where faulted terraces and have been trenched to assess the repeat period of earthquakes (Schwartz and Coppersmith, 1984; Piety et al., 1986). Analysis of the landscape involving digital imagery and numerical modelling has identified geomorphic features that indicate fault evolution (Jackson and Leeder,1994; Burbank et al., 1999; Champel et al., 2002; Densmore et al., 2003). The aims of the research are based on the identification of key geomorphic features that indicate fault growth style and propagation:

• To use geomorphology to reconstruct the spatial and temporal evolution of geologic structure propagation at fault segment tips.

• To compare fault tip evolution in continental extensional and compressional settings based on two case study areas; Star Valley, Wyoming, U.S.A. and Turpan Basin, Xinjiang, NW China.

These two study areas are in tectonically active regions, both indicating evidence of Quaternary deformation based on palaeo-seismic studies (Piety et al., 1992; Allen et al., 1994). An earthquake measured with a magnitude 4.4 on the Richter scale, occurred on

17 October 21st, 2002 near Alpine just north of Star Valley (Wyoming Geological Survey). Also, the Turpan Basin is in an active area of mountain building with regular earthquakes related to active uplift of the Bogda Shan Mountains. Effective and efficient mapping of geomorphology involves a multi-discipline approach to collecting, analysing and interpreting the data. The objectives for this research focus on the use of digital data integrated with targeted fieldwork to map the spatial relationships of geomorphology at fault tips in the two case study areas. The main objectives are:

• To map geomorphology and geology at fault segment tips using remotely sensed imagery, digital elevation models and fieldwork.

• To quantatively classify fault-generated topography using DEM.

• Identification of key geomorphic features in fault generated topography that demonstrates fault tip propagation.

• Construction of geologic cross-sections to indicate structural geometry.

1.2 Methodology The integrated approach involves the combination and manipulation of multi-data into a geospatial database or GIS. The following list outlines the simplified methodology workflow for compiling the database with acts as a platform for geomorphic mapping, analysis and interpretation:

• Data Acquisition • DEM generation • GIS compilation • Image Processing • Spatial Analysis • Interpretation • Development of conceptual models of temporal landscape development

18 1.3 Thesis Structure

The structure of the thesis is divided into the following:

• Chapter 1: Introduction to the research and outline of the thesis • Chapter 2: Tectonic Geomorphology background • Chapter 3: Methodology and Data Processing • Chapter 4: Tectonic Geomorphology of Star Valley, Wyoming • Chapter 5: Tectonic Geomorphology of the Flaming Mountain, Turpan • Chapter 6: Conclusions • Chapter 7: Further work • Chapter 8: Bibliography

Chapter 2 is a literature review of tectonic geomorphology in extensional and compressional basins with an emphasis on the recognition geomorphic features that indicate fault development and propagation. Chapter 3 outline the methodology employed in carrying out the research including the data generation, data sourcing, data compilation, data processing and analysis. Chapters 4 and 5 are the main research chapters. Chapter 6 is the conclusion chapter, which is a summary of the findings in Chapters 4 and 5.

19 Chapter 2: Tectonic Geomorphology

2.1 Introduction The identification and characterisation of terrain together with an understanding of surface processes is essential when analysing geomorphology in tectonically active terrain (Burbank and Pinter, 1999; Gupta and Cowie, 2000). Temporal evolution of the landscape ranges from daily transport of fine-grained particles in streams and hillslope processes, to occasional catastrophic landslides. It is therefore essential to recognise geomorphic features that record the response of erosion to tectonic displacement. Measurement and interpretation of drainage patterns, longitudinal stream profiles, catchment morphology and topographic profiles record the change of geomorphic processes to tectonic activity.

2.1.1 Landscape Development Landscape development results from the interaction of long and short-term tectonic and climatic processes. The latter consists of various physical erosional processes that occur in environments ranging from desert to glacial over differences in altitude. Water erosion is the common physical process that modifies terrain. Understanding water transportation pathways from high to low elevations gives insights into how the terrain is being modified by climatic and tectonic processes.

In order to qualify and quantify terrain morphology various field measurements and remote sensing techniques allow a classification of terrain. DEMs allow both an automatic and semi-automatic classification of terrain using a series of spatial grid functions that mimic the flow of water across a surface (Jenson and Domingue , 1988).

The landscape surface can be classified by the way water flows across its surface (Fig. 2.1). Areas where water drains from high to low elevations to a common outlet define drainage systems. These outlets or pour points define basins, otherwise known as a drainage basin, watershed, catchment and contributing area. The boundaries between catchments i.e. drainage divides, define different contributing areas. There is difference between topographic divide e.g. the different in elevation of terrain on either side of a mountain range front, and the drainage divide which defines the difference in flow direction of a mountain range. In most cases the two are coincident e.g. in steady state

20 mountain ranges where erosion and tectonic effects are in equilibrium, but where a mountain range is still developing, drainage divides, depending on the degree of headward migration into the range can differ to the topographic divide (Harbor, 1997).

Drainage development involves stream flow, across landscapes from high elevations to low elevations, finding the most efficient and easiest route resulting in a network of drainage patterns. Long term depending on watershed area, stream slope, bedrock rheology and rate of incision (Burbank et al., 2000), incise the terrain, forming drainage networks. Relative rapid changes in landscape morphology resulting from tectonic uplift often results in the re-organisation of drainage. Streams that are unable to incise through the uplifted fault-induced topography are deflected around the terrain to lower elevations. Stream patterns at the tips of normal fault segments and blind thrust anticline folds show deflected drainages around the structures (Jackson et al., 1994, Burbank et al., 1999, Burbank et al., 2000, and Keller et al., 2002).

Longitudinal stream profiles of main trunk streams in catchment basins indicate the degree of fluvial incision (Burbank et al 2000, and Keller et al 2002). In tectonically active areas the shape of these profiles indicates the temporal response of erosional processes in the catchment and the ability of the stream to incise the uplifted topography. Figure 2.2 is a sketch of the shapes of longitudinal stream profiles. Concave slope profile shape is typical of mature streams that have reached equilibrium or "steady state" and indicate long-term incision rates often as a result of climatic processes. Inflections in these curves, or , indicate the streams response to base-level changes at the range front that can be a result of changes in lake levels or fault displacement. Over time these knickpoints migrate upstream as long-term erosion removes this change in . Linear slope profiles indicate intermediate incision rates or the temporal inability of the stream to incise the topography. This may result from a small catchment in tectonically active areas where the stream has not been able to respond to tectonic uplift. Convex slope profiles indicate low incision rates and the inability of local hillslope processes to erode the terrain. In tectonically active areas these slopes often indicate the "fossilisation" of streams in uplifted terrain.

21 Watershed (Basin, Catchment, Contributing area)

Watershed boundaries (Drainage divide)

Catchment outlet (pour point)

Figure 2.1. Diagram showing drainage basin morphology, including typical drainage boundaries, in the Bogda Shan mountain range NW China. The boundaries and stream networks have been generated using an ASTER derived DEM (Davis et al., 2001).

22 ♦ A

C 0 ea Concave Linear Convex 0 w

Longitudinal stream length distance

Figure 2.2. Sketch graphs showing typical stream slope shapes (after Burbank and Pinter, 2000).

The enlargement of mountain range catchments or watersheds depends on the ability of to erode bedrock and transport sediment to lower elevations. This depends on a number of variables such as bedrock rheology, catchment size, slope of catchment, change in relief and water (Burbank et al., 2000; Densmore et al., 2003; Champel et al., 2003). Headward migration and enlargement of catchments depends on the temporal interaction of these variables. Catchment stream channel slope is important in the headward migration of catchments (Densmore et al., 2004).

2.2 Fault Segment Evolution

Fault segment evolution in terms of growth and evolution style, in extension and compressional regimes, vary in the way segments interact and coalesce. Models of fault segment growth, lateral propagation and linkage have been made by field observation, interpretation of seismic data and numerical modelling (Keller et al., 2002; Cowie et al., 2000, and Gupta et al., 2000). Mechanisms of fault segment growth and interaction involving variations in footwall displacement, hangingwall subsidence, and structural geometry is important when considering the coupled geomorphic response to uplift.

2.2.1 Extension

The development of fault segment arrays in a regional stress field depends on rock rheology, existing tectonic fabric and the interaction and coalescence of individual 23 segments. Fault evolution is quantified as the ratio of the maximum displacement (D) to the length of fault strike (L) (Scholz and Cowie 1990). Typically the maximum displacement is located at the fault midpoint with the displacement decreasing to zero at the tip forming a "bell shaped" curve profile (Scholz and Cowie 1990; Cartwright et al., 1995). An important primary mechanism of fault growth is by fault linkage / coalescence of multiple segments (Peacock and Sanderson, 1991, Cartwright et al., 1995; Cowie et al., 2000, Commins et al., 2005) and not by fault tip fold propagation.

The geometry of fault segments is taken from Cowie et al., 2000 and Crider 2001 (Fig. 2.3). In an overlapped en echelon configuration the front and rear segments have a separation i.e. the normal distance between segments, and overlap length i.e. the along strike distance common to both segments, forming a relay zone (Huggins et al., 1995; Cowie et al., 2000; Crider et al., 2001).

Overlap length Rear Segment 14-

Overlap 1 relay Separation zone.

Front Segment

Figure 2.3. Diagram showing the geometry of an en echelon normal fault (modified after Huggins et al., 1995; Cowie et al., 2000; Crider et al., 2001).

Palaeoseismic studies on faults particularly on the Wasatch Fault have provided insights into how faults grow and interact. Interpretation of fault scarp colluvium trench exposures, have indicated preserved displacements corresponding to individual earthquake events (Schwartz and Coppersmith, 1984). This has allowed the timing of fault growth, including repeat period of the faulting during the Quaternary. Further work on the Wasatch fault zone bounding the eastern boundary of the Basin and Range province has indicated that it is the longest and continuous active normal fault in the U.S.A. (Bruhn et al., 1987; Machette et al., 1991). Field and palaeoseismic investigations indicate that the fault along length is divided into ten discrete segments

24 with varying slip rates and recurrence intervals. The linkage style at the segment boundaries is divided into several categories including en echelon, cross fault, salient and gaps. Typical fault segments length in the Basin and Range range from 20 to 25 km. However, the central part of the Wasatch Fault indicates actives lengths of 52 km to 21 km in the distal parts.

Models of temporal fault evolution indicate scenarios of early and late linkage scenarios of fault segment linkage growth involving variation in the rate of lateral propagation, differential footwall uplift and linkage geometry (Cowie et al., 2000). Early linkage of fault segments at an early stage with no significant footwall uplift involves a decrease in displacement / length ratios. This linkage produces large hangingwall basins. Late linkage involves the lateral propagation and interaction of fault segments with significant footwall uplifts producing intrabasinal highs near to the overlap of the faults increasing in displacement / length ratios. This linkage type results in the development of small basins before breaching.

Late linkage can be sub-divided into two scenarios depending on fault segment growth geometry. The first is a gradual propagation of segments towards each other with an increase in length and displacement. Interaction of segments forms an intrabasinal high near the overlap. Hard linkage results in breaching of the relay between the segments. The second is the rapid lateral growth of the segments with an increase in length with little displacement forming the final overlap fault geometry. Differential displacement then occurs with no change in length with the formation of intrabasinal highs near the overlap. Again, hard linkage results in the breaching of the relay linking the en echelon fault segments. These two scenarios of late linkage involve two types of propagation histories depending on the displacement and length evolution and relative rate of displacement of the individual fault segments.

Fault interaction can affect earthquake rupture sequences involving "jumping" of seismic events across fault segment (Gupta and Scholtz, 2000). They have developed a model of interaction where growing fault segments respond to reductions in shear stress in adjacent segment with an increase in displacement at the tip. Also, a stress reduction region around faults prohibits segment propagation and initiation. They have developed a conceptual model of fault growth using a numerical model combined with field observations. They define a pre-interacting segment with symmetrical displacement /

25 length (D/L) profiles. Weak interaction between segments results in an underlapping geometry where an increase in shear stress at the tips may drive fault towards each other until a stress shadow is reached. Moderate interaction results when lateral propagation decreases when the tip enter the stress shadow. This results in an increase in displacement at the tip. Strong interaction of the segment results in an increase in symmetry of the displacement profile at the tips producing individual profiles that are dramatically different to individual D/L profile. Linkage results when propagation stops and minor structure develop in the overlap and the summed profile resembles that of an individual fault. Coalescence is the next stage in development where linkage is complete and propagation of the distal parts of the two segments occur.

Crider (2001) uses field observations made in the Oregon Basin and Range near Summer Lake, combined with numerical modelling, to form models of oblique slip and normal fault linkage geometry. The relationship between en echelon step sense, oblique- slip sense and position indicate the position in the relay ramp the linking faults / breach will occur. If the step sense and oblique are the same (right step, right oblique step) then the segments hard link in the lower part of the relay and there is the possibility for deformation ahead of the front segment and in the relay. If the step sense and oblique slip are opposite then the model predicts that the breach position will be in the upper part of the ramp. The model demonstrates that oblique slip influences fault growth and linkage.

Analysis of en echelon normal faults indicates how fault segments interact with each other over time with a change in displacement profiles (Gupta and Scholz, 2000; Cowie et al., 2000). As a result, en echelon fault arrays are the ideal focus for identifying tectonic geomorphic signatures.

2.2.2 Compression The development of intermontane thin-skinned folds in response to regional compression, results from the initiation, growth and lateral propagation of thrust faults (Mueller and Suppe, 1997; Champel et al., 2002). The relief above thrust faults is generated by fault propagation and fault bend folding (Suppe and Medwedeff, 1990). Fault propagation folding develops in the hangingwall of a blind thrust. The geometry

26 of the fault propagation folding is defined by the detachment depth, the detachment angle and the ramp angle (van der Beek et al., 2002). When the fault tip reaches the surface the deformation style changes to fault bend folding. The geometry of these folds is defined by axial surfaces whose angles result from the detachment angle and the ramp angle. Segmented thrust faults show similar overlap geometries to extensional relays where fault displacements decrease rapidly at relays zones with an increase in fold amplitude (Nicol et al., 2002).

2.3 Tectonic Geomorphology The physical processes that influence the three-dimensional shape of terrain vary in areas that are tectonically active relative to long-term climate erosional processes (Burbank et al., 2000 and Keller et al., 2002). The preservation of features such as fault scarps and wind gaps, indicate the style of deformation. The tectonic geomorphic features in extensional and compressional regimes are in some aspects similar in the way they have been formed, however there are distinct differences. The following sections list the features in both these tectonic settings.

2.3.1 Extension

Conceptual models have been developed for geomorphic evolution in active tectonic areas in response to fault segment growth (Burbank et al., 2000 and Keller et al., 2002). In areas of en echelon fault relays key geomorphic features indicate the lateral propagation of the fault segments. Figure 2.4(a) illustrates a sketch model of the focusing of drainage along relay ramps, as two fault segments propagate towards each other (Leeder et al., 1987; Jackson and Leeder, 1994). Drainages are deflected around the tips of the propagating fault segments. The model indicates that with progressive focussing of drainage into the relay, large sedimentary fans occur at the base of the relay ramp with associated large upstream catchments (Gawthorpe et al., 1987; Jackson et al., 1994; Burbank et al., 2000 and Keller et al., 2002). Jackson and Leeder, 1994, indicate a different coupled drainage and fault propagation evolution to the above model based on drainage patterns and geomorphology of the Pearce and Tobin fault segment en echelon relay in Nevada, U.S.A..

27 Large Catchment

Drainage Focused Along Relay

Tobin Segment

Relay Pearce Segment

Alluvium

Pearce Fault —IP! Tobin Fault Segment Tip Segment I

Drainage deflected around fault tip

Past Position of Pearce Segment Tip

Figure 2.4. a) Diagram of drainage and catchment development at extensional relays, b) drainage indicating fault propagation on the Pearce fault segment, Nevada, and c) sketch diagram of drainage indicating past position of fault tip (all diagrams modified after Jackson and Leeder, 1994). 28 Figure 2.4(b) is a sketch perspective view of the drainage and geomorphology of the Tobin and Pearce fault segments (modified after Jackson et al., 1994). Notice that the drainages cut across the relay and are not focussed along it. Fan development south from the Pearce segment tip indicates temporal incision of drainages that cross the relay i.e. the hangingwall of the Tobin fault. Drainage development at fan 1 indicates incision into topography generated by the latest lateral propagation of the Pearce segment. At fans 2 and 3 inherited streams resulting from northward lateral propagation, deeply incise the footwall resulting in . Fan 3 deposition next to the fault scarp indicates burial of earlier fan 2 type deposits illustrating temporal development of streams and fan north along the fault segment. Figure 2.4(c) summarises the above indicating that drainages that cut across the relay show the temporal and spatial position of the Pearce fault tip with relative lateral propagation to the north.

The geomorphic response to tectonics results in the re-organisation of drainage and change in local watershed dynamics (Eliet and Gawthorpe, 1995; Goldworthy and Jackson, 2000; Zelilidis, 2000). Drainage is commonly deflected around growing tectonic structures and in zones of overlap basins with large upstream catchments focus drainage along the relay zone. In segmented fault arrays re-organised drainage often produces reversed drainage where streams have responded to differential uplift along fault segments (Goldworthy and Jackson, 2000; Zelilidis, 2000).

2.3.2 Compression

In areas of fold growth related to thrust fault propagation and fault bend folding drainages and other key geomorphic features indicate fold growth and lateral propagation (Mueller and Suppe, 1997; Burbank et al., 1999; Burbank et al, 2000; Keller et al, 2002; and Champel et al., 2002).

The temporal interaction of stream development related to the growing fold results in various drainage patterns at the fold tip, in the upstream regional slope perpendicular to the fold axis and streams that cut across the growing fold. Figure 2.5 (modified after Burbank et al., 1999) is a model of drainage and catchment evolution based on the geomorphology of en echelon Naryn Basin blind thrust anticlines, Tien Shan, Kyrgyzstan. Like the normal fault model, drainages are deflected around the fold nose.

29 Antecedent streams that are able to incise through the growing topography form water gaps. With progressive lateral growth deflection of the drainages continues. The be- heading of water gaps due to uplift results in the deflection of drainages and results in the formation of wind gaps. These wind gaps become incorporated into the folded terrain. Continued lateral propagation results in the stepwise elevation of wind gaps away from the fold nose. The formation of wind gaps (Burbank et al., 2000, Keller et al., 2002) result in convex longitudinal stream profile shape as the original water gap concave to linear profile morphology is folded and hillslopes processes are unable to remove the effects of uplift and folding.

111 Future Anticline

Undeflected Drainage

Water Gaps

,,,,--_,-A,11116.0:11111111110.11111L

Deflected Wind Gap Drainage

I\ \

Figure 2.5. Diagram of water gap and wind gap generation during blind thrust fold growth (modified after Burbank et al, 1999).

These conceptual models indicate that key geomorphic features most notably drainage patterns and a decrease in range front catchment size towards the fold terminus, indicate fault segment propagation. It is therefore essential to relate these models to fault growth, displacement and lateral propagation evolution.

30 The plunge panel of a thrust related fold is a set of alluvial surfaces that form the relatively horizontal crest between the two fold limbs (Mueller and Tailing, 1997). The stepwise decrease in elevation of the plunge panel of the eastern end of the Wheeler Ridge anticline is interpreted as the topography of three thrust fold segments. Alluvial ridges suggest a geomorphic response to kink band migration as the fold develops at the range front (Mueller and Tailing, 1997). These ridges form a stepped elevation pattern in the range front of the fold.

2.4 Landscape Modelling Numerical model simulations of landscape development over time combine the latest research on fault growth development kinematics and erosional processes into a tectonic geomorphic model. The advantage of using such models is that the user can vary the input parameters of the model. By changing only one input and keeping all the other inputs constant, numerous model runs can produce a range of landscape scenarios e.g. by increasing the rate of fault propagation or increasing the amount of erosion. Such iterative experiments can be used to define a series of primary geomorphic features that are produced in certain tectonic and erosional regimes.

2.4.1 Extension

The Basin and Range region in the Western U.S.A. is a classic area for observing extensional basin landscape features. The excellent preservation of geomorphology has been used to define models of temporal fault evolution (see 2.3.1 above). Numerical model simulation of this area using the Zscape model (Densmore et al., 1998; Ellis et al., 1999) has produced landscape features similar to those observed in the field. The Zscape finite-difference model uses the three-dimensional effects of extensional fault kinematics and surface processes including hillslope and landsliding to generate a 3D computer generated landscape based on varying initial model parameters. Zscape has been used to explore different scenarios of normal fault block terrain evolution. In a series of experiments the role of bedrock landsliding incorporated into the model runs produced simulated terrain that closely resembled field observations (Densmore et al., 1998). Bedrock landsliding at the range front resulting from base level change is an efficient process for producing the type of catchment morphology observed in the Basin and Range. Modification of the Zscape model has included new insights into how fault segment arrays develop over time (Densmore et al., 2003) using modelling carried out by Cowie et al., 2000 (see 2.2.1 above). 31 Lateral propagation models described above (demonstrate that streams cutting across the relay zone indicate the relative temporal and spatial position of the fault segment tip (Jackson et al, 1994). Mechanisms of fault lateral propagation and linkage described in 1.3 indicate that there will be a range of geomorphic features that develop in response to the end member cases described in the late linkage scenarios (Cowie et al, 2000). The first fault linkage model scenario indicates fault segment length growth towards each other with a gradual increase in footwall displacement. Catchment and drainage response would produce similar to Figure 2.4(a) where drainage has enough time to respond to fault development and become focussed along the relay. Scenario two indicates initial lateral propagation is fast with the fault reaching its final overlap geometry. Footwall displacement then occurs with no change in length with a relative decrease in differential displacement along the segment, when compared to the first scenario. If we consider fast fault displacement of this second scenario then drainage in the relay has no time to respond to the displaced topography. Densmore et al, 2003, indicate that the dominant geomorphic processes relating to fast footwall displacement results in erosion development in the front fault footwall catchments. Catchment development and migration into the footwall and relay results in rear catchment capture by the front fault catchment. This footwall catchment capture hypothesis will be tested by analysing the geomorphology in the Star Valley case based on digital imagery and fieldwork.

2.4.2 Compression

Numerical models of landscape development in regions of fault-propagation folding attempt explain the first order variations in drainage patterns and sediment dispersal patterns to foreland basins. Two models are explained here, which focuses on 1) how the thrust detachment angle at depth controls drainage development on growing folds (van der Beek et al., 2002; Champel et al., 2002) and 2) how the three dimensional effect of erosion related to regional slope influences fold growth (Simpson, 2004). Numerical model experiments within simulated regions of fault propagation folds indicate that the thrust detachment angle strongly influences the drainage devlepment along the fold and the upstream regional slope (van der Beek et al., 2002; Champel et al., 2002). The numerical model used is a modified version of Cascade surface model (Braun and Sambridge, 1997) and is described in detail in Champel et al., 2002. It is a combination of the fault related folding kinematics and surfaces processes including

32 fluvial, hillslope transport and bedrock landsliding). The results show a marked change in drainage pattern development when the detachment angle of the thrust is change from horizontal to an inclined angle. Model runs where the detachment is horizontal show marked features. Slow convergence rates of propagating thrust folds result in the formation of water gaps and wind gaps. In a fast scenario, there is not enough time for water gap development with a series of wind gaps developing and diverted drainage around the fold nose. However, when the detachment angle is inclined this dramatically influences the drainage development on the growing fold. The propagating thrust produces an inclined slope, in the direction of propagation, upstream from the fold. The drainages are diverted in the regional slope behind the fold, shielding the fold from incision by these rivers. The diverted drainages form patterns in the direction of propagation in the regional inclined slope and, due to contined erosion, along the back limb of the growing fold. The main geomorphic features are that no water gaps or wind gaps are formed.

Simpson, 2004, uses a coupled three dimensional mathematical model to explore if erosion influences fold development. The model uses thin-skinned deformation coupled with fluvial and hillslope erosion processes. Two parameters (regional topographic slope and ratio of tectonic deformation to fluvial processes) are varied in the iterative models runs to record the effect of sediment dispersal patterns during fold growth. When the initial regional slope is flat and the rate of erosion relative to deformation is slow the folding is "two-dimensional" as it is not affected by erosion because the sediment loads created by erosion are imposed after fold growth. The result is that small scale drainage networks develop transporting sediment away from the developing anticlines. The regional slope is relative high (2 degrees) and the rate of deformation / erosion is high result in the development of large scale transverse drainage networks . These are maintained during folding and is able to influence the folding resulting in localised amplified structures resulting from sediment load removal during deformation. An interesting geomorphic feature resulting from this is that the greatest structural and topographic fold amplitude are associated with large transverse drainages. Relatively intermediate initial regional slopes and rate of deformation, erosion gives rise to a range of en echelon structures with complex drainage networks.

33 2.5 Geomorphic Indicators Remote Sensing, including digital elevation model analysis, together with field observations and numerical model terrain simulation have indicated the diverse geomorphic features that can form in extensional and compressional tectonic regions. The numerical model simulations have provided insights into how fault segment interact and grow over time with the identification of key geomorphic feature resulting from erosion in response to the tectonics. The day terrain is a snap-shot of the temporal evolution of a fault segment array. The model simulations allow the identification of key geomorphic features relevant to a specific tectonic scenario of temporal fault development. The observation of Quaternary structures and the identification of these key geomorphic features allow a new interpretation of the evolution state of actively growing extensional and compression fault segment arrays.

2.5.1 Extension In an extensional fault array, especially zones of overlap, characteristic geomorphic features indicate fault growth style evolution. Numerical model experiments employing varying temporal fault growth mechanisms result in characteristic terrain features (Densmore et al., 2003). The characteristic geomorphic features to be identified in normal fault arrays are as follows:

• Map range front geomorphology (triangular facets, faulted terraces / alluvial fans) • Map topography along fault segment strike. • The identification and analyse of relay zone geomorphology including the identification catchment capture events (Densmore et al., 2003). • Map change in footwall catchment morphology along fault strike. • Map drainage patterns in front and rear segments and in the relay • Map longitudinal stream profiles of catchments

2.5.2 Compression In a foreland basin thrust related fold array, like the extensional regime examples outlined in 2.5.1, characteristic geomorphic features indicate tectonic development. The characteristic geomorphic features to be identified in thrust folded terrain is as follows:

34 • Antecedent channels that traverse the fold (water gaps) • Preserved channels within uplifted terrain (wind gaps) • Mapping of drainage along the fold. • Diverted drainage at the fold nose and in the regional slope adjacent to the fold. • To map drainage patterns in regional slope upstream from fold. This implies detachment angle of thrust (Champel et al., 2002). • Map longitudinal water gap / wind gap profiles

2.6 Rationale

Field observations on active fault arrays involving palaeoseismic dating result in the assessment of earthquake potential in a given area by estimating repeat period events. These observations fail to identify the temporal style of deformation. Fault segments interact with one another differently depending on the age of the deformation and the proximity of interaction. The identification of tectonic geomorphic features along the structure can indicate the style of deformation. Numerical model simulations have shown that primary geomorphic features indicate the type of deformation in both extensional and compressional regimes. The geomorphic indicators listed in section 2.5, combined from previous field observations and augmented by numerical landscape simulation results, are applied to two the two study areas introduced in chapter 1. Both areas are in intermontane basin terrains, which have segmented fault arrays and active fault activity. From the literature en echelon faults are ideal areas for measuring the effects of geomorphologic processes on fault growth along structural strike (Gupta and Scholz, 2000; Cowie et al., 2000).

The study will focus on classifying the terrain using field observations and digital data to establish a model of integrated tectonic and erosional development. These models will be used to suggest the possible structural evolution during the Quaternary.

35 Chapter 3: Methodology and Data Processing

3.1 Introduction The multidiscipline approach to mapping the spatial relationships of geomorphology at fault tips involves the integration of many different data types. As a result, all data need to be spatially referenced in a Geographical Information System (GIS) database in order to manage and interpret the information efficiently. Data management and digital processing involves numerous procedures within several different software packages. The following table lists the image processing and GIS packages used to process and interpret the data.

Table 3.1. Software packages used in the research. Software Package Description ERMapper General image processing of satellite imagery and digital elevation models. ERDAS Virtual GIS Perspective visualisation PCI Orthoengine — ASTER DEM Module DEM and orthoimagery generation ArcView 3.2 and Arclnfo 9.1 with spatial Plotting and visualisation of GIS. Spatial analyst analysis of DEMs. Vector interpretation. Rivertools Hydrological spatial analysis of DEMs Excel and Access databases Collection and analysis of field data e.g. measured terrain profiles. Tabulation of GPS data together with attributes of field data e.g. structural measurements. Statistical analysis of terrain data. Generation of graphs for presentation Corel Draw Preparation of annotated figures.

The above table illustrates the potential problems in data formats when using the different packages. The formation of a GIS database centralises the processing results for interpretation and map composition. This chapter summarises the multi-discipline processing approach methodology with brief discussions on processing steps.

36 3.2 Digital Processing Methodology

The research methodology involves the acquisition, processing and interpretation of numerous data sources and derivative products (Fig. 3.1). The combined data was geo- referenced into consistent datum and projection in order to make spatial data layer comparisons. Some of the data was already in a "GIS" ready format while other data required processing such as satellite imagery, DEM generation and digitisation of field data. Once compiled in a GIS database analysis work was carried out. Initial work resulted in "value added" products such as slope maps and stream networks derived from DEMs, which lead to the identification of further work and analysis. This iterative process led to the production of a geomorphology map. Interpretation of the map enabled the identification of key tectonic geomorphic features. These were used to construct a conceptual tectonic geomorphology evolution model. Table 3.2 outlines the major processing steps which are described in more detail.

Table 3.2. Digital processing methodology descriptions. Data Acqusition Download available imagery and DEM. (orthophotos, 10m DEM, ASTER imagery). Digitise and georectify geologic and topographic maps.Import GPS point data and topographic profile location Pre-Processing DEM generation using ASTER stereo imagery Geo-Spatial Database Compilation of all data into a geospatial database in a consistent datum and projection Analysis Hydrological analysis of DEM giving stream networks, watershed boundaries, longitudinal stream profiles. Topographic profiling using DEM Results Geomorphology Map Interpretation Interpretation of geomorphology Conceptual Model Development of conceptual coupled tectonic geomorphological model.

37 Data Raw digital data Acquisition in Field data acquisition - requires "GIS ready" collection processing format

Data Digital input of processing field data into e.g. DEM database generation

GIS database compilation 0 41

Further work Value added products based on Analysis e.g. slope map analysis work derived from DEM

V Results "Geomorphology Map" Further work based on 41 interpretation

V Interpretation of Geomorphology Identification of key tectonic geomorphology indicators

Development of conceptual model of coupled tectonic / geomorphology evolution in study area.

Figure 3.1. Diagram outlining research methodology involving data preparation, analysis and interpretation.

38 3.3 Data Acquisition Data acquisition for both study areas involves the compilation of several different data types including raster, vector and hardcopy from a variety of sources. Most data, especially raster imagery is available via the web from ftp sties (see table 3.x). Raster data include satellite imagery, digital aerial photographs and gridded elevation (DEM). Vector products for this research are derived from the processed data and are described in the interpretation process below.

For the Wyoming case, raster and vector data was acquire via the GIS data clearinghouse website (web address). Geological information was obtained from the Geohazard section Wyoming Geological State Survey and fieldwork.

Table 3.3. Data types for Star Valley. Description Data Type / Resolution Source Orthophotograph Raster (.jpg). lm per pixel Wyoming GIS DEM Raster, 10m per pixel Wyoming GIS ASTER imagery Raster, Multispectral satellite USGS imagery. Geological Map Hardcopy. 1:50,000 scale Wyoming Geological Survey. Topographic Map Digital scans of hardcopy. Wyoming GIS Raster scans come complete with geo-reference file Fault Map Vector, Star Valley Faults Wyoming Geological Survey Fieldwork Location Vector points from GPS Fieldwork measurement

For the Turpan case data has been compiled from websites and from our Chinese collaborators at the Seismological Bureau together with fieldwork data collection.

39 Table 3.4. Date types for Turpan Basin. Description Data Type Source Landsat ETM+ Raster (.jpg). Multispectral USGS (cost approx. £400) imagery Corona stereo Raster, declassified military USGS (cost approx satellite photography, 5m including digital scanning per pixel. £100) ASTER imagery Raster, Multispectral USGS satellite imagery. Used to generate DEM Geological Map Hardcopy, 1:200,000 scale. Seismological Bureau Topographic Map Digital Raster Scan at Geopubs (£12.00 per map 300dpi, 1:250000 scale sheet scan) Russian. Fieldwork Location Vector points from GPS Fieldwork measurement

All digital and hardcopy data are in various datum and projections. The compilation of this data is briefly outlined below.

3.4 DEM generation The DEM processing has focussed on using new ASTER satellite stereo imagery for medium scale regional DEM generation. The following describes the ASTER instrument characteristics relevant for DEM generation and a summary of the process. The ASTER sensor is onboard the NASA TERRA satellite launched in December, 1999. Stereo imagery is collected by the Visible and Near Infrared (VNIR) subsystem (Abrams, 2000). A two telescope array is positioned so that one is pointed directly at the Earth's surface (Nadir) and the other tilted 27.6° to the nadir. ASTER stereo bands 3N (Nadir) and 3B (aft) have spectral bandwidths positioned in the NIR (0.76 — 0.86 1.1m). Each band image pixel has a nominal pixel resolution of 15m.

The DEM generation has been carried out using PCI Geomatica OrthoEngine software. The process involves several stages from input of the stereo imagery to the final gridding of elevation points to generate the DEM. PCI allows input of the hdf (ASTER data format) files directly into the OrthoEngine software. Information regarding real

40 coordinates i.e. the datum and projection are inputed. Matching of the stereo involves relative and absolute matching of the images. Relative matching involves a semi- automatic procedure where like points e.g. river junctions and road junctions, are measured in both images. These tie points must be distributed across the whole overlap area on both images. The absolute matching involves a manual process similar to the manual editing phase of tie pointing. Ground control points (GCP) are identified and measured in both images. These GCPs are terrain features that ideally do not change over time e.g. road junctions and building sides. A minimum eight GCPs are measured and distributed evenly across both images. The corresponding real world coordinates are measured from a topographic map. The computer is then able to triangulate the exterior orientation of the satellite using the image points and the measured xyz coordinates. This is known as the bundle adjustment and computes the photogrammetric model using the tie points, GCPs and the sensor ephemeris information.

The remainder of the processing is automatic. To efficiently calculate the elevation model the DEM process is separated into two processing steps. Epipolar generation involves calculation of the y offset of the stereo images. The epipolar images are then used to generate the DEM with image matching solving the x offset. The last phase is the geocoding of the relative DEM using the GCPs inputed in the earlier process. Errors in automatic DEM generation occur where matching of the images fails. Likely errors or blunders that produce spikes or pits in the DEM are the result of:

• Clouds • Mismatched areas due to tie point measurement errors • Poorly selected GCPs • Inaccurate manual measurement of GCPs • Topographic map accuracy used for GCP selection.

The geocoded DEM is then used to Orthorectify the ASTER image bands. Orthorectification removes central projection geometry distortions from the imagery resulting in the correction for terrain and elevation changes. Orthorectification in PCI is an automatic process with all three VNIR bands being corrected.

41 3.4.1 DEM Accuracy

The accuracy of DEMs can affect the results of derived products (Holmes et al., 2000; Kenward et al., 2000). The common measure of DEM accuracy is the root mean square error (RMSE), which gives a horizontal and vertical range of error.

The DEM used for Star Valley was obtained online from the Wyoming GIS data clearinghouse (see references for URL). The pixel resolution of the DEM is 10 m. The DEM is classed as the most accurate elevation model available from the clearinghouse and is equilivant to the high resolution USGS DEM products. The vertical accuracy of the DEM is +1- 7 m.

Two DEMs were used for the Turpan fold area. The ASTER 30 m pixel resolution DEM was generated using stereo imagery. The SRTM DEM has an absolute horizontal accuracy of 20 m and an absolute vertical accuracy of 16 m (see references for SRTM website URL).

The ASTER and SRTM DEMs provide terrain information at scales less than the 10 m DEM used in Star Valley. The larger pixel size has an impact on terrain measurements. The calculation of slope on the SRTM DEM lowers the slope angle effectively smoothing the values as the pixel size is large (Kenward et al., 2000)

3.5 Geo-spatial Database Compilation A range of data and imagery has been acquired, processed and manipulated. In order to compare information from different digital datasets all the data has been compiled into a consistent datum and projection in a GIS. This enables comparison of data types and allows the overlay of multidata for visualisation. ArcView GIS software package allows this integration and display. Datum and projection re-projection have been carried out using ERMapper image processing software. All digital data in Star Valley have been referenced to the orthophotography in NAD83 (datum) and UTM 12 (projection). In Turpan basin the data have been referenced to ASTER imagery reprojected using field control points in WGS84 (datum) and UTM 45N (projection). Hardcopy data in the form of maps have been scanned then geo-rectified using the projection grid or referenced to the imagery using ERMapper.

42 3.6 Spatial analysis

Spatial analysis of DEMs involves automatic user defined calculations based primarily on neighbourhood processing. Edge enhancements of the DEMs and imagery allow interpretation of general geomorphology delineating valley edges, ridges and dissected terrain. More involved DEM hydrological analysis processing simulates water flow over terrain resulting in the automatic characterisation of basins, watersheds and stream networks.

3.6.1 Hydrological Analyses Hydrological spatial analysis of DEMs involves a multi-stage process methodology (Jensen et al, 1988, and Tarboton et al, 1991). Rivertools software (Research Systems) and ArcInfo GIS 9.1 spatial analysis extension (ESRI) was used for hydrological analysis. They allow the automatic extraction of steam networks, watersheds / basins and stream profiles together with user defined parameters such as the location of specific basin outlets.

Assesssment of the DEM is carried out to establish DEM quality by determining artefacts such as sinks (holes in the DEM resulting in all flows converging on that pixel) or spikes (flows around spikes in the DEM, most likely caused by DEM generation "blunder" errors e.g. trees). The DEM is checked for errors and identified sinks are filled to the lowest neighbourhood pixel value.

Flow direction is a neighbourhood pixel surface calculation that determines the direction of flow from each DEM pixel (Jensen et al, 1988). The direction of flow is calculated by comparing the height value of a pixel to surrounding neighbourhood pixels within 3 x 3 kernel window. The flow direction is the vector linked from the centre pixel to the pixel with the lowest elevation value based on the following equation:

Maximum drop = change in elevation value / distance

If the maximum drop is the same for all surrounding pixels the neighbourhood is enlarged until the maximum drop is determined. There are eight possible directions relating to eight adjacent pixels (see direction coding in Fig 3.2). The spatial calculation output is a raster dataset of eight values (known as a D8 flow grid) of flow direction. 43 Flow Direction

78 72 69 71 58 49 2 2 2 4 74 67 56 49 46 50 2 2 2 4 4 69 53 44 37 38 48 1 2 4 64 58 55 22 31 24 128 128 1 2 4 68 61 47 21 16 19 2 2 1 4 4 74 53 34 12 11 12 1 1 4 DEM Flow Direction

3x3 direction grid

Flow Accumlation

10101101111111 0 0 0 0 0 0 0110011111 0 1 1 2 2 0 ==leill 11 0 3 7 5 4 0 1/8 MI=L111111 0 0 0 20 0 1 no=nrin 0 0 0 1 24 0 ====g11= 0 2 4 7 35 2 Flow Dir. - Flow Acc. Flow Accumulation Grid

Figure 3.2. Diagram showing calculation of flow direction across a DEM based on a 3 x 3 kernel window.

The flow accumulation surface calculation links these flow direction vectors forming an accumulated weight of all pixels that flow into each downslope cell. A flow direction raster is used to calculate the accumulated number of pixels that flow into each pixel (Fig. 3.2). Pixels with high flow accumulation are areas of focussed flow and can be interpreted as stream channels. Areas of zero local flows correspond to topographic ridges. A stream network is generated by thresholding the stream network. Strahler stream connectivity was used to generate the network using third order thresholding.

44

Flow direction DEM Sink calculation Fill sinks

Automatic generation of Depressionless basin DEM catchments

• V Flow Stream order Generation of accumulation Threshold applied (Strahler order basin wide swath calculation 3) using conditional function in profiles for spatial analyst to give stream overview maps network

Identification of pour points along range front. V Production of point file of Stream line watershed outlets Converted stream line to vector line using raster to vector conversion function

Generation of Generation of range range front swath front watersheds / profiles e.g. max. / catchments using min. slope. outlet file

V Generation of catchment properties e.g. area, flow length & average slope.

Interpretation of results. Spatial analysis products added to GIS database

Figure 3.3. Flow diagram showing DEM hydrological spatial analysis process methodology.

45 There are a number of user controls on the density of the generated network. Based on the flow accumulation, watershed or catchment basins can be automatically generated. In-order to analyse footwall basins each footwall basin outlet is manually measured and this point was used to calculate the upstream extents of that catchment area and water gaps. The flow diagram presented shows the spatial DEM hydrological process methodology (Fig. 3.3).

3.6.2 Slope / Aspect Calculations Slope and aspect neighbourhood pixel calculations are carried out on DEMs in Arclnfo using the spatial analysis tool. These calculations are based on a 3x3 kernel window that is passed over the DEM grid. Average slope is calculated from neighbouring cell values with the value replaced at the centre of the kernel. The aspect calculations like the flow direction is the neighbourhood elevation with the greatest drop in elevation compared to the central kernel value. This vector is then recorded corresponding to the facing compass direction.

3.6.3 Swath Profiles Swath profiles are graphs calculated from DEM regions. Maximum, mean and minimum profiles are the common profiles calculated from elevation values projected onto a common traverse. Regional swath profiles are calculated from rectangular cut regions extracted from ERMapper. These binary grids are imported in Arcview and exported as ascii grids, readable by excel where statistical measurements are made. Along row calculations are made for the entire grid. DEM grid extraction of footwall catchment regions and irregularly shaped sub-regions result in non-rectangular grids. In ERMapper a InRegion condition is applied where pixel values outside the required region are set to null. The extracted region is then rotated so that the traverse line for the swath is aligned N-S. For example, footwall catchments are rotated so that the trunk streams are aligned N-S. Again, these DEM grid region elevations and null values are exported to ArcView as grids and exported as ascii grids for import into Excel. In Excel along row cell calculations are made resulting in a column of summed data. The common along row calculations are maximum, average and minimum. The calculated column of data is graphed along DEM traverse cumulative pixel z (height) distance.

46 Slope angles have been calculated between neighbouring pixels for extracted profiles within each footwall / upstream catchment. For each DEM pixel along the stream channel the difference in height and slope angle change is calculated.

3.6.4 Concavity Concavity is the degree of curvature of a terrain surface or longitudinal stream profile. A simple calculation can be applied to elevation along a stream profile. A linear regression line is fitted to catchment relief (difference in elevation of the catchment outlet and upstream catchment elevation). The following equation is used to measure the difference in observed elevation relative to the linear regression line.

Simple concavity = observed elevation — linear regression elevation

For every elevation point measurement along a given traverse line, in this case the catchment trunk stream the observed elevation is compared to the linear regression line. Positive values indicate convex profiles while negative values indicate concave profile shapes (Fig. 3.4). Canyon max. Max Convex stream topography \ relief profile

0 Canyon outlet a) w Concave stream ••• profile

Linear regression line

Horizontal catchment profile

Figure 3.4. Schematic diagram showing the relationship of longitudinal stream profile to a linear regression line fitted to catchment relief.

3.7 Image Processing Digital image processing involved producing image base layers for the GIS. The generation of standard false colour composites using Landsat & and ASTER imagery show general geological and geomorphological features. There is now a need to develop 47 a methodology for analysing alluvial fans regions in the Turpan case, using new ASTER imagery. Directed enhancement using VNIR, SWIR and TIR can provide information on rock-type and can be used to discriminate alluvial / bajada fans regions.

Perspective or pseudo 3D visualisation has involved the generation of perspective "virtual views" within ERMapper and ERDAS Imagine image processing software. Remotely sensed image or topographic maps are draped over elevation models within the software to form perspective views. This interactive process allowed visualisation of the terrain that is not possible from fieldwork alone.

3.8 Field Mapping 3.8.1 Terrain Transects Two methods have been used to measure terrain profiles in identified field area locations. In Wyoming the DGPS was favoured in areas where there was direct view of the sky and the GPS network. This allowed accurate and quick measuring of the profile. However, the majority of the identified terrain profiles were covered with extensive tree canopies. The DGPS technique was not possible in these areas as there was not a clear view of the sky. The laser ranging levelling method was used in forested areas. The laser ranging levelling method was used for the Turpan fieldwork as permission for use of DGPS was not given as this part of Xinjiang is in a sensitive area.

3.8.2 Levelling Method Terrain traverse height profiles have been measured using a levelling surveying method using an electronic distance measurement (EDM) laser which records the angle / distance between two points relative to one point. The EDM used is a Laser Technology laser ranging device, which measures the distance with a laser and the angle of transmittance with an internal inclinometer. The EDM calculates the horizontal distance and vertical height based on these two ranged measurements. The device used has a distance resolution of 0.01m and +/- 0.1° inclination.

The surveying method involved recording height and distance differences between two surveying poles. An EDM mounted pole of fixed height and a target mark positioned at the same height on the target pole. The EDM is pointed towards the target pole reflective mark and the EDM laser is activated. Once the distance and relative height

48 measurements have been recorded the EDM mounted pole is moved to the target pole and the target pole moved to a new position along the traverse.

Bicycle reflector Laser Ranger Inclination angle \ He

------fit Horizontal distance

Figure 3.5. Diagram showing the levelling surveying method and calculation of height along a survey traverse using a laser ranging instrument.

This method is sensitive to the propagation of errors through the traverse. However, over short distances along the traverse the relative change the error is acceptable for 1:10000 surveying.

3.8.3 Differential GPS The DGPS involves the setup of a GPS base station which monitors and calculates the absolute position error or drift, over time, for its position triangulated from measurements received from the global network of spaceborne GPS satellites. A rover GPS communicates with the base station during the survey, which measures the relative distance to the base station. This is achieved by the base station transmitting the position correction. The handheld computer calculates the accuracy of the GPS measurement based on the real-time comparison with the base station. The horizontal and vertical measurements recorded were +1- 6cm.

49 3.8.4 GPS for fieldwork Hand held GPS units now have a purported horizontal accuracy of +1- 5m and vertical accuracy of +1- 10m (NIMA website) since selective availability was turned off by the American Military. Accuracies calculated by the handheld sets claim accuracies of 5m horizontal with no measure of vertical. When possible the GPS unit (Garmin Vista) was checked against benchmarks in the field which confirmed that the horizontal was within 5m and the vertical was within 15m. These accuracies are adequate for 1:10000 scale mapping.

3.8.5 Digital Mapping Digital mapping is now an essential tool for recording field observations is an essential and cost effective method for producing and interpreting maps and information in the field. Digital mapping allowed regular assessment and improvement of the field mapping plan by using targeted fieldwork identified from the imagery. Digital maps allowed an assessment of the geological mapping improving tracing geological boundaries.

The methodology used during the fieldwork was an intermediate stage to digital mapping in the field. There was no access to palm tops which would have allowed the recording of data digitally so observations were made using a traditional notebook. However, GPS points and digital photographs were recorded in the field. All data was collated in the evening and updated into the GIS allowing reassessment of the fieldwork plan on a daily basis.

3.9 Interpretation Interpretation involves interactive mapping of processed imagery as this is the most efficient and objective view of the spatial drainage patterns and morphology. Vector files are generated on screen where point, line or polygons are drawn and integrated into the database. These vector files can be overlain on the base images. Summaries of these data processing and mapping results will be described in Chapters 3 and 4. Analysis and interpretation will be made in Chapter 5.

50 Chapter 4: Tectonic Geomorphology of Star Valley, Wyoming, USA: Implications for Normal Fault Segment Growth at Extensional Relays.

4.1 Research Rationale This chapter describes the mapping and interpretation of the tectonic geomorphology of a continental extensional normal fault relay, which is used to test the hypothesis of catchment capture based on numerical model simulations of terrain development discussed in Chapter 2 (Densmore et al., 2003). The Grover relay, the region of en echelon overlap between the Star Valley normal fault segments, near Afton, Wyoming is the focus for this study. The faulted terrain is in the tectonically active Eastern Basin and Range Province where excellent preserved tectonic geomorphic features are accessible. To test the model of catchment capture and the implications for fault development at extensional relays this research focuses on the following aims: • To reconstruct the tectonic geomorphological evolution of the Grover relay, the zone of overlap between two normal fault segments, in Star Valley, Wyoming. • To develop a model of Fault Propagation in Star Valley. To address the above aims an inter-disciplinary research approach is applied that integrates remote sensing and fieldwork observations. The following outlines the main objectives that allow the development of a fault propagation model in Star Valley: • Geomorphology mapping of drainage and footwall catchments using a 10m DEM and orthophotography. • To map syn-extensional fault related sediments in the relay and at the tip of the southern fault segment to establish palaeo-geography during fault segment development and fault tip propagation. • To map structure within the relay to establish the degree of breaching in the relay. The chapter is divided into two main sections; 4.1 and 4.2 outlines previous work, regional geology and geomorphology described from previous research; Section 4.3 onwards describes the tectonic geomorphology mapping of the Star Valley fault terrain using remotely sensed imagery, digital elevation model and fieldwork. Interpretation of these results has allowed a reconstruction of coupled geomorphology and tectonic evolution. This has led to the development of a relative fault propagation model based on presented research results. 51 4.1.1 Location The Basin and Range province, western interior of the United States, is a region of asymmetric topography consisting of elevated terrain (range) and intermontane sedimentary basins (basin). This landscape morphology development is a result of Late Mesozoic compression related to the Laramide / Sevier Orogeny (Piety et al., 1992). Later Cenozoic extension of these Sevier foreland basins, involving normal faulting with associated footwall uplift and hanging wall basin evolution, formed this basin and range terrain. The preservation of geomorphology, continued fault activity and easy access makes this region an ideal area to study and measure the spatial and temporal relationships between erosional processes and tectonic deformation. The Grand Valley Fault in Wyoming is one of several Cenozoic normal faults in the Eastern Basin and Range Province. It is an arcuate, segmented fault that shows continuing tectonic activity. Figure 4.1 shows the location of this fault associated with a series of approximately N-S orientated mountain ranges, in relation to the Snake River Plain and Yellowstone volcanic province situated to the north. The Star Valley fault is the southernmost segment and palaeoseismic studies indicate latest Quaternary activity (Piety et al., 1992; Warren et al., 1992). This fault segment which itself is separated into two distinct segments with a right stepping en-echelon overlap is the focus of this study. The following heading summarises the previous research in the Star Valley region.

4.1.2 Previous Work Previous work carried out on the Grand Valley and Star Valley fault has concentrated on the palaeoseismicity or recurrence interval of faulting within the basin. Piety et al. (1986) carried out the first detailed and systematic mapping of fault scarps and Quaternary sediments. Their seismotectonic study was used to assess the local geo- hazard to the Palisades and reservoir, to the north west of Star Valley near Alpine, Wyoming. The results of this assessment indicated that the Star Valley fault is potentially active showing latest Quaternary deformation. Attention was directed to the geomorphology and sedimentology of Star Valley to assess the latest fault movement. An unpublished master thesis (Warren, 1992) focused on the detailed mapping of sediments and geomorphology of Star Valley with an emphasis on determining palaeoseismicity.

52 Figure 4.1. Sun-shaded regional topography image showing location of the Grand Valley Fault and Star Valley study area (Piety et al., 1992).

53 4.2 Regional Background The development of the basin and range terrain in western Wyoming and neighbouring states has resulted from the continuing interplay between tectonic events and erosional processes, including long-term climatic events, since the Late Jurassic to Early Cretaceous. The following summaries of these mountain building episodes indicate past activity, geomorphic development and reactivation of tectonic fabrics that have contributed to the present basin and range style landscape. The Wyoming and Idaho Overthrust Belt (Fig. 4.2a), made up of thrust sheets is one of a number of structural segments that form a sinuous trend extending from Mexico to Alaska. These structures formed in response to compression during the Sevier orogeny (150 to 55 Ma). The thrust belt propagated eastwards in a foreland basin setting and developed in five phases from the late Mesozoic to early Cenozoic (Lageson et al., 2000; Chester, 2003). A cross section (Fig. 4.2b) based on an approximately west-east seismic line across the northern part of Idaho and Wyoming shows the geometry of the Absaroka thrust at depth and indicates that the thrust is related to the topography in the Salt River range (Piety et al., 1992). The Star Valley fault is thought to be a re-activation of the Absaroka N-S trending thrust. The northeastern Basin and Range, centred on the eastern Snake River plain, is a region of extension containing numerous Cenozoic normal faults. The faults show evidence of ongoing late Quaternary (approximately 125ka) displacements and several of these display recent Quaternary (<15 ka) movement (Piety et al., 1992). The plan view fault map (Fig. 4.3) shows the north to northwest orientation where solid red lines indicate the position of recent Quaternary displacements. The faults display a variety of fault segment interactions and linkage with several faults having an en echelon right stepped geometry. The Grand Valley fault is one of these en echelon faults and shows differences in fault geomorphology south along range strike. The Grand Valley fault is approximately 140 km long and extends from the eastern Snake River Plain in Idaho into western Wyoming bounded to the west by the Snake River range and to the east by the Salt River Range. The fault is arcuate and is separated, based on palaeoseismicity into three fault segments; Snake River segment, Grand Valley segment and the Star Valley segment (Fig. 4.3b). Anders et al. (1989) indicate that extension on the Grand Valley fault began 10 million years ago. Activity on the Star Valley segment is thought to have started 2 million years after initiation of faulting in Grand and Swan valleys.

54 Sketch of Star Valley Half Graben

Tertiary Basin

Basement

Star Valley Normal Fault

b SW Salt River'Range E km Star Valley Fault ATF

10-

ATF : Absaroka thrust fault ' 0 km 10 (after Dixon, 1982)

Late Cenozoic seds. II1Jurassic/Triassic seds. Precamb. seds.

Cretaceous seds. Palaeozoic seds EPrecamb. basement

Figure 4.2. a) Overthrust Belt structure map showing location of the Absaroka Thrust, b) west — east cross section based on regional seismic line (after Piety et al., 1986) showing relationship between Absaroka Thrust and the Star Valley fault.

55

Fi ( and modif gure 4

R a) b) Snake Rive' TetonTeton an i ed 114° \ 113° , 112° 1 111°./ Plain .3 ge P 4 ,Basin , Range Jackson

af

. fr

a) C Siy Hole t er Pi rovi -45° —s— . 45°- en ej, et n ozoi ce y , b et c f al ) aul . re , 1992 gi t ma onal ) p .

of st r

th uct e S ural nak

geol e Ri

o Cenozoic Fault Map ver Pl gy -43° 43° m ai a p n L Eastern Snake River of re Plain gi I

th qs on i

e G Late Cenozoic fault with 4 3b / /)„ \ :.4 n suspected Quaternary

rand -1) Idaho% ing

th 1 , I . displacement

e E km 20 y,

42°

V Late Cenozoic faultg •th O\• k \ ast

all Quaternary displae nts Cenozoic sedimentary ern B

e (<= 15ka). and volcanic rocks \Normal Fault y F 113° 2° Palaeozoic and Mesozoic asi a basement ul III n t 4.2.1 Star Valley Geological Setting The location of the Star Valley fault system is shown in Figure 4.4. The westward dipping Star Valley fault, the southern-most segment of the Grand Valley fault, forms a N-S linear range front on the western side of the Salt River Mountains ranging in geographic extent from Longitude 110:46:11W to 111:4:33W (grid ref: E493823 to E518882) and Latitude 42:36:49N to 43:0:57N (grid ref: N4717936 to N4762585). The half graben is divided into two sub-basins, separated by elevated terrain east of the Narrows, north of the village of Grover, at grid ref E500911 N4744409. The plan view geology of Star Valley shows the fault geometry and stratigraphy of this basin and range structure (Fig. 4.4).

The Star Valley fault consists of two fault segments called the Northern Star Valley fault segment and the Southern Star Valley fault segment (Piety et al., 1992 and Warren, 1992). For this study, the two segments are defined as the rear and front fault segments respectively (see Chapter 2). These fault segments overlap to the east of Grover village forming the Grover relay. The footwall stratigraphy consists of Palaeozoic through to Mesozoic rocks, which are grouped together as basement (Fig. 4.4). These rocks are deformed and folded by the Absaroka thrust, part of a series of Late Mesozoic foreland thrust sheets (Piety et al., 1992; Chester et al., 2003). The hangingwall stratigraphy consists of Tertiary and Quaternary sedimentary rocks. The outcrop age is different in the two sub-basins, north and south of Star Valley. In the north, the oldest Tertiary rocks exposed consist of Upper Miocene to Lower Pleistocene silici-clastic sandstones and conglomerates. These rocks are not exposed in southern Star Valley. Upper Pleistocene to Holocene alluvium is present in both sub-basins.

The cross sectional perspective view of Star Valley shows the half graben geometry and the unconformable relationships of the Tertiary and Quaternary sediments overlain on the earlier deformed Palaeozoic and Mesozoic rocks (Fig. 4.5).

57 Alluvium (upper Corral Canyon Pleistocene & Holocene? Call Creek Loess (Upper Pleistocene) e Creek Basin Fill: Upper Prater Canyon Miocene to lower Pleistocene se* Palaeozoic and ceo26 Mesozoic rocks wee* os` Fault

fo Gtee''

Narrows ti Creek

Wi ow Creek

Relay zone

Front Segment

Rear segment

Anderson Canyon 73. (S, Swift Creek

Dry Creek I

Cottonwood Creek

0 5km

Figure 4.4. Summary geology map of Star Valley (modified after Piety et al., 1992).

58 3km

2km

t km

0

3km Absaroka Thrust 0 Approx. 2km Star Valley Fault Tertiary Fill Southern Segment

Figure 4.5. Perspective view of Star Valley with side panel showing west — east orientated cross section (modified after Ruby, 1957) latitude 42°47'30", showing half graben geometry.

4.2.2 Fault Segment Geometry The Star Valley fault trace has a N-S orientation with an overall cumulative length of 61 km measured from the DEM, which includes the northern section of the North Star Valley Fault (Fig. 4.6). The extent and position of the fault segments are defined by Piety et al. (1992) and verified by aerial photograph mapping and field observations. The rear fault segment in this study is defined as the section south of Prater Canyon. The rear segment extends from Prater Canyon in the north (E503686, N4759933), to four kilometres south of Willow Creek (E509789 N4739035). The range front between the southern tip of the Grand Valley segment and Prater Canyon is the northern section of the Star Valley segment, which has a less defined range front. Piety et al. (1986) names this section as the older North Star Valley Fault showing no evidence of Quaternary scarps. The front segment extends from Willow Creek in the north (E505440 N4743922) to Cottonwood Canyon in the south (E507350 N4719313).

59 Relay Fault

Grover Relay

KEY

Latest Quaternary iv Fault Scarp v/ Degraded Fault

Al Relay Fault

Vertical Displacement 9 Latest Quat. Fault Scarps

1,1

0 2 4

500000 510000 520000

Figure 4.6. Star Valley Fault map overlain on sun-shaded topography. Red tones indicate elevations over 2500m and blue tones around 1000m (modified after Piety et al., 1992).

60 The rear segment trace changes slightly along strike and, although generally continuous, there are a number of small steps along the range front, with a larger eastern step south of Willow Creek. The northern part of the segment between Prater Canyon and Dry Canyon has an orientation of 330° and length of 15 km. There is a distinct change in orientation between Dry Canyon and Willow Creek to 020°, and is 3.32 km in length. The fault trace reverts back to a N-S orientation and has an east step with a separation of 1 km. The trace extends south across the upper regions of the Phillips Creek catchment with a length of 4.14 km. The cumulative length of the rear segment from Prater Canyon to the upper catchment of Phillip Creek is 22.61 km. The front fault has an average N-S orientated plan view trace between Willow Creek and north of Cottonwood Creek near the village of Smoot, 3km north of Cottonwood Creek. The plan view trace from Jensen Canyon to Bradshaw Canyon changes from a N-S trend to an orientation of 340°. The overall cumulative length of the front segment is 25.26 km. Well-formed triangular facets dominate the morphology of the front segment range. The front and rear segments are overlapped in the middle of Star Valley centred at E508638 N4741470. The Grover relay, the zone of fault segment overlap, has a right stepping en echelon geometry with an overlap of 5 km and a separation of 4 km.

4.2.3 Latest Quaternary Displacements Identification and mapping of scarps in Quaternary fan deposits of late Pleistocene to early Holocene age, at range front outlets, south from Willow Creek to Cottonwood Creek near Smoot (Fig 4.6), indicate latest fault activity in the southern Star Valley (Warren, 1992; Piety et al., 1992). Table 4.1 (after Warren, 1992) summaries the morphology of latest fault scarps along the front segment and the lower portion of the north fault segment. The measurements in each table cell correspond to displaced Pleistocene and Holocene fan scarp measurements, respectively. Fieldwork confirms that fault scarps are not identified along the southern segment 3 km north of Bradshaw Canyon to Willow Creek. Figure 4.6 shows the location of the measured fault scarps and fieldwork confirms that the latest fault scarps disappear north of Bradshaw Canyon. The identification of latest fault scarps along the southern rear segment and the majority of the front segment indicates soft linkage of the two segments (Warren, 1992).

61 Table 4.1. Fault scarp characteristics along the southern Star Valley fault segment (after Warren 1992). Profile Vertical Surface Scarp height Maximum Slope of Slope of Location displacement(m) (m) slope angle upper surface lower surface (deg) (deg) (deg) Willow Creek 8.3 9.34 31.5 1.5-3.5 3-3.5 3.3 16.5 3-3.5 2.5-3 Phillips Creek 9 10 25 2-5 0-2 Swift Creek 11 12 34 1-3 2-4 7 11.3 25 9-12 2-8 Dry Creek 11.3 12.3 27 0-1.5 0-1 5 5.3 28 0 <2

4.2.4 Stratigraphy The oldest Tertiary sedimentary deposits outcropping in Star Valley are the basin fill conglomerates located in the northern sub-basin (Fig. 4.5). The Miocene to Pliocene Salt Lake Formation is mapped in the hangingwall of the Grand Valley fault segments. The outcrops further north of Star Valley near Alpine are subdivided into Upper Miocene to Lower Pliocene Salt Lake Formation and Lower Pleistocene and Upper Pliocene Long Spring Formation. The deposits in southern Star Valley at Swift Creek and Phillips Creek consist of subangular, subrounded and angular limestone clasts reaching cobble size and quartzite / sandstone pebble sized clasts (Piety et al., 1986). The conglomerate clasts are cemented with a pink coloured carbonate where limestone clasts are common but poorly consolidated where quartzite clasts are common. The low angle bedding dips and conglomeratic facies suggests that the mapped Salt Lake Formation in Star Valley maybe the Long Spring Formation which outcrops further north in the Grand Valley (Piety et al., 1986). The Long Spring Formation here is younger than the Salt Lake Conglomerates based on ash layer dates.

4.2.5 Geomorphology Star Valley is separated into two equally sized basins bounded to the east by the Salt River Range and to the west by the Caribou Range. The Salt River flows from south to north along the western side of the basin. The north and south basins are separated by the "Narrows", NW of Grover, where the Salt River flows along the western front of the Caribou Range (Fig. 4.7). The geomorphology of Star Valley is markedly different on the east side of the valley, which is bounded by the right-stepping segments of the north and south Star Valley faults. The relief ranges from 1830m to 3283m in the Salt River

62 Range. The most striking geomorphic feature of southern Star Valley is the linear faceted escarpment separating the valley fill and Mesozoic/Palaeozoic rocks of the Salt River Range. Steep , often V-shaped, diamond-shaped range-front facets and numerous scarps indicate active faulting. There is a change in geomorphology from high elevation glacial features formed by past Pinedale glaciation (Warren, 1992) to fluvial erosion at lower elevations along the Salt River.

Figure 4.7. Perspective terrain view of Star Valley, orientated north-east, showing range front, basin and mountain morphology.

63 4.3 Tectonic Geomorphology Investigation of Star Valley The tectonic geomorphology of Star Valley is now described using a combination of digital elevation models (DEM), remotely sensed imagery and field observations. The structural geology and sedimentology of the Grover relay is further described in this section using digital imagery and field observations. The geomorphology of Star Valley is described in section 4.4 using DEM analysis. The Grover relay geomorphology is then further described in the following sections.

4.3.1 Structural Geology The fault map of the Grover relay shows the orientation of the front and rear faults (Fig. 4.8a). The intrabasin high to the north of the front fault is interpreted as a fault tip propagation fold where the northern extension of the fault has deformed syn-rift conglomerates. In the relay a new fault has been identified and mapped in the field. This relay fault cuts across the relay topography with a SW-NE orientation. The morphology of the fault is a series of fault scarps that displace drainage. The scarp heights decrease to the SW from 56m to < lm.

The perspective view of the Grover relay (Fig. 4.8b) shows the relationship of the geology to elevation. The conglomerate unconformity is clearly defined on the east side of the back-rotated limb of the front fault.

Rear Fault Segment The rear fault southern tip from Strawberry Creek south to upper parts of Phillips Creek is in the study area. The plan view geometry as indicated above changes along strike. The section of the fault from Strawberry Creek to the outlet of Dry Creek has an orientation of 330°. There is a distinct change in fault plan view orientation to 020° at Dry Creek, which continues 1.15 km SSW of Willow Creek outlet near Turnerville village. The plan view fault trace then changes back to a N-S orientation with a step of 1.2 km to the east. This fault tip has a degraded range front, when compared to the front segment, with triangular facets disappearing in the upper quarter of the NE side of Phillips Creek catchment. A photomontage of the range front (Fig. 4.9a) shows the range front morphology from Strawberry Creek into the Grover relay. The range front triangular facets appear more degraded when compared to the well-formed linear faceted range front of the front fault.

64 a) 510000 520000 Rear Segment I

O O

h- O Fault Tip 0 Fold

Grover Relay Fault

O0 O

V O Front Segment • N Grover Village z Normal fault with / latest Holocene A scarps / Normal fault 0 km 4 / segment 7 510000 520000

b) Phillips Creek

Figure 4.8. a) Plan view fault map of the Grover study area with fault segment overlain on a sun-shaded DEM image, b) perspective view of the geology draped over the DEM showing the relationship of the rock types to elevation. 65

4.3.1.2 Front Fault Segment The northern 13 km of the front fault tip from Swift Creek to Willow Creek is in the relay study area. The plan view trace shown in Figure 4.10 of the front segment is generally N-S with a slight plan view convex curvature to the east centred on Swift Creek. There is a change in orientation to 330° between Jensen Canyon and Bradshaw Canyon. A secondary fault located 300m east of the range front mapped by Ruby (1957) extends for 2 km south to E505678 N4740124, 650m north of Bradshaw Canyon (Fig. 4.8c & Fig. 4.11).

Apart from the degraded tip north of Bradshaw Canyon, well-formed triangular facets dominate the morphology of the range front. A photomontage (Fig. 4.9) of the range front at the tip near Willow Creek south to Phillips Creek shows the morphology of the triangular facets with key footwall canyons marked. The 10m DEM and fieldwork will be used to further describe the front tip transverse and along strike topography.

4.3.1.3 Grover Relay Structure A previously unmapped fault has been identified in the relay zone, using the DEM and field observations. In plan view the relay breaching fault trace orientation is approximately 040° (SW-NE) with a slight bowed convex curvature to the SE (Fig. 4.8a). The plan view cumulative length is 1.5 km mapped between grid references E508420 N4740746 and E509187 N4741724. The southwestern extents are not traceable within the heavily vegetated back-rotated topography of the front segment. The fault scarps also die out to the northeast and are not traceable in the densely forested hangingwall of the rear segment.

66 Phillips Creek Astle Creek Willow Creek Grover Relay Willow Creek

E Canyon Dutson Facing Canyon Bradshaw StraWCreek Facing StraWCreek Willow Creek Front Segment Tip

Fault tip fold b) a)

Figure 4.9.a) Annotated photomontageof therear segmentori entated N-S facing E, b) annotated photomontageof the front fault. 4.3.1.4 Front Segment Tip The front segment tip extends northwards from Willow Creek (GR E504251 N4743316) deforming Long Spring Formation conglomerates forming a fault tip fold. The fault tip fold forms a broad triangular shape with a series of elevated and dissected terrain decreasing in size and elevation from 300m to Om (Fig. 4.17) to the NW over a length of 10.8 km (Fig. 4.10a). This topography is bisected by a series of antecedent streams i.e. Willow Creek, Strawberry Creek and Cedar Creek. The perspective view of the geology draped over the DEM clearly shows the decrease in topography to the NW (Fig. 4.8b). The fold tip structure is defined by the Long Spring Formation conglomerates outcrop dipping 20° to the east and 10° to the west. The trace of the structure has been determined by mapping the Long Spring Formation conglomerates along Strawberry Creek and the surrounding terrain (Fig. 4.10a). Along the creek and road-cuts there are excellent exposures but there is little exposure in the hilly terrain between the creeks. A cross section has been constructed by projecting bedding readings onto west-east maximum and minimum topographic profiles. A structural formline cross section orientated along Strawberry Creek shows an asymmetric structure (Fig. 4.10b). This has been constructed from geological mapping of the fold terrain to the north and along a detailed W-E section of Strawberry Creek. The eastern limb is well delineated but cannot be traced to the west. Poles to bedding have been plotted on a Wulf stereonet. Stereonet analysis of poles to bedding indicates two clusters of poles to bedding indicating a fold axis plunging 10° to 345°. This has the same orientation to the strike of the fold topography.

68 \ Fault tip structure trace

L • Thattne

6

X2 F \ \'10 \ 15 Cross section line -18 N B 14

IC 30 Strawberry Creek 10 '0 15- T8 7 )1?•• \ 12 12 4 a g 12 \ O Salt Lake Formation \8 \22 40 \ Conglomerates 14 30

3

Willow Creek 0 km 2

500000 504000

Figure 4.10. a) Geological map of fault tip propagation fold.

69

b) Strawberry Creek Cross Section

A

Eastward dipping ridge C0 ; 2000

• To • 41••.• " t 1800 ? Formlines Star Valley Fault (front segment) below surface Salt Lake E Formation Bedding 2 1800 0 200 600 1000 1400 1800 2200 2600 3000 3400 3800 4200

Horizontal Distance (m)

C) structure plunging 10° to 345°

Great circle of pole to bedding

Figure 4.10. b) formline cross section of Strawberry Creek and the area to the north showing bedding asymmetry and stereonet of poles to bedding indicating fold axial plunge.

4.3.2 Relay Zone Stratigraphy

The plan view hangingwall and footwall stratigraphy in the Grover relay is dominated by the unconformity of the Long Spring Formation silici-clastic syn-rift conglomerates deposited on footwall basement rocks consisting of deformed Triassic and Jurassic limestones and sandstones (Fig. 4.11).

70

Geol_m apshp Ca ME Cbm I ICw I Jn L n't P▪PP Ppr Qa Qal 1.1 Qd mg, Qm ▪ Otg TSL Tra ▪ Trd ▪ Trtb ME Trty Trw

0 2 Kilometres

Figure 4.11. Geological map of the Grover Relay (modified after Ruby, 1957).

71 4.2.3 Front Segment Stratigraphy The conglomerates in Phillips Creek have been mapped by Piety et al., (1986; 1992). Field mapping was carried out to further describe these Long Spring Formation conglomerates in the relay and along the front segment footwall. The well-exposed outcrops in Phillips Creek allowed mapping of the outcrop exposures and the stratigraphic relationships of the conglomerates to the front segment basement rocks. The plan view outcrop pattern of the conglomerates in Phillips Creek cuts across the catchment forming eastward verging "v" shaped geometry. Continuous conglomerate ridges were traceable across the north slopes. The conglomerates dip approximately 5 degrees to the NE.

Two types of conglomerate have been identified in Phillips Creek. A monomict conglomerate is mainly found juxtaposed to the basement limestone rocks. The conglomerate consists mainly of sub-angular to angular limestone clasts. The polymict conglomerate is located further up in the sequence and east along the catchment. This conglomerate consists of a range of clast shapes ranging from sub-rounded to angular, with a range of shapes from spherical to elongate. The clasts consist of red sandstone, limestone and quartzite and range in size from centimetre to thirty centimetre sized clasts. Occasional boulder (metre) sized clasts appear to be in situ. The conglomerate is both matrix and clast supported. The clast-supported units have associated erosional bases with fining up sequences ranging over sixty centimetres. The upper parts of these sequences are matrix supported. Fig. 4.13 is an example of the polymict clast supported conglomerate.

72 41 ,111.111

110(1141.) t• 0009Et

Figure 4.12. Plan view geological map of fault footwall basement and Long Spring Formation conglomerates centred on Phillips Creek.

73 Figure 4.13. Photograph of Long Spring Formation polymict conglomerate in Phillips Creek.

The conglomerate outcrop overlain on the terrain model shows the outcrop pattern cutting across Phillips Creek catchment and the trace of the conglomerate unconformity northwards along the front fault towards Willow Creek (Fig. 3.14a). The conglomerates have been rotated on the back of the front footwall. A longitudinal catchment cross section along Phillips Creek has been constructed to show the stratigraphic relationship of the conglomerates on the footwall (Fig. 3.14b). An unconformity is traceable along the northern side where the conglomerates disappear on the back of the footwall next to the red Triassic sandstones. The polymict conglomerates form ridges that are traceable across the northern catchment slope dipping to the NE. The cross section shows the termination of the conglomerates against the footwall basement rocks.

Further mapping at GR E507517 N4738110 along the north side of Phillips Creek shows the stratigraphic relationships of the monomict and polymict conglomerates to the basement rocks (Fig. 4.15). The section shows that the monomict conglomerate is

74 located next to the basement Jurassic limestones. The polymict conglomerate cuts across both the monomict conglomerate and the basement rocks.

The exposure of the conglomerate outcrop on the north side of the canyon has shown the onlapping unconformable stratigraphic relationships of the conglomerates to the footwall Triassic to Jurassic limestones and sandstones (Fig. 4.15). The section indicates the change from monomict through to a polymict conglomerate which is interpreted as a change in sediment source. The monomict conglomerate is sourced locally with eroded commonly angular clasts of the Jurassic limestones by hillslope process resulting in infilling of local topography. The polymict conglomerate represents an onlapping of conglomerates onto the footwall rocks sourced from the relay / rear fault. These are interpreted as deposits from landsliding and erosion from the rear fault catchments. There is evidence of recent avalanche deposits within the modern day terrain with large boulder size clast being deposited in the base of some of the rear and front segment catchments. This results from toppling of eroded bedrock from the sides of the canyon or transportation of rocks from higher up the catchment by avalanche "chutes".

75 Phillips Creek

a)

Phillips Creek Cross Section

Key Latest Quaternary

Salt Lake / Long Spring Formation Maximum Topography w E L Twin Creek limestone 1 Jurassic _ _ _ _ - - - -- Minimum Topography [ i Nugget sandstone

IMIAnkareh redbeds — — Line of section IM Thaynes limestone Triassic ion t 2700 - r-7 Woodeside redbeds I. Dinwoody formation Southern Star Valley Fault era = Phosphoria formations ] Permian I/ WI Wells formation Exagg ] Carboniferous l 2450- IIM Amsden formation iffl Braser and Madison limestones ica

t Unconformity Ver

2x 2100 -

). (m ion t 1800

va 300 900 1500, 2100 2700 3300 3900 4500 5100 5700 6300 6900 Ele b) Horizontal Distance (m)

Figure 4.14. a) perspective view of the geological map overlain on digital terrain model, orientated facing north, b) Cross section of Phillips Creek.

76 Figure 4.15. Photomontage of the north side of Phillips Creek showing the unconformity and stratigraphic relationships between the Long Spring Formation and the footwall basement.

77 4.4 Star Valley Geomorphology Terrain classification of Star Valley involved recognition of geomorphic features using a combination of automatic spatial DEM processing, image processing and perspective visualisation. The tools of investigation including digital imagery and field surveying techniques are described in Chapter 3.

The range front on both the rear and front fault segments can be divided up into a number of zones based on geomorphology and structural morphology, notably footwall catchments and hangingwall fans. Initial classification of the DEM identified areas for more detailed study in the Grover relay study area.

4.4.1 Footwall Topography Transverse range topographic profiles generated using the DEM (Fig. 4.16) show the footwall and hangingwall profiles along the fault segments. The profiles have been orientated normal to range strike and located in the middle of the triangular facets. The selection of four profiles in the basin shows the range transverse relief change of the rear and front faults. The scarps shown in Figure 4.16 are the eroded surface of the fault plane and therefore its slope is less than the fault plane dip. Profile 1, located near to Prater Canyon, clearly shows the difference in the footwall relief and hangingwall base level of 600 m. Profile 4 is similar but with a greater footwall relief. Profile 10 shows the west-east profile orientated across the Grover relay. The footwall relief of the front segment is shown indicating an eroded back-rotated footwall profile. Profile 18 shows the footwall relief and hangingwall baselevel further south towards the tip of the front segment. These transverse topographic profiles show the fault cross-sectional topography and indicate the topographic change along fault strike, notably in the Grover relay.

A swath profile orientated N-S parallel to the range front, generated using the DEM, shows the change in hangingwall and faulted topography from Alpine in the north to Smoot towards the south of Star Valley (Fig. 4.17). These profiles show the relationship of the fault segments and topography in Star Valley projected onto a north south profile. This profile shows the change in hangingwall profile along strike with a localised increase in elevation at the Grover relay. The maximum topography is generally constant along the range with a decrease at the segment tips.

78 14 Location of Transverse Profiles 10

18i_1,_

1. North Segment 3000 E 2800 W Front segment Lg 2600 -....- 2400 - Footwall .2 2200 Hangingwall basin 2000 - _DEM Profile 1800 ' 4 . 0 2 4 6 8 10 Line Distance (km)

4. North Segment 3100 2900 IN E Rear Segment i 2700 5 2500 Footwall 23°° Hangingwall basin w 2100 1900 — DEM Profile 1700 0 5 10 15 20 Line Distance (km)

10. Grover Relay 3300 3100 1W Rear Segment ? 2900 Grover Relay 2700 1 25„ Front Segment \ 2300 FW ITJ 2100 NA 1900 FW _ DEM Profile 1700 0 5 10 15 Line Distance (km)

18. Southern Segment 3300 3100 W Front Segment 2900 ! 2700 Hangingwall -4 2500 g 2300 Footwall IT' 2100 ir 1900 — DEM Profile 1700 0 5 10 15 Line Distance (km)

Figure 4.16. Transverse range topography profiles showing fault segment relief.

79 end f Fi r ont gur of e 4 of th

th .17 e e S 3500 - Max Basin . S t east S ar V - Mn Basin w

e 3000 ath - Front Seg. all rn

- Rear Seg. e sid profil y F - Breach Fault e a

of Rear Seg. (Relay) ul es tl 2000 t

S - Relay Topo south gen t ar 1500 - Front Seg.Topo erat V - Fold Tip Max

all 0 /10 20 30 of 40 50 60 70 81),

ed .. Fold Tip Min e S N y Horizontal distance (km) m usi r oot an n gi g . n

th 3500 3500 g f

e 30 3300 3300 rom D 3100 3100

m 2900 2900 DEM r c y C 2700 --c 2700 2 2500 • 2500 re

ori >cu• 2300 ek E 2300

i ent ru 2100 ILI 2100 n

at 1900 1900 th ed e 1700 1700 north

parall 1500 1500

25 35 40 45 50 55

t el o Horizontal distance (km)

south t o r a n ern ge The fold topography north of the front segment shows a decrease in elevation to the north. The rear and front segment profile elevation at similar. The rear segment fault within the relay is positioned higher than the segment trace to the north. This is interpreted as a result of this part of the fault being abandoned due to breaching of the relay zone forming hard linkage of the rear and front segments (Fig. 4.17 bottom right diagram). There is an associated increase in local relay topography with the overlap of the rear and front segments.

4.4.2 Drainage Spatial hydrologic DEM based analysis carried out using Rivertools software (see Chapter 3), has allowed the semi-automatic calculation of upsteam footwall catchments from the range front outlet along the fault trace and the delineation of drainage in Star Valley. The size and shape of the catchments vary significantly. The largest catchments are located at Strawberry Creek, Swift Creek and Dry canyon, which coincides with the centres of the two fault segments. The larger catchment shapes in plan view are asymmetric at the higher elevations with southward extensions. The morphology of the footwall catchments in the Grover relay show a change south along the range front and will be the focus for further study in section 4.5.

Drainages upstream from the range front show right angled bends at higher elevations. From the range front to the basin the drainages vary in the southern and northern parts of the basin. In the southern basin, drainages flow NW into the Salt River focussed into the Narrows area. In the north basin the drainages mainly flow west into the Salt River. There are drainage deflections around localised topography around the front segment fold tip. In the relay area there are SW-NE oriented drainages where some of the stream patterns show changes in orientation upstream along catchment length. The asymmetry in the catchment and drainage orientations at the higher elevations of the upstream catchments in the Salt River range, indicates a change in local morphology. Upstream migration resulting in headward incision of the footwall catchments may indicate capture of existing upstream mountain catchments forming the asymmetric shape of the upper catchment morphology and drainage patterns.

81

504000 510000

Drainage

Catchments km2 0 - 3 -3- 9 9-40 1 40 - 80

0 km 4

504000

Figure 4.18. Star Valley geomorphology; upstream range front catchments extracted from DEM with overlain drainage.

82 Longitudinal stream profiles, measured using the DEM show the change in channel slope from the basin to the upper mountain range catchments (Fig. 4.19). Twelve longitudinal stream profiles show changes in main trunk stream channel slope upstream from the range front outlets. The profiles show concave to linear shapes with upstream inflections at the higher elevations. Profiles 4, 9 and 11, show concave to linear profiles at low elevations with slope inflections at higher elevations. Profile 9 measured along Willow Creek shows a change in slope at higher elevations. These inflections occur at stream junctions further upstream and may be related to changes in geomorphic processes at the higher elevations especially in the glaciated terrain at the top of the catchments. Longitudinal profiles measured in the Grover relay indicate a change in shape south along the range front.

4.5 Observations of Topography and Morphology of the Grover Relay The tectonic geomorphology mapping results in 4.4 identified the Grover relay area for further study. The study area extends from Swift Creek, near Afton, north to Willow Creek and includes the Grover relay and the front segment fold tip to the north of the front fault. Spatial analysis was carried out on the 10m DEM using Rivertools software (see Chapter 3) to map the topography of the fault segment footwalls and drainage within the footwall catchments and surrounding terrain. Catchment drainage outlets identified on the front and rear footwalls were used for calculating the upstream terrain morphology. Thirty-six footwall catchments were mapped in the relay resulting in a catchment map (Fig. 4.20). Small catchments developed within the triangular facets are not included and are interpreted as secondary geomorphic features. The plan map will be used to qualitatively and quantitatively map the changes in geomorphology within the Grover relay and will be used as a reference in later sections.

83

. . i

., i Location of Longitudinal Stream Profiles See Figure 4.18

_ .

4.

3100 2900 - E 2700 - ,ve 5toPe g 2500 conc- 173>. 2300 - 5119P" ifi'l 2100 - 1900 - , —DEM Profilel 1700

0 1 2 4 5 i Line Distance (km) E 9. Willow Creek

2700

2500 Glaciated Upper Catchments ------0. E Range c 2300 Slope Inflection —II' o Front r; 2100 - \ I 2 ia 1900 - 'I — DEM Profile 1700 , 0 2 4 .5 8 10 Line Distance (krn) VV E

11. Swift Creek

3100 - 2900 - Slope Inflection i2700 ------Ok- g 2500 - Range Front 2300 Lii 2100 -

1900 DEM Profile 1700 -- , -- 0 2 r^ 171 '2 14 '6 Line Distance (krn)

Figure 4.19. Star Valley longitudinal stream profiles measured using the DEM.

84 Visual inspection of the plan view morphology shows a change south along the front fault footwall. The shapes of the canyons south from Willow Creek to Astle Canyon on the front fault generally show elongate forms. The larger canyons in plan view, i.e., Bradshaw Canyon and Dutson Canyon, have wider, irregular upstream shapes. The size and shape of the catchments along the front fault generally increase southwards with a dramatic change at Philips Creek (catchment 12). The shape of this catchment is not elongate and is "wine glass" shaped. The size is significantly larger than any other footwall catchment in the relay. Similar catchment shapes, decreasing in size, occur further south at Jensen Canyon, Blaney Canyon and Anderson Canyon (Fig. 4.20).

Drainage patterns in the front fault footwall show linear streams orientated along the catchment south until Phillips Creek (Fig. 4.21). Drainage patterns in this catchment change significantly with anomalous bends in the stream pattern further upstream (Fig. 4.21). These anomalous changes in stream patterns suggest a change in erosion within the upstream regions. Stream patterns orientated SW-NE located to the north east part of the front fault footwall, are not in the same orientation as modern day patterns of erosion which are aligned west-east on the back rotated front segment footwall. These SW-NE drainage patterns may represent fossilised early drainage formed in the footwall and relay topography. Subsequent footwall displacement and uplift have rotated these drainages into their current orientation.

85 505000 510000

Bradshaw Canyon

Dutson Canyon

Astle Creek

Catchment size km' { - 1 1 - 2 2 - 3 7 3 - 1 0

Figure 4.20. Grover relay fault footwall catchment plan view map extracted from the 10m DEM using Rivertools software. The main catchment names are marked.

86 SW-NE drainage

Bradshaw Canyon

Dutson Canyon

Astle Creek

Figure 4.21. Grover relay drainage map extracted from the 10m DEM. The yellow circle identifies the change in plan view drainage patterns at Phillips Creek.

87 4.6 Front Segment Footwall Catchments

The plan view map of the front fault segment catchments in the Grover relay shows marked differences along fault strike. Further DEM analysis is used to describe the west-east longitudinal footwall terrain shape and the topographic change along the north-south fault trace.

4.6.1 Longitudinal Stream Profiles

Thirty-six trunk stream profiles (see Appendix 1) have been measured along the front fault south from Willow Creek to Swift Creek based on the 23 footwall catchments (Fig. 4.20 and Fig. 4.21). Seven stream profiles are now used to show the change in profile shape along the front fault (Fig. 4.22). The smaller catchments to the north at the front segment tip at Willow Creek have linear shapes similar to profile 3, which is in catchment 2. Further south, the slopes shapes are still linear with "convex up" slope profiles at the upstream elevations close to the drainage divide at Bradshaw Canyon and Dutson Canyon. At Phillips Creek the profile shape changes to a concave shape with similar profile shapes at Jensen and Blaney Canyon. A composite graph plot of the longitudinal stream profile streams (Fig. 4.22b) clearly shows the change from linear to concave shapes along fault strike. These profiles preserve the temporal erosion history of the catchments and the ability to erode the affects of footwall fault displacement. The smaller catchments with linear profile shapes indicate the inability to erode the effects of footwall displacement. There hasn't been enough time or stream power to erode the catchment to form a typical concave stream profile form. The larger canyons i.e. Bradshaw Canyon, Dutson Canyon and Astle Creek also show linear profile shapes along most of the catchment again indicating that these streams have not had enough time and / or stream power to erode the tectonic overprint on the streams formation. The change in slope shape near to the drainage divide corresponds to a change in lithology where the catchments have cut back past the syn-tectonic conglomerate unconformity. The large catchments at Phillips Creek, Jensen Canyon and Blaney Canyon all show concave profiles. These larger catchments have enough stream power to erode the effects of footwall uplifts.

88 Location of longitudinal 3. stream profiles south along front segment. Linear slope

Pn:0101

I2 I 0

Do:lance .Ium I E 7. Bradshaw Canyon 13. Dutson Canyon 2.3 2,50 _—. 1,..0 Convex Slope _ - -- Linear Slope i 2a0

15.52 2 1950 1—eski 0,01,1 1003 --- 1— 1304 Prtillel 850 14,3 •5 2 2:5 [5 I '5 2 25 3

Lim 01010000 Owl Lin. DaitlIKII 0001 E W E 14. Astle Creek 16. Phillips Creek

Z2C.J 3030

21. 2800 i 2100 I 2030 Concave Slope 1 2000 Linea i 2400 .2290 __.------2002 _ _ _ 1— 000 Ch0h101 '850 v000 . - _ . 5 B W Line 0.00ce 00, E' W [me Ckstance ikno E 20. Jensen Canyon 24. Blaney Canyon 2000 _ 2800 2600 2200 Concave Slope k0- MO Concave Slope p300 i .00 2E00 2000 - - -- - I — DEM I.I00e1 _ - - - r— ochi nowl 1900 2 3 . . . Una Dintonce(100 , E W i_,L.:',:.:, rr. e

2800 — Stream 3 — Bradshaw Canyon 7 Dutson Canyon 13 2600 Astle Creek 14 — Phillips Creek 16 --Jensen Canyon 20 c 2400 — Blaney Canyon 24

C

2200

2000

1800 0 1 2 3 4 5 6 Horizontal Distance (km)

Figure 4.22. a) Longitudinal stream profiles measured using the 10m DEM, b) composite graph show the change in profile shape south from the fault tip.

89 4.6.2 Footwall Catchment Topography

West-east topographic profiles show the longitudinal shape of the footwall catchment. West — east swath profiles are generated using the 10m DEM where the catchment boundaries in Figure 3.20 are used to define the individual catchment extents. Each catchment is isolated from the surrounding 10m DEM pixels using an inside polygon test in ERMapper (Chapter 3) with the cells outside the catchment set to null values. The extracted catchment is then rotated so that each of the catchment trunk streams are orientated North — South. Exported as ascii text files, the cross-sectional maximum and minimum (in effect the stream profile) swath profiles are calculated along the rows of the extracted height values using a spreadsheet database. The maximum - minimum difference i.e. calculation of the vertical difference in elevation for each pixel along the catchment, shows the change in footwall transverse topographic shape. The maximum profile is effectively the longitudinal footwall profile that has been modified by erosion. Landsliding including hillslope erosion and competition with adjacent catchments reduces the longitudinal topographic profile i.e. lowering the idealised normal fault footwall longitudinal profile that would form assuming no erosion.

These graphs show the west - east longitudinal topographic shape of the front segment footwall. Eight of the footwall swath profiles (see Appendix 1 for the 23 swath profiles) are used to show the change in longitudinal footwall topography related to topographic divide of the footwall. Canyon 2 maximum and minimum swath curves are broadly linear with its difference curve showing a slight asymmetry / irregular shape with an elevation difference of 30m (Fig. 4.23a). Canyon 4 is asymmetric with a well-formed cross-sectional shape of the range front triangular facet to the west. Its difference curve clearly shows topographic asymmetry with a maximum elevation difference of 90 m. Bradshaw Canyon swath curves are different showing the same asymmetry but with a maximum difference of nearly 200 m. There is a difference in the position of the topographic divide and the drainage divide, which is positioned 500 m to the east. Canyon 8 swath is typical of the smaller catchments between Bradshaw Canyon and Dutson Creek. The curve is strongly asymmetric with the maximum elevation difference at the range front. Dutson Canyon is similar in shape to Bradshaw Canyon with an asymmetric difference curve but with the maximum elevation of nearly 200 m towards the range front. Again, there is a difference in the position of the topographic divide relative to the drainage divide, which is located 900 m to the east. Astle Creek shows the most pronounced asymmetric swath profile along the front fault with a maximum 90 difference of over 200 m at the range front. Phillips Creek swath is different to the canyon swaths to the north. Its profile has an asymmetric bimodal shape with its maximum profile having a similar asymmetric range front topography to the canyons above but with a dramatic change in the upstream topography. The maximum difference curve clearly shows the bimodal shape with a peak of nearly 600 m to the east of the catchment. Jensen Canyon and Anderson Canyon have similar swath and difference curves but are not as pronounced.

Comparison of the graphs shows a dramatic change in transverse catchment profile shape along fault strike (Fig. 4.23a-c). The profiles show a change from irregular / linear shaped curves to a more asymmetric curve with steeper slopes / peak on the western range front resulting in a bimodal shape curve. The canyons north of Phillips Creek show a change in transverse footwall shapes. The footwall canyons near to the tip show the shape of the footwall in the western side of the footwall with the topographic divide at the same position as the drainage divide (Canyon 2 & 4 in Fig. 4.23a). However, Bradshaw Canyon and Dutson Canyon show a relative increase southward in topographic and drainage divide position. Astle Creek is the only transverse profile that shows a topographic cross section across the footwall topography. This curve is asymmetric and shows the profile of the back rotated footwall.

Interpretation of these swath curves indicates the imprint of the fault development over the terrain evolution. However, temporal erosion modifies this structurally dominated topography producing asymmetric curves with an eroded footwall form. Bradshaw and Dutson Canyons show this asymmetry but with an irregular curve towards the upstream part of the catchment. This is interpreted as inherited topography that has become preserved within the footwall. Phillips Creek swath profiles are dramatically different to the rest. The maximum and minimum profiles show a change in catchment longitudinal shape especially in the upper catchment area and the catchment length is much longer. The maximum — minimum difference profile is different, showing a bimodal shaped curve. The first peak corresponds to the front fault scarp and the second corresponds to the topography of the drainage divide in parallel with the rear fault topography. This graph shows the pronounced change in minimum curve slope angle. Similar changes are observed in the larger catchments to the south i.e. Jensen Canyon, Blaney Canyon and Anderson Canyon. Canyon 20, which is an example of the smaller catchments to the

91 south, has similar transverse topographic profile shape and size to the smaller catchments near to the tip.

Canyon 2 2200 300 I- Max-rnin Irregular surface 250 2150

-g 200 2100 C 0 o 150 ▪'>1'i 2050 (1)• 100 LIJ 2000 50

1950 0.6 0.5 0.4 0.3 0.2 0.1 0 0.6 0.5 0.4 0.3 02 0.1 Horizontal Distance (km) Horizontal Distance (km) Canyon 4 2300 300

2250 Asymmetric 250

E 2200 topography E 200 Fault scarp Inherited topography C C .9 2150 9 150

- 2100 EuN- 100

2050 • 50 - max - ram 2000 0 0.8 0.6 0.4 0.2 0 0.8 0.6 0.4 0.2 0 Horizontal Distance (km) Horizontal Distance (km) Canyon 6: Bradshaw Canyon 2300 300 I--1 2250 Topographic 250 Topographic divide - 2200 divide drainage divide 200 2150 C C .2 2100 .2 150 To 7>6 2050 . 100 LL 2000 Drainage divide 50 1950 - max -min 1900 0 2 1.5 1 C.5 0 2 1.5 0.5 0 Horizontal Distance (km) Horizontal Distance (km) Canyon 8 300 - max-rn 2250 Asymmetric 250 2200 topography \ Fault scarp E zoo Eroded footwall *--2150 0 profile = 2100 tso CD 2050 La 100 2000 50 1950 max ".. "mm 0 191 6 1 4 1.2 1 08 0.6 OA 0.2 0 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 Horizontal Distance (km) Horizontal Distance (km)

Figure 4.23. a) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography.

92 Canyon 10: Dutson Canyon 2350 300 ma.... 2300 Topographic Topographic divide - 250 2250 divide drainage divide 22°° E 200 c 2150 0 0 150 tT5 2100 w 2050 • 100 111 tu 2000 Drainage divide 50 1950 min 1900 0 2.5 2 1.5 1 0.5 0 2.5 2 1.5 1 0.5 0 Horizontal Distance (km) Horizontal Distance (km)

Canyon 11: Astle Creek 2400 300 2350 Asymmetric 1----1 250 Eroded footwall 2300 topography ? 2250 profile E 200 2200 0 2150 to• 150

41) 2100 100 11 2050 1.1.1 2000 - max 1950 - min , 1900 4 3.5 3 2.5 2 1.5 1 0.5 0 3.5 3 2.5 2 1.5 0.5 Horizontal Distance (km) Horizontal Distance (km) Canyon 12: Phillips Creek 3200- 600 Bimodal topography Twin peaks 3000 14 500 2800 Grover Relay E 400 c 2600 4 Asymmetric TO 300 > 2400 topography co a) Lys 200 - 2200

rna. 2000 - 100 Chan • e in slope an • le - mm

18°°5 4.5 4 3.5 3 2.5 2 1.5 0.5 0 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 Horizontal Distance (km) Horizontal Distance (km) Canyon 14: Jensen Canyon 2900 600 - mac-mm I 2800 Bimodal to•o•ra.h 500 2700 Twin peaks -E 2600 Fault scarp E 40o • 2500 0 2400 300 ▪> 2300 2 1"•••,,,..,, 11.1 2200 uj 200 2100 (j\--/-Peak caused by - Max 100 2000 dogleg in catchment han•e in sloe an le - mm 1900 0 3.5 2.5 2 1.5 1 0.5 0 3.5 3 2.5 2 1.5 1 0.5 0 Horizontal Distance (km) Horizontal Distance (km)

Figure 4.23. b) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography.

93 Canyon 18: Anderson Canyon 2800 600 I— ,r4K rr.,, I 2700 Bimodal topography Twin peaks 500 2600 E 2500 Fault scarp E 400 c 2400 0 .•0 300 ▪co 2300 CO

El 2200 12• 200 2100 100 Change in max 2000 slope angle 1900 2 1.5 1 0 25 2 15 0.5 0 2.5 0.5 Horizontal Distance (km) Horizontal Distance (km) Canyon 20 2600 300

2500 Asymmetric 250 E 2400 topography — 200 •O 2300 0 co '17, 150 >w 2200 [171 • 100 2100

50 2000 --- max mm 1900 0 1 6 1.4 1.2 1 0.8 0.6 04 0.2 0 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 Horizontal Distance (km) Horizontal Distance (km)

Figure 4.23. c) Maximum / minimum transverse swath profiles and maximum / minimum difference profiles of front footwall catchment topography.

4.6.3 Front Segment Topography

Maximum topographic profiles measured from Swift Creek to Willow Creek shows a general increase in topography south along fault strike (Fig. 4.24). However, at Phillips Creek there is a decrease in topography. South of Phillips Creek the maximum topography of the range front is less clear and is difficult to distinguish from the drainage divide. The drainage divide topographic profile also shows an anomalous change at Phillips Creek with a dramatic change in elevation. The increase in relief is located on the north side of the Phillips Creek catchment. The drainage divide is different to the maximum topography in the fault tip area north from Phillips Creek to Willow Creek with the diversion in the plan view centred on Bradshaw Canyon. This change in the topographic and drainage divide south from the fault tip is similar to the observation of the transverse swath profiles made above. The topographic anomaly at Phillips Creek suggests that this catchment has developed differently to the ones further north.

94

Fi tu gure 4 Front Segment Footwall sua Topographic Profiles Maximum Front Segment Topography ain .24 1. p n . F s r! 2700 ront E i.

I 2500 2 2300 se 1; gment

0I N >w 2100 — DEM Profile

111 Ca 1900 2 4 6 8 10 12 14 maxi 1U

.1 Line Distance (km) N S mum

t Topographic anomalies at Phillips Creek catchment o po gr

a 2. Front Segment Drainage Divide Topography phi 3200 c

prof 3000 —DEM Profile

il t. 2800 e

and d ,-° 2600 15 tt 2400 r

ai r\ 2200 n a

ge di 2000 0 2 4 6 8 10 12 14 16 18 20 vid N Line Distance (km) e pr ofi l e 4.6.4 Catchment Comparisons Quantitative analyses of front fault footwall catchment basins using the DEM have been carried out in Rivertools. Individual basin characteristics have been calculated allowing a comparison south along front fault strike between Willow Creek and Swift Creek (Fig. 4.25a)). Footwall catchment area (km2) shows variation along strike with a large catchment size at Phillips Creek. Catchment maximum diameter varies south along fault strike with the largest diameters centred again on Phillips Creek. These two graphs show at least two populations. One population is related to a change in catchment shape and size with Phillips Creek, Jensen Canyon, Blaney Canyon and Anderson Canyon all showing similar values. These graphs show a change south along the fault towards Phillips Creek. The size and shape then decrease overall towards Swift Creek.

Catchment relief, the difference in height from the upstream catchment elevation and the range front outlet elevation shows an increase in relief south along fault strike and shows a similar shape to the drainage divide topography profile (Fig. 4.24). At basin 12, Phillips Creek, there is a dramatic increase in relief, which then decreases towards Swift Creek. Again the graph shows at least two data populations. The catchments with the greater relief i.e. catchments 12, 14, 17, 18 19 and 20 are to the south of Phillips Creek. The graph of longest channel length shows an overall increase south along fault strike towards Phillips Creek with an overall decrease south to Swift Creek. Total Channel Length shows a similar trend with an anomalously large value located at catchment 12.

Comparison of the transverse catchment topography has been qualitatively described above based on change in shape along fault strike (Fig. 4.23a-c). The difference in maximum and minimum values along the catchment is calculated by measuring the vertical difference in height value along each 10 m swath pixel. The maximum and average difference values are then calculated for the whole catchment, which allows a comparison of catchment along strike. The graph shows a gradual increase south along strike with a large increase at Phillips Creek (Fig. 4.25b). There is a decrease towards with Jensen Canyon, Blaney Canyon, and Anderson Canyon having similarly large values.

96

Catchment Area (km2) Catchment Diameter (km)

4

335 4 to Value 13 LoValuel

a 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0 11 12 3 14 15 16 17 18 19 20 21 22 23 Catchments South of Fault Tip Catchments South of Faun Tip

Catchment Relief (km) Longest Channel Length (km)

7

4 Invatoer 13 0.4 2

n n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 9 20 21 22 23 Catchments South of Fautt Tip Catchments South from Eau t Tip

Total Channel Length (km)

Figure 4.25. a) Catchment characteristic comparison along front footwall strike. Calculations have been made using the 10m DEM using Rivertools software.

97

Maximum / Minimum Swath Difference

600

o Maximum Difference 500 E ■ Average Difference sa 400

300

:a 200 W 100

0

35 ❑ Average Slope 30 • Mode Slope 25 o Median Slope

13 20 co c 15

0- 10 Co 5 I 0 1t mm r r1 S A h r ct. N.ni N. v6 <1 0 0 10 15, ,173 .‘v Canyon Number

Figure 4.25. b) Composite graphs of swath calculations of maximum / minimum difference and minimum slope angles.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 50 k -50 11-111[11 PP

-150 - c

I -250 • to Max w • Min -350 ❑ Median o Mean -450 Catchment number Slope calculations are made using the minimum profiles and extracted stream profiles. The slope angle change along the main trunk stream is calculated from every adjacent 10 m DEM pixel along the stream. The maximum, average, mode and medium values are calculated from the adjacent pixel slope calculations. The graph shows several trends in slope angle along fault strike. We would expect the slope angle to increase along fault strike away from the tip assuming there was no erosion along the footwall. Differential erosion along the footwall is controlled by catchment area and catchment steepnest. We know that landsliding at the fault scarp drives footwall catchment development (Densmore et al., 1998) hence we would expect a decrease in slope angle along strike with increasing displacement. It is clear from the composite graph that there is a decrease in slope angle from 15 degrees at the tip to 5 degrees at Phillips Creek. However, there is a change in the expected trend towards Swift Creek with a general increase in slope. Catchments 14 and 18, Jensen Canyon and Anderson Canyon have similar catchment mode slope to Phillips Creek. The general increase in slope angles south from Phillips Creek produce the largest average slope angles of any catchment, with the exception of canyons 14 and 18, may indicate preservation of earlier catchments within the footwall at the expense of more rapidly developing adjacent catchments.

The curvature or concavity has been calculated for each catchment trunk stream from Willow Creek to Swift Creek on the front segment (Fig. 4.25). Maximum, minimum, mean and median values of the difference of observed elevation to a linear regression (see chapter 3 for method) line fitted to the catchment relief indicate changes along front segment strike. Phillips Creek (catchment 12) shows a high negative minimum value indicating a concave profile shape.

4.6.5 Rear Segment Catchments

The rear segment catchments show similar stream profiles and topography to the front. In plan view the canyons are elongate with wider upstream diameters. The main difference in the rear catchments is the change in relief. The rear stream profile curves are again linear but the swath profile is different to the front segment catchment. The maximum curve is asymmetric but has a steep box-type shape at the high elevation.

99 a) Canyon 32

2800

2600 -

2400 - ` Linear profile o 2200 -

ir4 2000 -- DEM P,Aie 1800 0_5 1.5 2 2.5 3 3 5 Line Distance (km)

Canyon 30

2800

2600 E Linear profile g 2400

2200

w 2000 DEM Profilel 1800 0.5 2 2.5 3 Line Distance ikm) E

b) Canyon 30 3000 600 2900 500 2800 Rear Segment 2700 Facet 400 0 2600 8 300 > 2500 80 co irj 2400 LTI 200 2300 -max 1 100 2200 --min

21°°0 0.5 1 15 2 2.5 0 0.5 1 1.5 2 25 Horizontal Distance (km) Horizontal Distance (km)

Figure 4.26. a) Longitudinal stream profiles of rear footwall, b) swath profile of canyon 30.

4.6.6 Catchment Capture

Catchment morphology and drainage patterns in the Grover relay do not show the same geomorphic features consistent with the relay drainage focusing model discussed in Chapter 2 (Leeder et al.,1987 and Jackson et al., 1994) and more importantly the lateral propagation model indicated by cross cutting relay drainage (Jackson et al., 1994). There is no well-developed relay ramp drainage and small upstream catchments. There is not a large fan at the base of the relay with large hangingwall fans located either side of the relay. There are drainages that cut across the relay but their catchment and drainage morphology does not match the linear shaped one described in the Jackson and Leeder model. Footwall catchment size and shape varies in the Grover relay area. There 100 is a dramatic change in catchment properties and anomalous upstream drainage patterns at Phillips Creek. This catchment cuts across the relay. Drainages in this catchment change upstream from straight elongate morphology, similar to the smaller catchment streams north along the front fault, to more densified patterns with some stream junctions having right angled bends. The "wine glass" shape of the catchment differs to the more straight elongate catchments north along the front fault. The geomorphic results show the change in footwall catchment along range front. Footwall catchment development depends on catchment size, relief, channel slope, baselevel fall at range front and stream erosion (Burbank et al., 1999; Burbank et al., 2000; Densmore et al., 2003). Large catchments with greater relief and an increase in hangingwall / footwall displacement are more able to erode headward upstream. As a result these catchments are more likely to erode further headward into the footwall with the drainage divide passing the maximum footwall topography. Increased of the catchments erodes the topography normal to the footwall range. The increase in the difference between maximum footwall topography and the drainage divide indicates the temporal change in catchment development south along the front fault. The dramatic change in catchment properties at Phillips Creek indicate that this catchment developed differently to the ones further north. The location and morphology suggests that the headward incision of the catchment has resulted in the capture of catchments in the relay and rear segment. Jensen Canyon, Blaney Canyon and Anderson Canyon show similar catchment morphologies to Phillips and are interpreted as similar catchment capture events. These indicate a temporal catchment capture migration northwards where several capture events have occurred.

101

3100 Bimodal topography Merged Astle Creek 2900 14 topography and rear segment catchment 2700 30 swath profiles 8 .17i 2500 Asymmetric footwall 14 11 2300 /V t Overlap of merged 2100 Change in — Max topography. minimum slope --- Min 1900 0 1 2 3 4 6 7 Horizontal Distance (km)

3200 Bimodal topography Phillips Creek catchment 3000 14 swath profiles 2800 E migration? 2600 O0 Asymmetric footwall > 2400

—Max 2200.

2000 Change in minimum slope 1800 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Horizontal Distance (km)

600 Merged Astle Creek 500 topography and rear segment catchment 30 400 .g swath profile difference 300

LTJ 200

100

7 Horizontal Distance (km)

600 Phillips Creek swath profile 500 difference

400

5g 300

200 111

100

0.5 5 2 2.5 3 3.5 4 4.5 Horizontal Distance (km)

Figure 4.27. a) Addition of Astle Creek and canyon 30 swath and difference profiles.

102 The comparison and combination of front / rear longitudinal stream profiles and swath profiles show striking similarilies to the Phillips Creek catchment morphologies. Astle Creek profile, which has a topographic profile across the front fault footwall including the back rotated footwall, combined with rear catchment 30 have a combined bimodal swath topographic shape (Fig. 4.27). The combination of the minimum profiles (longitudinal stream profile) shows the combination of two linear shape profiles. Over time this is interpreted to give the same concave profile as Phillip Creek because of the increase in catchment size resulting in increased stream power to effectively erode the stream channel. The combination of the differenced curves also show striking similarities to Phillips Creek. These graphs indicate that with further fault development it is possible that Astle Creek or Dutson Canyon will be the future sites of catchment capture.

4.6.7 Front Segment Topography Transition The front segment longitudinal stream profiles, transverse swath profiles and fault parallel topography indicates that there is a transition in front footwall erosion south from Willow Creek to Phillips Creek. The smaller catchments to the north have eroded into the west side of the front range front where the topographic and drainage divides are the same. Bradshaw Canyon and Dutson Canyon have different profiles with plan view geometry and swath profiles indicating that the drainage divide is 500 m to 900 m, respectively, further east. Astle Creek is the largest catchment north of Phillips Creek and cuts across the front segment footwall. This change in front segment footwall catchment development is interpreted as a spatial and temporal transitional change in headward footwall catchment migration south from the front fault tip. Similar changes in footwall topographic and drainage divide on a larger scale are observed in the eastern basin and range (Habor, 1997). The position of the topographic and drainage divide indicates the erosional history of the catchments, in particular the headward migration of the catchment into the fault footwall.

103 4.7 Grover Relay Geomorphology

The Grover Relay zone is approximately 2 km wide and 4 km long (Fig. 3.28). To the west there is the back-rotated topography of the front segment footwall and to the east there is the range front of the rear segment with degraded triangular facets. A north- south orientated stream is located between the topography of the front footwall and the relay slope and is interpreted as the axial relay catchment stream. A series of southeast to northwest orientated drainages join this axial stream and are linked to the rear segment footwall catchments. The axial stream, which in modern times is dry, follows the topography of the front footwall northwards joining Willow Creek. To the north of the front footwall there are linear drainage patterns oriented SW-NE juxtaposed to more west-east oriented patterns which join the relay axial stream.

508000 510000

SW -NE O Ct" drainage N- A

11111

Axial strea W-E lbh drainage. 0 O

rotate 00 .11

y fault / /

508000 510000

Figure 4.28. Plan view map of the Grover relay with drainage overlain on a sun-shaded DEM.

104 The aspect and slope maps derived from the DEM show the morphology of the relay topography (Fig. 4.29). They clearly delineate the front back-rotated footwall and rear fault topography. The slope of the front back-rotated footwall dips 10-20 degrees to the east-north east. The rear fault footwall range front scarp are clearly defined dipping 30+ degrees to the west.

The relay topography gradually increases in elevation from 1-2 degrees in the north to 10 degrees to the south. The aspect of the relay topography dips northwest. The slope and aspect of the relay drainage that joins the axial stream indicates an asymmetric transverse fluvial channel morphology. This will be further described in section 4.7.2.

The topography of the relay increases to the south to the edge of Phillips Creek with an elevation change of 200 m (Fig. 4.30). The west- east topographic profiles define the cross-sectional form of the relay indicating the decrease in elevation from the front footwall to the axial stream. There is a stepped increase in elevation eastward at the relay fault and the rear fault. A west — east orientated profile at the edge of Phillips Creek has a "U" shaped curve which corresponds to the position of the axial stream (Fig. 4.30a). This is interpreted as a wind gap that formed from the beheading of the axial relay stream.

105

508000 510000

0 0 0 ct

Slope I 10-1 L_J 1-3 0 MN 3-5 0 MO 5-10 \ I I 10-20 MN 20-30 EN 30+ - no data

A 0 km 1

508000 510000

508000 510000

0 0 CP O O O Aspect Flat (-1) North t0-22 5,337 5-360) I I Northeast (225-67.5) East (67 5-112.5) 1111 Southeast (112.5-157 5) I 1 South (157.5-202.5) Southwest (202.5-247_5) ME West (247 5-292.5) ▪ Northwest (292.5-337 5) No Data

0 0 o vrt O O

0 km 1

508 000 510000

Figure 4.29. Aspect and slope plan view maps generated using the DEM.

106

a) Profile 1

2900 — DEM Profile 2700

e 2500 Slope Inflection

2300

• 2100

1900

2 4 N ° Line Distance (km) Profile 2

2400 Rear Segment Facet -g- 2300 "N4, C Relay Fault Scarp 2 2200

2100 — DEM Profile' 2000

0 0.5 1 1.5 2 2.5

Una Distance (km) E Profile 3

2300 "U" Shape Profile ? 2250 . Wind Gap? ▪ 2200

C la 2150 I —.. DEM El"..31jda 2100

D 5 1.5 Line Distance (km)

b) Stream 1

2250 I DEM Profile' 2200 Linear - 1- 2150 Convex S/ope 5 2100 2950 2 2000 1950 1900

0.5 1 2 2

Line Distance (km) E Stream 2

2200 - DEM Profile 2150 -

c 2100 Concave Slope 2050

▪ 2000

1950

0.5 1.5

W Line Distance (km) E

Figure 4.30. Topography and stream profiles in relay measured using the DEM.

107 4.7.1 Front Segment Footwall Drainage The anomalous drainage in the tip of the front fault rotated footwall shows different plan view orientation to drainages to the south (Fig. 4.28 and 4.29). The more west-east drainages flow eastward into the relay axial stream. The longitudinal stream profiles for these drainages show linear — convex shape curves at the tip to more concave curves to the south (Fig. 4.30). This change is interpreted as a change in fluvial erosion. The north tip drainage in interpreted as fluvial channels that have been preserved within the terrain and are no longer active. The west — east drainage has concave profiles indicating active fluvial erosion. This drainage is interpreted as streams that have developed more recently in response to the back rotated front footwall.

4.7.2 Grover Relay Drainage The SE-NW drainage that joins the N-S relay axial drainage to the west is generally connected to large, more west-east orientated, rear fault footwall catchments (Fig. 4.29). The change to a NW direction indicates the change in topographic slope within the relay. Three well-defined channels are clearly outlined in the aspect and slope maps (Fig. 4.29). Topographic profiles orientated normal to the drainages show cross sectional asymmetric channel morphology with a steep western side and a gradual decrease in slope to the east (Fig. 4.31). The morphology of these streams is different to typical "u" or "v" shaped incision channel morphologies generally observed in modern day stream incision (Burbank et al., 2000).

108

A Slip drainage 1 nrti

SW NE

2200 2180 - Profile 1 2160 - 2140 - ro 2120 LU 2100 2080 02 D8 1 12 Horizontal Distance (km)

2070 2060 Profile 2 Asymmetric topoography 2050 t 2040 channel 4---- SD2 channel w 2030 SD1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Horizontal Distance (km)

2050 2040 Profile 3 Asymmetric topoography 2030 t 2020 - SD2 channel 2010 4-- SD3 channel 0 0 2 0.4 0.6 0.8 1 1.2 1.4 Horizontal Distance (km)

2030 2020 Asymmetric topoography f, 2010 ------..-----4-_, ------....„,y------m 2000 SD2 channel 4------1990 Profile 4 SD3 channel 1980 0.2 OA 0 6 0.8 1 1.2 1.4 1.6 Horizontal Distance (km)

Figure 4.31. Relay drainage map with SW-NE transverse topographic profiles measured using the DEM showing cross sectional fluvial morphology.

109 4.7.2 Drainage evolution

The spatial and temporal topographic development within the relay zone related to the displacement history on the two fault segments has resulted in the development and change in slope of the relay over time. The drainages that cut across the relay have interacted and responded to this temporal change in relay topography.

A model of coupled relay topography and drainage evolution is proposed (Fig. 4.32 a- c). At an early fault stage the soft linkage between the front and rear forms shallow sloped relay topography dipping approximately to the north- northwest. Early fault stage axial drainage forms and drainage related to the developing rear catchments incises into this proto-relay slope.

As the fault segments grow and accumulate displacement, the topographic slope of the relay increases in elevation. The orientation of the regional relay slope tilts towards the northwest. The drainage response to the accumulated change in relay slope results in incision on the NE side of the fluvial channel. This NE migration of the channel results in a change in the plan view orientation with the drainage "slipping" to the N-NE forming asymmetric channel cross-sectional morphology. This "slip drainage" records the spatial and temporal interaction of the channel incision into the evolving fault controlled topography.

As the relay develops a relay fault cross cuts the relay interacting with the slip drainages. The NE drainages at the front fault tip become preserved in the back footwall topography forming linear — convex stream profiles. New active drainage forms on the footwall with more west-east oriented streams flowing eastward into the relay joining the axial stream.

110 Slip Drainage Evolution

1: Early Extension Drainage Development in Relay

me Seg

(1) Rear

Drainage at ing (i) Fault Tip

LL

Drainage Development in Relay

2: Drainage Migration in Relay resulting from Front Segment Footwall uplift and Rear Segment subsidence.

Drainage Migration Forming Slip Drainage

Figure 4.32. Model of Grover relay terrain development, 1) early extensional fault development and 2) drainage migration.

111 3: Latest Holocene ruptures along Breaching Fault has displaced terraces and slip drainage geomorphology.

•

iln-active streams preserved in back rotated footwall

Change in drainage pattern of Modern drainage

Breaching Fault displaces Slip Drainage

Evidence for latest I Fault (no evidence for latest Holocene Ruptures rupture)

Figure 4.32. Model of Grover relay terrain development (continued), 3) latest Holocene ruptures cross cut the drainage.

112 4.8 Grover Relay Fault The plan view trace of the relay fault extends for 1 km oriented SW-NE and located in the centre of the relay (Fig. 4.33) between the front and rear segment footwall. The latest Holocene rupture trace at Willow Creek (after Piety et al., 1992) is in a similar plan view orientation to this new relay fault. Also, a line of circular depressions, several metres wide, has been mapped in a similar orientation at the upper southern reach of Astle Creek near to the drainage divide (Fig. 4.33).

An oblique aerial photograph orientated SW-NE facing NW (Fig. 4.33) shows the forested breaching fault study area. The trace of the fault is delineated by a row of trees along the footwall, which in places is clearly defined next to flat un-vegetated terrace areas. The photograph also shows the northern section of the relay and the relationship to the back rotated footwall of the front fault. The along strike morphology of the fault is expressed in a series of faulted alluvial terraces and triangular facets which cross cuts the fluvial channels in the relay. DEM analysis and topographic profiles, together with field mapping are used to describe the relative change in fault morphology along strike. Field mapping involved a series of transverse and parallel EDM (laser ranged) levelled topographic profiles (see Chapter 3). The fault scarps dip northwest and are generally covered with dense vegetation (Fig. 4.34). The footwall and hangingwall profiles are defined by changes in vegetation, often with rows of trees lining the footwall topography (Fig. 4.34).

113 a

5050Cc

b) Oblique aerial photograph view of relay. Facing NW. Back Tilted Front Segment Fault Tip Footwall Fold

Figure 4.33. a) Plan view map of the breaching fault, generated using the DEM, showing the hangingwall and footwall trace, b) annotated aerial photograph of relay fault.

114 Figure 4.34. Photographs of fault scarp in the field near to triangular facet three (TF3) (see Fig. 4.35 for location).

115 4.8.1 Grover Relay Fault Topography The hangingwall and footwall trace of the breaching fault has been mapped from the DEM and modified by field mapping (Fig. 4.35a). The presentation of the following graphs and sketches of the fault have been orientated NE-SW, facing SE at the fault scarp (Fig. 4.35). SW-NE orientated topographic profiles generated using the DEM, indicates the footwall and hangingwall morphology along fault strike. These profiles have been projected onto a line orientated 040° (the average along fault strike orientation) for comparison and show a decrease in footwall topography to the south west (Fig. 4.36a). A displacement profile resulting from the subtraction of the hangingwall profile from the footwall shows a complex displacement profile along strike (Fig. 4.36b). It shows a series of triangular facets (TF) and terraces (T) along fault strike. A schematic sketch based on these DEM profiles and field observations summarise the fault morphology along strike (Fig. 4.37). This diagram shows the positions of antecedent streams (STR) and faulted terraces (FT). The hangingwall profile has a broad asymmetric hill shape with steeper sides to the south west similar in shape to the cross-sectional fluvial morphology of the slip drainages. The footwall profile along strike shows a more complicated shape with three dominant triangular facets decreasing in elevation to the south west, marked TF 1-3 in Fig. 4.37. The facets themselves are asymmetric in that on the north east side there is a flat horizontal terrace. There are three flat terrace areas marked FT1-3, with a fourth area marked as FT4. Antecedent streams cross normal to fault strike and are located to the SW of each facet, marked as STR 1-4.

Field mapping in these areas has been carried out to further describe the fault morphology. The latest fault displacements are expressed as a series of faulted terraces along strike. In order to clarify this profile shape (Fig. 4.37), targeted field mapping of the scarps has been made to check the change in relative shape along strike.

116

508500 509500 b)

Photograph of Relay: Facing SE.

Grover Relay Fault

Figure 4.35. a) Plan view map of relay fault showing locations of measured fault traverses, b) annotated photograph showing relay morphology and preferred orientation of fault for presentation.

117

a) Footwall — Hangingwall Profile

2150

2100 - 1=1

2050

200 400 600 800 1000 1200 1400

b) Displacement Profile

so 10 - IJ 20 - 0- 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300

Figure 4.36. (a) Graph of hangingwall and footwall profiles orientated NE-SW measured using the DEM locations shown in Figure 3.35, (b) graph showing the subtraction of footwall profile from hangingwall profile resulting in a displacement profile along strike.

The best area for observing the geomorphology of the fault scarp along strike is between triangular facets TF1 and TF2 (Fig. 3.37). This sketch shows the along strike three- dimensional relationships of the geomorphology of the fault scarp and faulted terrace. The most striking feature is the asymmetry of the fault scarp footwall topography along strike. Fluvial incision by the antecedent stream has cut two localised terraces marked as Ti and T2 forming a "v "shape channel geometry. There is a step in the terrain (T3) to the west forming a flat terrace area for 30 metres along strike (FT2). At the outlet of the stream at the fault scarp there is a change in the topography forming a broad shallow lobate feature interpreted as a localised sedimentary fan. Along the base of the scarp the topography changes in angle gradually flattening out into the hangingwall. This is interpreted as localised talus resulting from erosion of the fault scarp.

118 and

h STR1 FT1 TF1 Facing SE e 4 an gi

.37 NE STR2 FT2 n TF2 gwall . S STR3 FT3 SW

ch TF3 STR4 FT4 emati geom c

or Slip drainage 1 pers ph Minor Fault linked to large

ol scarps upstream pecti o

gy b catchment. ve

ased Slip drainage 1 linked to Slip drainage 3

sk Local drainage large inboard fault ridge Slip drainage et

on fi upstream catchment. ch 0 m 100

of eld th ob e b servati r eachi NE SW on n g f s

and TF1 aul DEM t sh owi profil n Stream STR3 g th es e f Hangingwall 0 m 20 . oot Fan Stream STR2 wall

4.8.2 Hangingwall - Footwall Topography Profiles Hangingwall and footwall profiles measured using the EDM levelling method (see chapter 3) along the fault scarp to the east of triangular facet 1 (TF1) and the area between triangular facets 2 and 3 (TF 2 & 3) verify the change in topography along fault strike (Fig. 4.38). Each of the footwall profiles show the same westward step from the stream channel up to a flat footwall terrace (FT 1-3). FT 1 and 2 have an elevation change of 10 m with FT3 only showing a 2 m. This indicates that the scarp footwalls decrease in elevation to the west. The corresponding hangingwall profiles show a decrease to the west.

4.8.3 Transverse Fault Topography Profiles The displacement curve measured using the DEM above (Fig. 4.36b) was verified by measuring a series of transverse profiles along strike. Fourteen transverse fault profiles were surveyed using the EDM laser ranging levelling method (Fig. 4.39b). Positioning of the profiles were selected at points along the fault to show the relative change along strike (Fig. 4.35a). Sections normal to the maximum elevation point on the triangular facets and the midpoint of the faulted terraces were selected. The EDM profile start and end points were positioned to pick up the hangingwall and footwall slopes either side of the fault scarp. The surface-slope angles, scarp-slope, scarp heights and vertical surface displacements were measured using the methods of Bucknam and Anderson (1979), which is the same method that previous authors used to map the terraces at the footwall outlets in Star Valley (Piety et al., 1986; Warren, 1992).

120

FT1 - Footwall traverse

HT1 - Hangingwall traverse

NE SW 15 - E o- STR1 FT1 8 5 •

-5 -10

20 40 60 80 100 120 Horizontal distance (m) Footwall traverse 1 — profile

E 5 5 0 -;-- .> -5- ...... -..---- m2 -10 w 0 20 40 60 80 100 120 Horizontal distance (m) Hangingwall traverse 1 —Profile 25

20 • i 15 • STR2 FT2 C I 1 5 • ur 0 -5 0 50 100 150 200 Horizontal distance (m) Footwall traverse 2 Profile

E 5 g 0 I - 0' "5 Ili -10 0 50 100 150 200 Horizontal distance (m) Hangingwall traverse 2 - - Profile

1 4 5 2 STR3 FT3, = o E -2 i 0 50 100 150 Horizontal distance (m) Footwall traverse 3 --- Profile

£ 5 c 0 o ----- 1 -5 : -10 W 0 50 100 150 200 Horizontal distance (m) Hangingwall traverse 3 — Profile

Figure 4.38. Footwall and hanginwall profiles measured using the EDM verifying the change in shape along fault strike.

121 The scarp profiles show typical fault profiles indicating a steepening in profile slope at the fault trace. There is a change in slope length and shape between the terraces (FT) and the triangular facets. The graphs of the triangular facets show a number of inflections along the scarp from the hangingwall to the footwall crest. Triangular facet 2 profile (Fig. 4.39a) shows a change in slope on the fault scarp. There is an inflection on the fault scarp, which flattens out towards the footwall crest. These inflections were observed in the field while carrying out the survey. These are interpreted as temporal slope erosion to multiple earthquake events. The footwall terrace profiles show typical transverse fault profiles with slope inflections indicating the eroded profile surface of the fault (Fig. 4.39a). The graphs for all measured scarps can be found in Appendix 1 and the measurements are summarised in Table 4.2.

The transverse fault profiles allow the measurement of the vertical distance between the projected slopes of the hangingwall and footwall (Bucknam et al., 1979) indicating vertical scarp displacement. The vertical distance plotted against position along fault is made to show a composite graph of scarp height change along fault strike. The graph shows that the faulted terraces decrease in vertical displacement to the SW. Also there is a decrease in the triangular facet vertical distances towards the SW. This indicates a complicated fault displacement history. The displaced terraces show the latest fault activity probably indicating several earthquake episodes. The triangular facets probably indicate fault activity over a longer period of time. There morphology is similar to the triangular facets along the front fault. Their position and spacing may be related to temporal slip drainage migration over time coupled with temporal earthquake activity.

122

a) NW SE 60 - Profile 9 Triangular Facet 2 Change in scarp slope angle E 40 - c 0

0 0 50 100 150 Horizontal distance (m) Maximum scarp-slope angle = 0 - - - Secondary scarp-slope angle = 0' Upper slope angle = U0 Lower slope angle = Le Scarp height = H Vertical surface displacement = V

NW SE 40 - Profile 10 Footwall Terrace 2

•• • E •0 ..••• . 0 " o ›,I 20 - 0 w

50 100 150 Horizontal distance (m)

b) SW 60 0 o Fault scarp 50 11 TF1

9 TF2 0

5 TF3 0 0 12 FT1 10 FT2 FT3 10 FT4 6 2 1 8 7 4003 0 0 0 r 0

0 100 200 300 400 500 600 700 800 900 1000 Horizontal distance (m)

Figure 4.39. EDM levelled traverses measured normal to the fault scarp, b) composite graph of vertical scarp displacements along strike calculated from transverse scarp

profiles.

123 Table 4.2. Summary of Grover relay transverse fault scarp measurements. Profile Vertical Scarp height Maximum Upper Lower Location surface (m) slope angle surface slope surface slope displacement (degrees) (degrees) (degrees) (m) 12: FT1 10 12.5 35 10.5 5 11: TF1 52 56 23 1 3 10: FT2 14 16.5 28 6 3.5 9: TF2 26.5 29.5 32.5 6 2.5 11 23 8: FT3 2.25 3.5 13.5 6 5 7: FT3 1 2 10 5 6 6: FT3 3 4 16 3 5.5 5: TF1 15 16 20.5 4 7.5 4: FT4 2.75 4.62 14 5 5.5 3: FT4 2.5 2.75 19 4.5 5 2: FT4 3 3.75 22 4.5 5.5 1: FT4 2.5 3.25 16 3.5 4 13 4 6 28 8 15 14 lower 4 6 12 7 5 14 upper 7 14 19 11 7

4.8.4 Transverse Stream Profiles The longitudinal stream profile of STR2 has been measured using the EDM levelling method (Fig. 4.40) and is located to the west of triangular facet one (TF1). This graph shows a change in shape in the middle of the profile, which steepens and flattens out to the previous profile slope. The approximate position of the fault scarp is marked. This change in stream profile slope is interpreted as a change in erosion resulting from a change in base level of the faulted stream course. Smoothing of the faulted stream channel has produced this intermediate stream profile shape. There has not been enough time to fully erode out the affects of faulting. The vertical difference of the stream at the fault scarp is 2m. STR3 has a linear profile across the fault scarp. The stream was able to erode out the small displacement as the adjacent scarp record scarp heights of 2-3 metres.

124 Longitudinal stream profile STR 2

C 0 c Fault scarp trace Knickpoint migration

co t (m)

n ion c t e lev

E NChange in slope angle [ Profile 0 0 20 40 60 80 100 Horizontal distance (m)

Longitudinal stream profile STR 3

12 - 10 E Linear shape 6 iu 4 — Profile 2 0 0 20 40 60 80 100 Horizontal distance (m)

Figure 4.40. Longitudinal stream profiles, measured using the EDM, of two streams cutting across the fault scarp, locations shown in Figure 4.35(a).

4.8.5 Quaternary Grover Relay Fault Evolution The observations of topography and drainage related to the Grover relay fault indicate that this fault has similar geomorphology to the Holocene ruptures mapped by Piety et al. (1992) and Warren (1992). A composite scarp displacement graph generated by plotting the data above with published scarp data (Piety et al., 1992, and Warren, 1992) shows similar scarp displacements. However, to the east of the relay fault the scarp heights are siginificantly larger than at Willow Creek, Phillips Creek and Swift Creek. This indicates that displacement on the relay fault is greater resulting from accumulated displacement on the rear and front faults. This data clearly indicates that the Grover relay has been in a state of hard linkage during the Holocene. The larger triangular facets indicate earlier deformation on the fault possibly indicating an age of an order of magnitude greater that the Holocene scarps, approximately 150k years. TF1, TF2 and TF3 scarp heights indicate that the relay fault has been in hard linkage possible at an early stage in the fault segment interaction history.

125 60 0 scarp height 0

t (m) 50 -

40 - lacemen disp l 30 - 0 ica t

ht ver 20 - ig 0 0

he 0 10- 0 0 arp 0 0 Sc 0 0 0 0 0 0 o

0 1\ t b 1' tel A 6 '5 t), 3ti° \ N1-- Nk- * 4 dle oo e' N. q* • d .4' t cf; e\'• Scarp location

Figure 4.41. Composite graph of measured scarp heights with published Holocene rupture scarps heights (after Piety et al., 1992; Warren 1992).

4.9 Front Segment Tip Topography The front segment tip fold or intrabasin high is located between the two Star Valley sub- basins and is the northern extension of the front fault segment. In plan view the fold tip topography extends from Willow Creek to 2 km north of Thayne village. The topography gradually decreases with approximately three hundred metres drop over 10 km. Two large fluvial stream channels, Willow Creek and Strawberry Creek cut across the structure and both have large upstream rear segments catchments (Fig. 4.42). A dominant ridge is located on the east flank of the fold tip structure. This is clearly marked in the aspect and slope images (Fig. 4.43).

126 505000

41.

0 N 0 ff

0C> O 0 Lc)

N

50500C

Figure 4.42. Plan view map of drainage overlain on sun-shaded DEM showing fault tip topography. Blue arrows indicate drainage diversion.

127

4 755000 4750000 1745000

0

00055/P 00000t 0005PIt

4755000 4750000 4745000

8 0

000551 t' 00009Lt

Figure 4.43. Aspect and slope maps generated using the 10m DEM.

128

495000 504000

Asymmetric topography (prof.3) 1,4 — 1950 ridge C 0 Salt river — 3 w — 2 Gentle convex topgraphy (prof 2) — 1 1750 1 2 3 4 5 6 7

Line distance (km)

2200 ridge —Prof 41

.2• 2000 Salt river w

1800 0 1 2 3 4 5 6 7

Line distance (km)

Figure 4.44. Fault tip transverse topographic profiles showing the change in topographic morphology.

129 Five SW-NE profiles orientated normal to the segment tip fold structure axial topography show the decrease in topography to the northwest (Fig. 4.44). The profiles also show a change in asymmetry to the northwest. Profile 3 clearly shows the position of the prominent scarp, shown in Fig. 4.42, that parallels the fold axial trace. Profiles 1 and 2 show a gentle fold topography. Profile 3 shows a transverse fault tip terrain asymmetry with an eastward dipping surface and a horizontal westward surface. The profiles also indicate a decrease in topographic relief, i.e. difference in elevation from basin level to fold tip maximum height.

There is a definite change in transverse topography from Willow Creek to the terrain north of Strawberry Creek. The change in asymmetry to the NW may indicate the change in intrabasinal topography, generated from the interaction of the front and rear segments, to the topography generated by the front segment extension forming a tip propagation fold above a NW laterally propagating fault segment.

4.9.1 Willow Creek Drainage

Drainage patterns in Willow Creek indicate a localised diversion north of the front fault (Fig. 4.45a). In plan view the stream pattern changes from a linear to meandering form with a change in stream orientation north of front fault. The yellow dots indicate a detailed longitudinal stream profile surveyed along the south side of the stream using a differential GPS (see Chapter 3 for survey method). Point elevation measurements were recorded along the stream. The longitudinal stream profile shows inflections along the surveyed channel length. Two inflections in the profile correspond to changes in the plan view stream pattern. The inflection marked on the graph coincides with the northward projection of the front fault trace. A west east topographic survey along Willow Creek indicates a minor scarp with a scarp height of 3.2 m. This indicates that there is a continuation of the front fault with a degraded minor fault scarp extending northwards across Willow Creek. The change in drainage is in response to faulting diverting the drainage in the direction of propagation.

130

50400C 506000

Northward migration o stream channel

504000 506000 b) 35 Longitudinal stream profile 30so 1- 25 W E 20 Change in plan view 15 V3 stream course to 5 Inflection in stream profile — Survey Point -5

0 0.5 1 1.5 2 2.5 Survey Line Distance (km)

E Topographic survey 1915 - Scarp height 3.2m E Minor fault scarp 0 in Willow Creek

>0 1905 •- • • * • Elevation

1895 0 500 1000 Line distance (m)

Figure 4.45. a) Plan view orthophotograph map of Willow Creek showing change in drainage pattern to the north of the front fault and locations of differential GPS survey, b) longitudinal stream and topographic DGPS survey. 131 4.9.2 Willow Creek Terraces Terrain features that have similar morphology to terraces have been identified in the fold tip topography north-east of Willow Creek (Fig. 4.46) Two terraces have been identified with aspect and slope measurements showing that these surfaces dip 5-15° to the NW. These surfaces are interpreted as deformed and rotated stream terraces that have become preserved in the growing front tip propagation fold.

Front Segment Intrabasin High

Willow Creek

Figure 4.46. Perspective view of the terrace features on the east side of the intrabasin high/fault tip fold structure.

4.9.3 Front Fault Tip Fold Development Modern day drainage and topography at the tip of the fold tip structure does not show incision by drainage flowing west from the rear fault range. The distance from the range front is 4-5 km and it is interpreted that only streams with relatively large upstream catchments flow this far west into the basin. The positions of the terrace-like structures north of Willow Creek are 2-2.5 km from the range. It is interpreted that the early fault tip fold topography that formed as the blind front fault tip propagated northwards was modified by drainage from the rear range front including Dry Canyon which has a 132 medium size upstream catchment but is not as large as Willow Creek or Strawberry Creek.

Drainage patterns north of Willow Creek before the front fault segment propagated northwards are interpreted flowing westwards to the Salt River. Drainage from Willow Creek, Dry Canyon and from the rear range front flowed from east to west (Fig. 4.47a). Deposits of Salt Lake / Long Spring Formation formed to the north of the front fault. Propagation of the front segment northwards formed proto fault tip topography deforming the Sallt Lake / Long Spring Formation conglomerates. Assuming constant erosion the course of Dry Canyon drainage and range front minor were diverted and eroded the proto topography to the NE of Willow Creek. Continued incision into the topography resulted in the new stream course migration into the topography. Continued displacement and propagation of the front tip resulted in an increase in topography. The balance between tectonic uplift and incision resulted in the defeat of the established streams linked to Dry Canyon and the range front, forming an abandoned fan surface and associated fluvial terrace morphology. An increase in topography caused a migration of the stream incision eastwards as it responded to the increase in structure size (Fig. 4.47b). This indicates a relative reverse stratigraphy with the older terrace-like surface positioned at a higher elevation than the later terrace surface.

133

1: Pre Front Tip Fold Drainage Strawberry Creek

Dry Canyon

Willow Creek

Grover Relay

2: Drainage related to proto fold Strawberry Creek topography

co 0)

'3o

Willow Creek

Grover Relay

Figure 4.47. a) Diagram showing early drainage development related to the monocline evolution. Yellow tone indicates deformed and uplifted Long Spring Formation conglomerates. 134 3: Continued drainage Strawberry Creek diversion

co

Nco

Dry Canyon

Preserved: terrace

Willow Creek

1 O Grover Relay co co 3 cD

4: Recent drainage

Strawberry Creek

Dry Canyon

'Preserved\ I , terraces I I Willow Creek

O

c.n Grover Relay CD cfl

CD

000

Figure 4.47. b) Diagram showing formation of terraces in monocline topography.

135 4.10 Results Summary Landscape mapping in Star Valley and the Grover relay study area have indicated key tectonic geomorphology features. Analysis of the front footwall catchments involving spatial hydrological DEM calculations, longitudinal catchment swath profiles and trunk stream longitudinal profiles has shown changes in fault terrain morphology south from the tip at Willow Creek to Swift Creek near Afton. This coupled with observations of the unconformable onlapping Long Spring Formation conglomerates on palaeo- topography of the front footwall indicates catchment capture of Grover relay stream and rear segment footwall catchment by headward migration of the front segment catchment.

The geomorphology in the Grover relay shows a response streams to tilting of the relay as it developed over time. Asymmetric cross channel morphology indicates northeastward incision into the hangingwall on the rear segment. A previously unmapped relay fault identified in the relay indicates that the front and rear segments are in hard linkage.

Analysis of terrain of the front segment fold tip has shown deformation of Long Spring Formation conglomerates. The topography changes from a transverse asymmetric shape from Willow Creek northwards to a more symmetrical shape at the tip.

4.11 Discussion: Star Valley Landscape Evolution The interpretation of the geomorphology and geology observations made in Star Valley and the Grover relay study area show that fault array has undergone several stages of drainage and terrain change in response to fault development since the Upper Pliocene / Lower Pleistocene. A model of coupled drainage and fault evolution is proposed.

1.Initial overlap geometry Stage 1 shows the final fault geometry reached by the fault segments following initial rapid lateral length propagation with little footwall displacement. Soft linkage occurs due to interaction of the overlapped fault segments forming a relay ramp. No antecedent drainage has been focussed into the relay. Minor drainages develop upstream of the relay ramp due to incision into proto-topography generated by the interaction of the two

136 segments. The catchments at the centre of the segments have already started to migrate past the maximum footwall topography.

2.Increase in fault displacement Significant fault topography develops. Stage 2 continues the fault development with no change in the segment length but an increase in footwall displacement. Topography development now occurs in three places. The footwalls have increased differential footwall topography, the relay ramp topography increases and minor intrabasinal topography develops at the north tip of the front fault in response to differential hangingwall and footwall displacement on both fault segments. Footwall catchments develop on both segments. 1-leadward catchment migration is greatest at the fault segment centre. More of the catchments either side of the centre catchment are able to erode pass the maximum topography due to increased differential footwall displacement resulting in a greater change in baselevel change at the range front. There is a gradation in the headward catchment migration with a decrease towards the segment tips. Small catchments develop upstream of the relay ramp.

3. Long Spring Formation deposition Long Spring Formation (LSF) is deposited in a series of early syn-rift fans fed by catchments along the range front and from erosion to the west (Carribou Range). In the relay localised debris flows fill in localised topography and form a continuous layer covering the filled in areas. These sedimentary deposits onlap against the early formed back rotated front footwall.

4. Early Catchment Capture Differential fault footwall displacement continues with associated catchment development. The catchment to the north of the centre footwall catchment with less headward migration starts to interfere and interact with the relay drainages. This catchment has migrated east enough to capture local relay drainages resulting in a bend in drainages in the area of capture.

5.Catchment capture at Phillips Creek Stage 5 illustrates the development of the front capture hypothesis. The catchment north of the first capture has migrated east and captured more of the relay drainages and the fault tip catchment of the rear segment. The capture of the rear catchment cuts across

137 the old fluvial stream channel in the relay forming a wind gap. This capture event results in a larger capture catchment as more relay drainage and rear catchments have been incorporated. This results in an increase in capture catchment development towards the front segment tip. Continuation of stage 4 results in the hard linkage and breaching of the relay ramp.

6.Front fault tip propagation Front fault tip propagation deforms the Long Spring Formation conglomerates to the north of the front fault. The blind normal fault deforms the overlying conglomerates forming a fault tip monocline.

7.Latest Holocene earthquakes Continuing fault activity during the Holocene forms a series of fault scarps along outlets at the range fronts on the front fault and along the southern portion of the rear near Willow Creek. Hard linkage has occurred in the Holocene and possibly before, with a relay fault cutting across the relay ramp.

8.Transition of headward migration The difference in drainage divide and max topography along fault strike indicates a transition zone of headward migration and catchment capture. This final part of the model cartoon summarises the current plan view landscape in Star Valley in recent times.

138 1: Intial Fault Geometry

•

2: Fault Displacement Proto relay topography divide Catchment development in rely topography

3: LSF Deposition Debris flow deposits ponded against outboard fault topography

Long Spring Formation fan deposition in hangingwall

4: Early Catchment Capture Catchment capture

LSF basin fill extent

\, Drainage Normal fault segment

Footwall Footwall maximum topgraphy Catchment

Figure 4.48. a) Conceptual sketch model of coupled catchment and drainage development in response to fault evolution in Star Valley

139 5:Phillips Creek Catchment Capture Rotated LSF in relay Catchment capture

6: Outboard Fault Tip Monocline Tip propagation

Quaternary sediments

7: Latest Holocene Tip propgation folding LSF fault ruptures

Relay breach fault

• - ---- _ - _ _ ----- 8:Catchment Key Migration Transition — Latest Holocene ruptures Topographic divide transition

- - _ _ _• ------• • _

Figure 4.48. b), Continuation of conceptual sketch model.

140 4.12 Implications for Normal Fault Growth The coupled erosion and fault evolution of the end two end member modelling scenarios of fault development are considered in Star Valley (Cowie et al., 2000 and Schlishe et al., 1996). The topography surrounding en echelon fault segments indicates the linkage history of the faults. Early fault linkage involves the coalescence and breaching of relay zones between faults with no significant development of topography at the overlap. Late linkage of fault segments developing over time allowing differential footwall and hanginwall displacements in the overlap zone, form intrabasinal highs near to the relay zone (Cowie et al., 2000, and Schlishe et al., 1996). The identified topography north of the front fault indicates an intrabasinal high related to the Grover relay implying that the faults have not linked at an early phase (Cowie et al., 2000). The identification of a new fault scarp in the relay area together with the measured late Pleistocene to early Holocene fault scarps on the front fault and rear fault at Willow Creek indicate a transition between soft linkage to hard linkage or breaching of the relay. Two scenarios of fault development are based on fault displacement and length evolution over time based on simple fault geometries (Cowie et al., 2000, Schlishe et al., 1996 and Densmore et al., 2003).

1: Relatively slow fault growth displacement and lateral propagation. Fault segments gradually propagate towards each other with incremental displacement.

2: Relatively fast lateral fault propagation and displacement. Fault segments grow and reach final overlap geometry with relatively little displacement. Displacement then occurs with no change in length (Schlishe et al., 1996).

Based on these two end member fault growth models, a conceptual model of geomorphic development in Star Valley is considered. It is unlikely that the first case occurred. The gradual propagation of the fault segments towards each other would allow drainage and catchments to respond to the increasing topography generated by footwall uplift. Drainage would be focused into the relay area. Footwall catchment development would change as the fault propagated. There would be relatively smaller footwall catchments at the fault tips with larger catchment development at the fault segment centre. This differential footwall displacement would indicate that catchments would be able to erode pass the maximum topography at the centre of the segment.

141 The second case is more likely to occur based on the observed geomorphology. If the fault segments rapidly grew with little displacement there would be no focusing of drainage in the relay. Footwall uplift development would be different with more accumulated displacement distributed along the segment not just at the centre. As a result the hangingwall and footwall would have more exaggerated footwall topographic profiles with more of the range front having larger changes in baselevel. This increase would result in more of catchments either side of the segment centre to erode further upstream. The relatively high displacement rate on the fault footwall facilitates the headward migration of the front catchments into the relay. This change in high displacement at the range front increases the potential for landsliding and hillslope erosion at the base of the catchment. Landsliding is the dominant process for allowing active headward migration of catchments (Allen et al., 2000). The identification of the new Grover relay fault has implications on the style of linkage of the front and rear segments. The plan view fault pattern in the relay between the front fault and the rear fault shows a stepped pattern suggesting a sense of shear in the relay area. The measurement of the relay fault scarp heights indicates a displacement gradient decreasing to the south west. This suggests that the rear fault is propagating towards the front implying breaching in the upper part of the ramp. The en echelon step-sense i.e. right step, the position of the relay fault, and the dip of the relay ramp show similar plan view geometries to numerical modelling results on oblique fault linkage of normal faults (Crider, 2001). The plan view map suggests that the Grover relay is a right stepping left oblique fault. However, the hard linkage / breach of the relay is interpreted as a transitional stage where the rear has not fully linked to the front.

142 4.13 Conclusion In conclusion, the Grover relay is an excellent site for observing the coupled erosional and fault evolution in and developing continental fault normal fault array. Analysis and interpretation of the digital data and field observations has described the varied landscape morphology in Star Valley. The preserved terrain features have allowed the spatial and temporal comparison of terrain feature along fault strike which indicates the relative style of erosional response to normal fault development.

The fault evolution in Star Valley has produced the following unique erosional / terrain feature processes :

• Catchment capture of rear segment and relay catchments by headward erosion of front segment catchments. This has diverted drainage away from the relay ramp. • Hard linkage of the relay ramp by a new to science fault which shows similar scarps displacement morphology to Holocene scarps mapped along outlets on the front segment. • Front segment tip fold has a change in terrain morphology towards the north indicating that the front fault has possible grown by a new segment developing to the north of Willow Creek.

Catchment capture implies fault growth where the segments initially grow fast in length but have little displacement i.e. under displaced. The fault segments then grow by a displacement gradient along the fault. Increased topography at the range front and associated baseline fall drives headward erosion into the footwall topography. A point is reached where the front catchment erodes into the relay and captures existing drainage.

The morphology of the relay fault indicates that the fault has latest Holocene scarps. Larger scarp displacements on three ridges indicate that the fault has been active over a longer period of time with the front and rear segment being in hard linkage possibly at an early stage in development.

The identification of key geomorphic features indicates the style and temporal evolution of the north and south Star Valley Faults over a longer period than previous palaeo- seismic studies.

143 Chapter 5: Tectonic Geomorphology of the Turpan Blind Thrust Fold Belt, Turpan Basin, Xinjiang, NW China.

5.1 Research Rationale This chapter describes the tectonic geomorphology of a continental en echelon blind thrust anticline fold belt in the Turpan Basin, Xinjiang, NW China. Key tectonic geomorphic features and their interpreted evolution have implications for relative blind thrust evolution in the basin. The results are compared to numerical models of blind thrust propagation (Champel et al, 2002; van der Beek et al., 2002; Simpson, 2004) and fold growth in foreland basins, which suggest that the surface morphology indicates fault geometry at depth.

The Turpan Basin is in an active tectonic regime along the north and south of the eastern Tien Shan (Chinese Tien Shan). The blind thrust folds are overlapped but unlike other folds in the region (Molnar et al., 1994; Burbank et al., 1999; Bihong et al., 2003), the individual thrust fold segments are separate and have not fully coalesced. This together with excellent preserved geomorphic features makes the Turpan Basin an ideal place to study drainage and erosional processes interacting with neo-tectonics. To compare the Turpan Basin foreland folds to numerical tectonic geomorphology models of blind thrust propagation and the implications for blind thrust development this research focuses on the following aims:

• To reconstruct the tectonic geomorphology of the Flaming Mountain Anticline in the Turpan Basin. • To develop a model of Flaming Mountain blind thrust anticline lateral propagation at the western tip.

Like chapter 4, this research encompasses various tools of investigation involving interpretation of multi — data, i.e. remotely sensed satellite imagery, DEMs, and fieldwork observations, to map the geomorphology and geology of the Turpan Basin.

144 The objectives of the research are: • To map the regional geomorphology using SRTM DEM. • To generate ASTER DEMs for fold tip regions to improve DEM spatial resolution. • Geomorphology mapping of drainages and catchments, in the Flaming Mountain. • To map the relationships of Quaternary sedimentation along terraces in the Flaming Mountain.

This chapter outlines the results of geomorphological mapping using remotely sensed imagery, DEMs and fieldwork, and the development of a blind thrust fold model of lateral propagation in the basin during the Quaternary.

The chapter is divided into several sections. The first part (5.2) describes the geological setting and outlines the regional geomorphology. The following sections present the research results with the focus on the investigation of the western tip. An evolution model is established in section 5.6 and is compared to published numerical models of landscape development with implications for Turpan foreland thrust anticline growth evolution.

5.1.1 Location

The landscape of Central Asia is separated into a series of mountain ranges and intermontane basins formed as a result of compression in response to the India-Eurasian plate orogeny (Tapponier et al., 1986). The study area is situated in the Turpan Basin, Xinjiang Ugyur Autonomous region of NW China. This basin is one of several intermontane basins in central Asia and is located to the northeast of the Tarim Basin, the largest in the region (Nishidai and Berry, 1991; Allen et al., 1994; Bihong et al., 2003). The Turpan Basin is located to the south of the Bogda Shan mountain range which is the eastern extension of the Tien Shan (Fig. 5.1).

This region is an excellent natural laboratory for analysing terrain evolution in an active area of mountain building and intermontane basin development. Cenozoic deformation in response to the Himalayan Orogeny (Allen et al., 1993; Shao et al., 1999) has resulted in the formation of foreland basin folds above blind thrusts that have modified the drainage and Bajada fan development during the Quaternary.

145 •

g8r. E8$ EN' 61° Junngar Basin Urumqi 4 -ter •- Z • •' 41417,vs •;P.e ; • - 4-0.%""(- I •; • • •

fp f-4 in

'- 0 km 100

Figure 5.1. Sun-shaded topography image showing location of Turpan Basin study area.

146 5.1.2 Previous Work

Previous work in the Turpan Basin has focused on understanding the structural geology and stratigraphy of the basin mainly for oil exploration. Little published literature has addressed the tectonic geomorphologic evolution of the Turpan Blind thrust fold belt. Geological investigation of the Turpan basin and surrounding regions including the Bogda Shan Mountain has concentrated on the development of the tectonic and stratigraphic evolution of early Palaeozoic to the Quaternary units (Allen et al., 1993; Shao et al., 2001). Geological mapping in the basin has recently undergone a revision. Mapping at 1:50000 scale (Deng et al, 2000) was carried out to establish the tectonic evolution of a series of fold structures in the Xinjiang region. This work details the structural and sedimentary evolution of these fold structures but does not link the two into a tectonic geomorphology evolution.

Remote Sensing investigations have concentrated on basin analysis focussing on petroleum exploration (Nisihdi et al, 1991). Davis et al (2000) show that interpretation of ASTER derived DEMs show changes in erosional processes from high glacial terrains to mountain front regions in the Bogda Shan. This work demonstrates that ASTER satellite imagery is a powerful tool for mapping tectonic geomorphology especially in remote regions of the world. Remote Sensing of fault related folding in the Turpan and adjacent basins has indicated Quaternary deformation (Bihong et al., 2003).

147 5.2 Regional Background The Turpan Basin forms a dynamic neotectonic region where the interplay between erosional processes and tectonic deformation is preserved in the geology and geomorphology within zones of thin skinned (foreland basin blind thrust anticlines) and thick skinned tectonics (block uplift of the Bogda Shan mountains). The geology and stratigraphy of the Turpan is summarised below with an introduction to the geomorphology of the region.

5.2.1 Geological Setting The basin geology consists of a range of foreland basin fill sediments. The oldest sediments are Permian with a stratigraphic sequence consisting of Triassic, Jurassic, Cretaceous and Tertiary. These rocks are exposed along thrust faults and folds striking west-east at the Bodga Shan range front (Fig. 5.2). Cenozoic compressional deformation has resulted in the development of intra-basinal thinned skinned thrust structures. These foreland fold thrusts form west-east elongate arcuate structures and linkage of the individual segments shows a range of overlap and transfers zones. The structural geometry in the northern regional slope of the Turpan Basin is a series of west-east blind thrust anticlines, verging south with north dipping hinterland reverse faults (Allen et al., 1994). A cross section (Fig. 5.2) of the Turpan basin based on a seismic line profile (Nishidai and Berry, 1991) indicates the transition from thick to thin skinned structures towards the centre of the basin.

148 750000 750000 front Section lin Fig 4.2b 700000 700000 Flaming Alyountain Eocene Blind Thrust -154m Aiding Lake Permian-Cretaceous alo range Permian-Cretaceous alo range 650000 650000 600000 600000 Turpan Foreland 550000 550000 500000 500000 O

Figure5 .2.a) Summary of Turpan Basingeol ogy overlainon a Landsat / ASTER mosaic image( modifiedaft erAll enet al .,1994 ). Flaming Mountain Bogda Shan Range front Bogda Shan N Aydingkol Lake Thrust at 5 (km) surface S1 MSL 2 4 V1/4 6 Neogene: undifferentiated Jurassic-Cretaceous-Eocene: Undifferentiated 0 km 10 Permian: Undifferentiated 'INk Faulting

Figure 5.2. b) North — south orientated cross — section of the Turpan Basin based on seismic section (modified after Nishidai and Berry, 1991).

The Flaming Mountain is the largest of the west-east orientated blind thrust anticlines and itself is formed by the coalescence of a several smaller anticlines. The Flaming Mountain blind thrust is approximately west — east with a sigmoidal or stepped change in orientation to a more northwest-southeast trend in the central part of the structure. The plan view trace of the fold topography along the southern limb is 140 km. The plan view trace of the mapped thrust at surface, mainly along the southern range front in the centre of the structure, is 80 km. The structure of the Flaming Mountain changes along strike (Fig. 5.3). The differential change in thrust geometry along strike is shown in Figure 5.4, where schematic cross sections are superimposed onto perspective views of the terrain generated by draping the satellite imagery over the DEM. Near to the fold centre the thrust outcrops along the southern range front and cuts Jurassic, Cretaceous and Eocene rocks on the southern fold limb. Here the thrust anticline asymmetry is very pronounced, commonly with the removal of the southern limb by faulting. The thrust does not outcrop at the range front in the westernmost twenty kilometres where the deformed rock types young towards the tips especially in the westernmost twenty kilometres.

150 4 770000 4740000

8 0 0 rn

0 00 OLL

Figure 5.3. Geological map of the Flaming Mountain (modified after Deng et al., 2000).

151 a) z

0 km 5 L I

b)

Thrust outcrops at surface

Figure 5.4. Schematic perspective cross sections across the Flaming Mountain showing the structural geometry at the a) fold tip and b) the fold centre (see Figure 5.3 for location of cross sections).

152 The stratigraphy of the Flaming Mountain is sub-divided into three stages the Mesozoic, Tertiary and Quaternary deposition (see generalised stratigraphy column in Figure 5.3). The oldest Mesozoic rocks outcrop in the centre of the Flaming Mountain fold with Jurassic limestones and sandstone outcrop in the fold core. There is then a sequence of Cretaceous sandstones, limestones and marls. The Tertiary rocks indicate a phase of continued basin filling with a change in palaeo-environment to a more terrestrial setting. The Eocene rocks comprise mainly red marls and sandstones, commonly interbedded with gypsum layers. The lower Miocene (N1) is similar with metre bedded red sandstones. The upper Miocene (N2) rocks show a change in sedimentation with sandstones commonly interbedded with decimetre-thick conglomerates. The boundary between NI and N2 is gradational. The Quaternary deposits are located in the western twenty kilometres of the Flaming Mountain and locally along the structure at the north and southern range fronts and are described in more detail in section 5.3.

5.2.2 Geomorphology Setting

The basin is dominated by zones of erosion, deposition and neotectonic topography. To the south of the Bogda Shan, large bajada fans form at the range front. Aydingkol Lake is located in the centre of the basin at an elevation of —154 m below sea level. West - east blind thrust fold topography bisects and interferes with the bajada fan morphology. This topography is geographically located between the range front and the centre of the basin.

153 5.3 Flaming Mountain Geomorphology The geomorphology of the Turpan basin is described using a combination of the SRTM DEM and ASTER image interpretation. Spatial analysis of the SRTM DEM has involved the pixel neighbourhood calculation of aspect and slope, which is shown in Figure 5.6 (a) and (b). These maps are used to define the morphology of these foreland basin regions identified in Figure 4.5. These maps, together with topographic profiles and drainage, will be used to describe the geomorphology.

Large bajada fans have formed to the south of the Bogda Shan range. In plan view (Fig. 5.5) the size of these fans range from 5-30 km in width to 20-30 km in length. The three dimensional geometry is describe using aspect and slope maps (Fig. 5.7) and west - east swath profiles measured using the SRTM DEM. Neighbourhood analysis of the DEM involving aspect calculations shows the overall plan view form of the fans indicating the facing direction of the proximal and distal parts of the fan. The fans form large lobate convex up deposits. Neighbourhood slope calculations using the DEM show the regional dip of the basin slope and the change in dip, in plan view, of the proximal to distal parts of the fans which range from 5° adjacent to the Bogda range front to less than 1° to the south at the distal end regions of the fan. West-east profiles show the cross sectional morphology and demonstrate that the regional fan slope has varying topography (Fig. 5.7).

Towards the centre of the Flaming Mountain large Bajada fans on the regional slope north of the structure change in morphology towards the northern limb. Aspect and slope indicate the ponding of sediment against the fold topography along the northern limb where slope angles change from 2°-5° to near horizontal on the slope north of the fold topography in the central parts of the fold (Fig. 5.6b).

154 .l i''111111V1

8

0 0 0 0 !, L

Figure 5.5. Sun-shaded pseudocolour DEM image indicating Turpan foreland basin topography.

155 Fi S gure

IN 73000 750000 r I 5 I W -, .6 ,

. AlaC

a) ,.• 1 if re f.( :/* '41 A f • , . . e. . s pect i m a ge generat ed usi n g s pati

• e al

nei e frotiffahs „...‘,„ et ghb Aspect of SRTM DEM 1 Flat (-1)

ourh =North (0-22.5,337.5-360) I __I Northeast (22.5-67.5) 1 East (67.5-112.5) ood _I Southeast (112.5-157.5) South (157.5-202.5) I r. • anal ___411 Southwest (202.5-247.5) MI West (247.5-292.5)

ysi NM Northwest (292.5-337.5) • MI No Data f

s 0 km 10 ...C•-•••" Z" 4441 of I-,

700000 th 750000 e l ell aq va

S 700000 ti 750000 WIN

O cr cin 'WH

cn cn

co

sv CD Cfg CD 0 CD 1-1 to sv CD

Ua

sv r73 DEM slope (degrees) _10-1 1-2 2-3 ar OE 3-4 9 El 5-6.9 MI 7-9.9 O 15-20 0 111 15.20 20 - 30 ME 30+ tz, an No Data

700000 750000 O

Profile 1

E — 1000

.171 a) Fan2 Fanl — Profile 500 0 20 40 60 80 100 Line distance (km)

Profile 2 800 - w E 0

a) Fanl Fan2 Fan cross section — Profile 300 0 20 40 60 80 100 Line distance (km)

Figure 5.7. West — east profiles generated using the SRTM DEM showing cross sectional Bajada fan morphology.

158 The Flaming Mountain displays a dramatic change in aspect along fold axial strike where the fold topographic divide faces north and south (Fig. 5.6a). However, there is a noticeable change in aspect at the western tip where there is a more gradual change in the aspect of the fold topography. There is also a clear change in slope angles normal and parallel to the fold axis. The northern limbs are generally less steep than the southern limbs, which in the central parts are commonly dissected due to erosion on steep slopes. The western tip shows a change in slope along strike with relatively steep southern limbs and a decrease in slope angle from approximately 15° to near horizontal towards the tip. North — south swath profiles (Fig. 5.8) were generated by selecting rectangular polygon regions oriented with the long axis normal to the range front positioned at the water gaps locations i.e. fluvial channels that cut across the fold.

Calculation of the maximum topography and minimum topography (water gap channel) across the mountain range allows an interpretation of the N-S topographic change along fold axial strike. This together with the aspect and slope maps show the asymmetry of the fold topography mimicking the structural geometry described above. West — east swath profiles are generated by digitising an irregular polygon along the thrust hangingwall trace at the southern range front and in the area of change in fold topography along the northern range front (Fig. 5.9). This region is then extracted from the DEM. Three regions have been extracted parallel to the range due to the sigmoidal shape of the Flaming Mountain. The maximum and minimum topography calculations show the fold axial maximum topography and the range front profiles (Fig. 5.9). This indicates the minimum fold hangingwall topography, the footwall topography and the increase in topography towards the centre of the fold. The positions of the water gaps are clearly shown. There is an overall decrease in topography towards the tips of the fold. The difference in water gap incision depth and related maximum topography indicates a decrease towards the tips. This indicates a change in fault displacement along strike.

159 qr „rryr_ 11,1117,1,1”: - -r • 79 ,i414;Oir-

BODSWath 1 Swath '2' ' 600 Swath 3 400 Swath 4 200

400 A 200m Contours 0 km 10

.., 900 ------._ — MaX E 700 ----____. Fold tip topography g 500 —min ------__ 1 300 - — co 100 ir -100 0 5 10 15 20 25 1; Fold western tip Swath line distance (km)

-,- 900 — E 700 - ___ - ____ g 500 - —Max .r. _-\-- >a 300 —Min 2 100 to -100 0 5 10 15 20 25 Swath line distance (km) 2: Water gap 5

1400 -g- — max i 900 -,-----______Asymmetric topography —min 03 > 400

-100 0 5 10 15 20 25 30 35 40 45 3: Water gap 6 Swath line distance (km)

1400 1 Asymmetric topography Steeper slope _ —max c 900 ------______o - . % ,on southern — min 1 400 - / range front >w ------LTi — ___,,_ -100 --- 0 5 10 15 20 25 30 35 40 45 4: Water gap 7 Swath line distance (km)

Figure 5.8. North-south swath profiles generated using the SRTM DEM orientated normal to the Flaming Mountain showing fold topography.

160 t o "r1 po gr a ph

y to a nd *rl r an ge f 0\z, ront 0 pr West East ofil 900 s. WG7 WG1 WDG1 WG3 WG5: WG6 — Max e 0 WG8 WDG WG9 . 0 - 700 4 5 \W G2 \WDG2/ WG Min 0 c 500 -- Max-Min

.11> 300 cn (1, cn • 100 (D -100 0 10 20 30 40 50 60 70 80 90 1001 gi) Swath line distance (km) WG Water Gap WDG Wind Gap *CJ 0

ci•

0

as 5 1=6,

• Drainage in the Turpan basin has been generated by spatial neighbourhood analysis of the SRTM DEM (Fig. 5.10). North — south drainage patterns are common on the north regional slope with drainage related to catchments located on the southern range front of the Bogda Shan Mountains. Changes in drainage pattern are related to the Turpan fold belt. There is diverted drainage around the fold tips at either end of the Flaming Mountain. At the western tip of the Flaming Mountain, drainage is diverted to the west around the plunging fold tip. Reverse drainage resulting from drainage re-organisation on the rotated northern fold backlimb is also observed on the northern limbs of the folds and indentified in Figure 5.10. The southern range front along the central portion of the Flaming Mountain fold is dissected with drainage patterns related to the fold structural morphology. Axial drainage patterns are present that pick out incompetent Triassic through to Eocene rock layers and have outlets along the range front to the south. Drainages that are able to incise through the fold topography form water gaps (as described in chapter 2 and above). Nine water gaps are identified along the Flaming mountain axial strike.

A graph of water gap spacing along strike of the fold (Fig. 5.11), measured using the DEM, shows a 5 km and 15 km water gap spacing along fold strike. A plot of the maximum difference between the maximum and minimum topography N-S swath profile located at the water gaps (Fig. 4.9) shows the decrease in fold topography towards the fold tips. Dominant channels on the southern limb along the central part of the fold may indicate positions of possible wind gaps. The graphs in Figure 5.11 clearly shows that at the western end of the Flaming mountain anticline there is an increase in the number of water gaps indicating that the erosion here is lithologically driven.

Interpretation of the SRTM DEM and ASTER imagery for the north regional slope of the Turpan Basin focusing on the Flaming Mountain has indicated the western end for further study. The western twenty kilometres of the Flaming Mountains has a greater density of water gaps, complex drainage patterns and a decrease in range topography.

162 Fi gure 5 700000 750000 .10 . D rai na ge overl ai n on sun- sh ad ed S RTM DE M

. DRF Drainge from 0 range front Axial Drainage Large upstream LUC catchment A WG Water gap Drainage Axial Drainage

0 km 10

700000 750000

a)

900 800 • •elevation 700 • E 600 • •o 5°0 400 300 • • 200 • 100 0 0 10 20 30 40 50 60 70 80 90 Distance from west fold nose (km) b)

35000 • Possblelftand Gaps 30000 •Water Gap • 25000 NNW Gaps p 20000 0 15000 - • • • 10000 • • • •• 5000 • • • A • ■ 0 • In 0 10 20 30 40 50 60 70 80 90

Distance from west fold nose (km)

Figure 5.11. a) Graph showing relief of water gaps, b) graph of water gap spacing along fold axial strike .

164 5.4 Investigation of the Flaming Mountain Western Tip The western 20 km of the Flaming Mountain is described in further detail with a systematic description of the terrain features using a 30m ASTER derived DEM, ASTER imagery, Corona declassified panchromatic imagery, and field observations.

The geology, including the structural geometry and stratigraphy, is verified and further described using field observations. This is essential as comparisons of the structural geometry are compared to terrain features later. Geomorphic features are described using a combination of field observations and image analysis.

5.4.1 Blind Thrust Fold Geometry

A geological map of the study area is the combination of published maps (Deng et al., 2000) that have been modified by ASTER imagery mapping (Fig. 5.12). Quaternary conglomerates and sandstones ranging from Q1 (Lower Pleistocene) to Q4 (recent) in age dominate the study area. The latest Q4 deposits form the latest water gap channel deposits. Q3 (Holocene) outcrops are located on terraces along the water gaps. Q2 (Upper Pleistocene) and Q1 units are folded with Q2 forming part of the northern limb.

Lineaments have been mapped in the western part of the fold. On the north limb there are several evenly spaced pervasive lineaments that can be traced across water gap 3 and the northern edge of the terrain between water gap 2 and water gap 3 has a linear west-east range front. Similar lineaments are observed between water gap 1 and 2. In the western tip area there are three west-east trending lineaments. The north limb front between water gap 2 and 3 is linear and can be traced for three kilometres.

Along the southern range front the Q1 deposits form a series of curved features that appear overlapped near to the mouth of water gap 2 (marked on Figure 5.12). To the west of water gap 3 there is a fault that cuts across the fan deposits. This verifies similar observations by Deng et al., 2000, and is interpreted as a recent fault. These field observations are used to describe the change in fold geometry and geology towards the east.

165 4766000 4760000

8 CO

8 8 r -1 O Co CNI CNI aaaZz

E

O

00099LP 00009Lb

Figure 5.12. Geological map of the study area interpreted from ASTER and corona imagery (modified after Deng et al., 2000).

166

5.4.1.1 Water Gap 1 Exposure of the Quaternary deposits and the structural geometry in water gap 1 is restricted to the eastern edge of the fluvial channel. Observations of the eastern cliff show the cross sectional geometry of the fold. The terrain and Q1 / Q2 bedding angle orientations indicate a gentle fold with marginally steeper angles towards the south as shown in the N-S photomontage (Fig. 5.13).

5.4.1.2 Water Gap 2 A similar cross sectional observation is made of water gap 2, which is a well developed agricultural valley where outcrop exposure is limited. A photo montage (Fig. 5.14) indicates an increase in the dip of the northern fold limb. Bedding outcrop exposures in the eastern cliff indicate a change in dip towards the south. Again the terrain shape and bedding layers indicate a gentle fold with asymmetry to the south.

5.4.1.3 Water Gap 3 Excellent outcrop exposure and accessibility in water gap 3 allowed a N-S structural traverse. The strike and dips of bedding within the N2, Q1 and Q2 units were recorded along the water gap at regular intervals. This shows the change in cross sectional dip across the fold range and pin-points the plan view fold crest position at 1.2 kms north of the southern range front. Stereographic projection of poles to bedding indicates an a fold verging south and plunging 5° to the west (275°). This plot graphically shows the steeper bedding dips at the southern range front with more gentle 10°-20° dips on the northern limb. The mapping allows the construction of a N-S structural cross section orientated normal to the fold axis (Fig. 5.15) and show fold vergence to the south. Accommodation structures are observed in a deformed zone on the southern limb. Minor faulting is common, displacing sandstone layers within the Miocene N2 units. Minor folding and displacement along fault plane visable in the western cliff edge indicate back thrusting with related fold vergence to the north as shown in Figure 4.16.

167 Figure 5.13. Photomontage of the eastern side of water gap 1 showing apparent dips of gentle asymmetric anticline.

168 Figure 5.14. Photomontage of the eastern side of water gap 2 showing apparent dips of asymmetric anticline. 169 687000 690000

687000 690000

Figure 5.15. a) Structural traverse map showing location of bedding measurements made along a north-south traverse of water gap 3.

170 b) Stereonet

+ n=21 (P) Num total: 21

Equal area projectionlower hemisphere

C)

N Vertical exaggeration x1 S 2 Maximum topographic swath profile

E Minimum topographic swath profile 0

0 0- 171 Thrust at depth

-1 1 2 3 4 5 6 8 Water gap section line distance (km)

Figure 5.15. b) stereonet plot of poles to bedding, and c) cross section of water gap 3 based on imagery and traverse.

171 . • S Facing West

Verging North

•

Figure 5.16. Photograph of back thrusting south of water gap 3.

5.4.1.4 Water Gap 5 Water Gap 5, like water gap 3 has excellent outcrop exposures and forms a natural N-S geologic cross-section. A photomontage (Fig. 5.17) shows the fold crest near to the southern range front. Here the bedding dips within the red sandstones of N1 and N2 age indicate the fold axis with steeper dips to the south. A map of bedding layer dips measured at regular intervals along the channel (Fig. 5.18a) shows the change in dips and the position of the fold axis, which is 2 km north of the southern range front. The stereonet plotting of poles to bedding (Fig. 5.18b) shows the asymmetric fold geometry with steeper southern range dips of 75° to near vertical. The fold plunges 10° to the west (280°). These observations allows the construction of a structural cross section (Fig. 4.18 c) which show the N-S asymmetric fold southern vergence with a blind thrust at depth.

172 Figure 5.17. Photograph mosaic of folded lower Miocene to Upper Miocene rocks and the unconformable relationships of the Q3 and Q4 conglomerates.

173 693000 696000

N- 0 0 0

22/ 8 22/ 1Q// 20/ 101 10 20 ;\ 24\ 38 45

64 0) r- 80 0 75 Inclined

Vertical

Horizontal

693000 696000

Figure 5.18. a) Structural traverse map, b) stereonet plot of poles to bedding, and c) cross section of water gap 5 based on imagery and traverse.

174

b) Stereonet

+ nm21 (P) Num tote 21

Equal area proiection, lower hemisphere__

c) N Vertical exaggeration xl Crest S 2 Maximum topogra • hic swath Hinge profile Minimum topographic swath profil

Thrust at depth

1 2 3 4 5 6 7 8 9 Water gap section line distance (km)

Figure 5.18. b) stereonet plot of poles to bedding, and c) cross section of water gap 5 based on imagery and traverse.

175 Figure 5.19. Photograph of normal faulting in Lower Miocene sandstones.

Normal faulting in the lower Miocene sandstones (Fig. 5.19) indicates accommodation extensional structures within in the fold hinge zone. The faulting does not deform the overlying Q3 conglomerates.

The fold axial swath calculated in Figure 4.9 shows the decrease in topography to the west. By calculating the difference in the maximum and minimum topography allows the construction of a minimum hangingwall height profile. This displacement profile is the result of hangingwall thrust displacement, which has been modified by erosion over time hence minimum displacement. The profile shows the decrease in thrust fold displacement towards the west (Fig. 5.20). The graph is irregular in shape resulting of erosion modifying the curve. From the western tip of the fold, there is a displacement of 300m over 51cm towards the east.

176 600 600 - ..... E4C 40000- /I f 300 - - .I . 200 • T.! 100 - , E"----- Max-Mint 0 , , 0 5 10 15 20 Distance from fold nose (km)

Figure 5.20. West — east profile of maximum — minimum topography difference.

5.4.2 Stratigraphy The Quaternary conglomerates and sandstones are divided up into four units (Deng et al., 2000). Q1 is the oldest unit and is comprised of a thick sequence coarse conglomerates consisting mainly of basic igneous clasts. There are few interbedded sandstones. The Q2 boundary is not well defined and locally is gradational. The main difference between Q1 and Q2 deposits is that the latter has more interbedded sandstone layers further up the sequence. Both Q1 and Q2 have been deformed and are folded. Q1 outcrops on both limbs of the fold with Q2 outcrop on the north limb (Fig. 5.12).

There is an angular unconformity Q3 deposits between Q1 and Q2 and older rocks. A photograph of the angular unconformity between Q1 and Q3 deposits is shown in Figure 4.21. Q3 is similar to Q1 deposits in that there are coarse conglomerates at the base of the sequence. More interbedded sandstones then occur further up the terrace sequence. Q3 deposits outcrop discontinually along water gaps and along fold strike. Q4 deposits are recent fluvial and colluvial sediments that outcrop along the water gaps. These recent deposits are coarse conglomerates located in the current fluvial channel. Colluvial deposits are mainly local fans of eroded sediments from the surrounding fold and channel topography. Many recent Q4 fans are preserved in the main fluvial channel (Fig. 5.22). Localised landsliding occurs where erosion at the channel edges results in block and topple failure.

177 S Facing West

Modern day fluvial channel, Q4

Figure 5.21. Photograph of angular unconformity between Q1 and Q3 deposits in water gap 5.

Figure 5.22. Photograph of Q4 (recent) fans within water gap 5 fluvial valley floor.

178 The sedimentary deposition in the water gaps results from a complex interaction between ephemeral erosion from the Bogda Shan and localised erosion from the side of the fluvial channel and surrounding fold topography. Examples of the interbedded layered deposits are present in all of the water gaps. Water gap 5 has particularly well preserved examples.

A photograph of Q3 / N1 unconformity shows a vertical section through the sedimentary section (Fig. 5.23). There is a bedrock hiatus indicating an erosional event. This horizon is then covered by rounded to well rounded coarse conglomerates mainly basic in composition. These are interpreted as channel fill conglomerates sourced from the Bogda Shan. These deposits removed evidence of local deposition forming a terrace. Overlying these course conglomerates there are a series of local fan deposits. There are layers of interbedded course conglomerates, which are interpreted as interbedded channel fill within localised fan sediments indicating periods of erosion from the Bogda Shan. This can be seen in the modern day valley floor as shown in Figure 5.22.

Figure 5.23. Photograph showing channel fill conglomerates and fan deposits.

179 Preserved in the current water gap fluvial channels are deposits that range in time of formation. Infilling of channels by coarse conglomerates may take hours or a few days to be deposited resulting from ephemeral erosion from the Bogda Shan. Localised fan deposition, normal to the water gap channel, are formed over a larger time span. Some fans are the result of storm wash from the surrounding topography. Some fans show evidence of incision indicating baselevel changes in the fluvial channel. At the southern range front in water gap 5 there are three well exposed Q3 terraces. A levelled traverse, Figure 4.24, measured using an EDM, shows the change in height of these three strath terraces. These terraces indicate erosional time slices during thrust fold displacement.

120 100 Height of T3 80 T2 60 T1 5215 40 20 Water Gap 5 Q4 Channel

-20 • sesiesil -40

0 0.2 0.4 0.6 0.8 1 Profile line distance (km)

Figure 5.24. Levelled traverse orientated normal to water gap showing position of terraces.

From the observations made above and comparing the terrace deposits to the Q4 recent fluvial deposits an interpretation of the fluvial depositional history of the water gaps is proposed (Fig. 5.25). There are two main sources of sediment input into the water gap channel:

• sandstones and mainly course conglomerates from the Bogda Shan • localised erosion from the side of the water gap channel and surrounding topography.

180

a) N

S Localised fans Strath "*-- terrace

Q4 Recent channel

b) Schematic section Mainly localised fan deposits

Interbedded channel and fan deposits

Strath conglomerates Q4 recent channel fill conglomerates

c) Erosion from Bogda Shan Sketch of water gap channel erosion

Fan deposit sourced from channel and surrounding topography

Figure 5.25. Schematic diagram showing water gap depositional history.

181 5.5 Water Gaps and Wind Gaps The geomorphology in the western 20 km of the Flaming Mountain has a variety of drainage patterns, topography and terrain features that change along fold strike. This region differs greatly from the more eroded / incised landscape in the central parts of the fold further to the east.

The water gap channel morphology and adjacent topography varies significantly in the study area. The locations of five water gaps and two wind gaps are shown in Figure 5.26. The wind gaps have similar fluvial plan view geometries to the water gap channel but from stereographic viewing of the stereo ASTER imagery the channels are no longer connected to the upstream Bajada catchment and are preserved in the landscape. These will be discussed later.

Automatic generation of water gap, wind gap and range front watersheds / catchments have been carried out using ArcGIS 9.1 with the spatial analyst extension. Outlet points identified along the south range front were used to extraction upstream catchment areas within the fold topography and the Bajada fan area to the north (Fig. 5.26b). Water gap 2 has the largest watershed area. The watershed area of water gaps 3 to 5 decrease to the west.

Axial swath profiles (Fig. 5.27) generated from the polygon regions outlined in Figure 5.26, demonstrate a dynamic fluvial landscape. The maximum topographic relief decreases to the west. There is a stepped decrease in maximum water gap elevation to the west. This is also mimicked in the minimum footwall profile. Also there is a decrease in wind gap elevations to the west. Fold axial topographic swath profiles (Fig. 5.27) orientated west - east indicate water and wind gap positions, and cross-sectional morphology. They show that water gap channel baselevel decreases to the west. Water gaps 3, 4 and 5 are at the same elevation while to the west, water gap 1 and 2 are at a lower elevations. The wind gaps are at an elevation above water gaps 3, 4 and 5.

182 Figure 5.26. a) ASTER 321 BCET DDS (Liu, 1991; Liu and McM Moore, 1996) image geomorphology map of the western 20 km showing locations of water gaps.

183 00008Lt 00009Lt

00008!V 00009LV

Figure 5.26. b) Map of water gap upstream watershed boundaries automatically generated using the SRTM DEM. The process calculates the area of the upstream watershed based on the outlet of the water gaps at the range front. The watershed calculation area is restricted to the fold topography and the Bajada fan area to the north at the range front of the Bogda Shan mountain.

184 800

700 WNG2 600 Fold Nose WNG1 WG3 WG4 WG5 — 500 WG1 WG2

t- 400

—Max 300 —Min

200 - WG outlets

100 Range front profile Elevation difference 0

0 5 10 15 20 Distance from fold nose (km)

Figure 5.27. West east swath profiles generated using the SRTM DEM.

Drainage in the western end of the Flaming Mountains shows complex patterns. Figure 5.28 is a greyscale Corona satellite image with overlain drainage, water gap channel extents, local drainage divides and wind gaps. The drainage patterns vary along strike from west to east. The drainage is deflected around the fold nose at the west tip. North — south sinuous drainages identified with a yellow ellipse in Figure 5.28 join water gap 1. Further east similar N-S drainage patterns (yellow ellipse) are identified in the uplifted topography to the east and join water gap 2. At the northern end of water gap 2 numerous streams are deflected into the channel. Water gap 2 has the largest upstream catchment. There are local drainage divides along the north limb of the fold related to the upstream water gap catchments.

The relatively steeper slopes of the southern range front show increased erosion and incision (Fig. 5.28). The dissected range front (DRF) shows an increase in based on image textural information. Localised erosion at the southern range front between the water gaps forms small colluvial fans.

The diversity in geomorphology in the study area requires detailed mapping. Digital elevation models and imagery are used to calculate minimum and maximum topographic profiles. Targeted fieldwork was used to verify digital mapping results. The following paragraphs systematically describe the geomorphology of the five water gaps from west to east. 185 overlain drainageandwater channelmorphology. Figure 5.28.GreyscaleCorona imageofthewesterntipFlaming Mountainwith 186

Local Divide 5.5.1 Water Gap 1 Water gap 1 is located to the western tip of the Flaming Mountain (Fig. 5.29) and is the westernmost fluvial stream to incise through the blind thrust topography. The incised valley floor is 5.1 km long and is oriented approximately SW-NE. The valley floor in plan view is not straight (Fig. 5.29a). The upper 2km of the valley from the northern range front has straight plan view morphology. There is a kink in the general orientation of the valley 2 km south along the valley where there is a 90° change in the stream course. At this intersection the valley plan morphology is sinuous. Further south to the range front the valley becomes straight. The width of the fluvial valley floor increases southwards, widening at the intersection, ranging from 50 m in the north to 150 m in the middle and 350m at the southern range front. The topography on either side of the valley shows badland type erosion. The east side of the valley is approximately 100 m high at the position of the fold crest with steep cliffs, often near vertical.

Undercutting by the main fluvial stream has resulted in a steep cliff on the eastern edge where the channel has migrated laterally to the east. There are numerous localised debris fans at the steep side cliff bottoms. The valley is asymmetric with the eastern side having steeper sides with near vertical cliffs (Fig. 5.13) while the western side is less steep. The N-S swath profile (Fig. 5.29b) shows that the maximum topography adjacent to the eastern channel edge is gently curved with a convex up shape profile.

5.5.2 Water Gap 2 Water gap 2 is located 7.5 km to the east of the fold tip. The incised valley is 6.03 km long and is orientated approximately N-S in plan view (Fig. 5.30). There is a slight bow to the plan view valley morphology with a change in inflection in the middle. The water gap again has an asymmetric transverse profile (Fig. 5.30) where the eastern side has sub-vertical cliffs. The terrain is dissected and shows evidence of landsliding especially on eastern side of the fluvial valley. The terrain to the NE of the channel has a scalloped morphology and is well dissected. The western side is generally less steep and slopes towards the fluvial valley as shown in figure 4.30(c). Valley floor widths generally increase south, ranging from 300m in the north to 400m in the middle and 1 km at the southern range front.

187 6 e

I co

hy ion it rap os

og E l p

top C hanne nvex

c C t n co le

t 0 0. E Prese Gen

files o

Pr le

fi 8 CNI hic rap h Pro

t 117)-

og 0 Swa

2 f; Top tu) too IWA•13 W (w) uoulmem b)

c) 0 0 4768000 4766000 4764000

00

00089/1, 0009911, 00099/1, Figure 5.29. Geomorphology of Water Gap 1 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross fluvial valley morphology measured using the DEM.

188 N .S E E

2 0. S 0 a) ET) S 04 7.1) co a) 0 Q. 0 a)

km) km) ( (

nce

ta 10 tance lley dis dis

a ine v line l

ic file file tr Pro Pro mme Asy

0 kr) O files

Pro

le a) fi hic

Pro ap 2 2 r 0 0

h 0 0 0 0

t LO • lf)

og 01 04 0 0 0 0 Z 2 0 0 0 0 0 (w) uoneAai3 UJ N v- O ▪( tu) uogenal3 Top Swa

4766000 4764000 4762000

0 0 <40 000

0

00

Figure 5.30. Geomorphology of Water Gap 2 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross fluvial valley morphology measured using the DEM.

189 5.5.3 Water Gap 3 Water Gap 3 is located 13.2 km to the east of the fold tip. The incised channel is 6.8 km long and is orientated approximately N-S in plan view (Fig. 5.31). The channel bifurcates 3 km from the northern range front forming a "Y" shape in plan view. Towards the south the channel bows in plan view with a convex shape to the west. Transverse profiles are not as asymmetric as water gaps 1 and 2. The profiles show stepped shapes showing previous terrace levels. The valley widths are generally the same along valley length, although slightly decreasing at the southern range front, ranging from 430m at the northern range to 530m at the mid-valley to 290m at the southern range front. The bifurcated northern part of the valley shows complicated landscape morphology. Drainage patterns show four main fluvial valley incisions, two form the existing water gaps, and two orientated N-S and SW-NE are no longer incising and are abandoned. The orientation of the SW-NE stream is similar to the orientation of the upper main fluvial valley in Water gap 1. Dendritic drainage joining the main fluvial channel in Figure 5.31(a) is orientated 90° to the main valley. There is well-developed dissection of the surrounding terrain indicating badland type erosion.

5.5.4 Water Gap 4 Water gap 4 is located 15.9 km from the western fold tip. The incised fluvial valley is 6.8km long and is orientated N-S in plan view (Fig. 5.32a). The valley bifurcates approximately 2 km north of the fold crest and 1 km north of the southern front there is a change in valley floor shape with a westward diversion coincident with the fold crest. Two main channels are observed, one that joins the main fluvial channel and the other is orientated SW-NE in plan view. The transverse valley profiles are mainly symmetric. A series of stepped terraces are identified. The valley widths are generally the same, ranging from 160m at the north to 237m in the middle to 230m at the southern range front. The SW-NE valley to the north is no longer incised by streams from the north, appearing abandoned with recent streams being diverted either side of these channels. There are two tributaries that join this SW-NE channel which also appear abandoned. There is well developed dissection of the surrounding terrain indicating "badland" type erosion. The N-S swath profile shows a more asymmetric terrain profile.

190

3 - file - ro - - - - - _ . - p . _

5 2.

2

hy km) ( ap r tance og 5 1. dis

top line ic tr file Pro mme

1 Asy

s 5 le 0.

fi

---....._ Pro

2 file hic file ro ---- )

Pro ....a.

rap • 0 0 0 0 0

h 0 0 0 0 0

t CO ,0 • 0, CV og 2 (w) uoReAela W CI Top Swa

4768000

0 0 0 0, CO

0 0 0 CO CO CO

00099L4

Figure 5.31. Geomorphology of Water Gap 3 a) greyscale Corona image showing geomorphology in plan view b) swath profile of longitudinal valley profile measured using the DEM and c) transverse profiles showing cross-fluvial valley morphology measured using the DEM.

191

usi geomor w n g sea

th Swath Profile m Asymmetric topography ph 690000 692000 694000 a e DEM Eroded hinge zone ol p 600

n 400 — max o s

g 200 — min y i 0 max-mm ti! 0 1 000 2000 3000 4000 5000 6000 7000 8000 9000 s n a nd pl atp

an c) Nga vi t .I ran e w b

sve To.o.ra shic Profiles

) I 550 -IN rse Valley floor Old valley floor preserved in northern limb sw g 450 ath profil 350 - Stepped terrain — profile

w 250 prof 0.5 1 1.5 2 2.5 es Profile line distance (km) il sh

e DEM erofile 1 owi of 700 W l n on 600 - g gre 0 500 gi Valley floor cro 400 - t yscal udi 300 — — profile

ss-f W 200 nal 0 0.5 1.5 2 2.5 3 e C l uvi Profile line distance (km) v DEM profile 2 all orona i al e vall y 700

pr 8 500 Dissected / eroded fold topography — profile e

ofil ma Valley floor y 300 mor ge

e 100

m 0.5 1 1.5 2 2.5 3 sh ph eas Profile line distance (km) owi ol DEM profile 3 ur o n gy ed g

5.5.5 Water Gap 5 Water gap 5 is located 19.3 km from the western fold tip. The incised fluvial valley is 7 km long and is orientated N-S in plan view (Fig. 5.33). The maximum N-S swath profile (Fig. 5.33b) is bimodal in shape with a decrease in elevation at the fold hinge zone. This is interpreted as erosion of the fold hinge zone (crest and hinge area of the fold). The valley splits into 4 smaller valleys north of the fold axis (Fig. 5.33a profilel). The transverse profiles are generally symmetrical although there is asymmetry towards the southern range front located at the fold crest. Profiles show at least three terrace levels. The valley widths are generally the same ranging from 196m in the northern front, to 216m in the mid-point of the valley to 234m at the southern range front and valley outlet. The surrounding terrain is well-dissected showing "badland" type erosion.

193

geom

- 694000 Swath Profile Asymmetric topography or ns mu ph 500 ai North limb Eroded fold crest 400 max p ol 300 min o 200 sn gy i u 100 max-min G 0 e 21. 1000 2000 om 3000 4000 5000 6000 7000 noon n N° S pl ay or Steeper sloped an ct ph southern limb vi

ol Topographic Profiles wa 0 -

ew b 0 o

gy 8 0 ? 450 N. c 400 Valley Valley ) Valley of

sw 350 W 1, 300 ath — profile w 250 at

e 0.5 1 1.5 profil 2 2.5 r G Profile line distance (km) DEM profile 1 a 0 e p 0 of l

5 0 750 a) o g 550• • • — profile n gi gr t 350 t e udi yscal ▪w 150

nal 0.5 1 1.5 2 2.5 e C Profile line distance (km)

ch 0 0 DEM profile 2 an orona i co r-- nel ? 500 E 4-- Q3 Terrace 400 Terrain step

profil Dissected hinge zone io 300

m Valley floor 200 a — profile ge

e ILI 100

mea 0.5 1 1.5 2 2.5 sh

owi Profile line distance (km) sured DEM profile 3

n 694000 696000 g

5.5.6 Wind Gaps The terrain east of water gap 2 is an area of valley morphology, in plan view, similar to the water gaps (Fig. 5.34). Swath profiles orientated north — south show that the terrain is convex up with a steep dip to the south and a sub horizontal to two-degree dip to the north. This is verified with the aspect and slope maps and observations in the field. In plan view five main drainage patterns are mapped. The general orientation of these drainages is SW-NE. The orientation angles for each stream are 042°, 025°, 010°, 030° and 018° from west to east. The dissection of the terrain is similar to the terrain to the west. Edge enhancement of the Corona imagery shows a zone of well-developed dissection at the southern range showing a badland type erosion. The north part of the terrain is not as dissected and the channel morphology are well defined. The northern range front is linear and plan view drainage is mapped following west to water gap 2 and east to water gap 3. This forms a local drainage divide marked on Figure 4.34. Two fluvial channel morphologies have been identified, using ASTER and Corona imagery, east of water gap 2 during image mapping. Plan view channel morphology and dimensions are mapped and shown in figure 4.34 (a). The first channel is 7.3km long and the plan view shape is bowed towards the south and is slightly convex to the west. The other channel is 6.9km long and is also convex in plan view to the west. West — East profiles measured using the DEM, shown in figure 4.34 (b), indicate cross sectional concave fluvial channel morphology. The profile also shows a step down in channel base elevation to the west.

The two main channels clearly have a different fluvial evolution history to the rest. Plan view mapping of the terrain verified in the field shows areas of recent erosion. At the north end of the terrain there is evidence of recent incision into the terrain (Fig. 4.34). The main fluvial channels that are de-lineated by straight edged sides are no longer connected to the upstream Bajada fan catchments to the north, which is verified with ASTER stereoscopic satellite image viewing. Longitudinal stream profiles were measured using the DEM and verified in the field with a laser ranged profile (Fig. 4.35a). These profiles generally show a convex profile to the fluvial channels streams. The drainage divide in the stream channel occurs 1 km from the northern edge. Further inspection of the profile shows a change from a convex shape flattening out to a straight line. This straighter part of the curve co-insides with the northern extent of the dissected zone. This is interpreted as a knickpoint where recent erosion is cutting back into the terrain. The convex profiles shape and the fluvial channel morphology and the straight 195 longitudinal terraces edges indicate the preservation of fluvial water gap morphology in this terrain. It is interpreted that these two channels were formally water gaps that have been defeated. The channels were disconnected to the main Bajada catchment to the north and have form wind gaps.

w

8

hy ) kin ( rap og tance is d top ic line tr file Pro me m Asy

(.0 0)

2 le fi E 0_ 5 h Pro t oor) 2 Swa Zi W 4768000 4766000 4764000 4762000

00099L8 0004914 000Z9L

Figure 5.34. Geomorphology of Wind Gaps a) greyscale Corona image showing geomorphology in plan view, b) swath profile of longitudinal channel profile measured using the DEM c) transverse profiles showing cross fluvial channel morphology measured using the DEM. 196

ph 675. ot a) o N 400 gra Divide c•D Recent Incision —LR Convex ph 350 4 GPS _:-.-''-.7-----,:------.7_.--_. ..r. Extent • DEMI mont ag 300 .,... / c ,.,..,,,.... Linear

a r- 250 t.:- ......

ge l 0 0 (jc? M 200 ooki 150 n g NE 100 0 1000 2000 3000 4000 5000 600

t Longitudinal profile line distance (km)

o SE Fr

,. b) W ,111' E 8 40 CD 1.1 20 Wind Gap 1 a 10 ...... • § 0 - T3 10 No..* CD 5 -20 ';r CD -30 0 0.5 1.5 2 2.5 3 CD Horizontal Distance (km) I Surveyed Points)

til c) Terrace Dipping S Terrace Dipping N Horizontal Terrace cD CD CD

scC'

'07t1 cp fa. b F ound i gu 685000 690000 685000 690000 re ari 5 Fluvial terrace crosses drainage divide .3 es h 5 . d a ve b ) as een pect add and ed

e) t o sl sh o pe ow to ma th e l ps ocati ge ner on at of ed th

usi e d n r g ai

th 1 na Aspect 685000 690000 685000 690000 e S ge di LJ Flat (-1) RTM IN North (0-22.5) N Slope (Degrees) vid Northeast (22.5-67.5) ❑ ▪ 0 - 3.76 DEM es 0 East (67.5-112.5) A ▪ 3.76 - 8.10 . • Southeast (112.5-157.5) 0 South (157.5-202.5) 0 2 4 Kilometres 0 8.10 - 14.63

. W I I in Southwest (202.5-247.5) ■ 14.63 - 23.52

at ▪ VVest( 247.5-292. 5) ■ 23.52 - 50.41

ersh II Northwest (292.5-337.5) U North (337.5-360) ed 5.5.7 Water Gap / Wind Gap Comparisons Combining the above observations in a series of graphs allows comparisons of the five water gaps and wind gaps. Graphs of water gap and wind gap longitudinal stream profiles along channel length, show linear and convex profile shapes (Fig. 5.36). The water gaps are linear and water gap 2 shows a decrease in elevation to the west when compared to water gaps 3, 4 and 5. The convex up shape of the wind gaps are also at an elevation above all the water gaps. These graphs indicate a decrease in elevation of the water gaps to the west.

Figure 5.37 is a graphical plot of valley floor width at the southern range front, mid channel and north of channel, and shows a decrease from water gap 2 to 5 in valley width. This change to the east indicates that thrust displacement resulting in fold uplift influences erosion in the water gaps. The morphology of water gaps 3, 4 and 5 with increasing topography show a decrease in valley width. Increasing topography combined with a decrease in valley width and incision into older Lower Miocene rocks in water gap 5 implies that at some point if erosion does not keep up with fold development the valley will be defeated upstream. Water gap 5 is interpreted as a marginal wind gap.

Maps of Q3 deposits in water gaps 3 and 4, show the same valley plan view shape morphology to the present day morphology of water gap 2. This is further evidence for westward propagation. Also, comparison of the northern east region of water gap 2 is similar to the drainage pattern in water gap 1.

A plot of maximum and mean difference calculations made from the N-S swath profiles of the water gaps confirm the general decrease in water gap topographic relief to the west (Fig. 5.37b). Also the wind gap measurements are less than the water gaps and may indicate infilling of the channels.

The area measurements of the upstream watersheds derived from outlet points along the range front indicates a general decrease in area from water gap 5 to water gap 1 (Fig. 5.37 c). However, there is a difference in water gap 2 which has a very large upstream watershed. This area measurement only applies to the Bajada fan area. The watershed area is much larger as it is linked to a major catchment within the Bogda Shan mountain. 199

500 — Water Gap 2 Convex Water Gap 3 400 Water Gap 4 E Water Gap 5 c 300 - 0 Wind Gap 1 — Wind Gap 2 >5 200 - N 100 - Linear

0 0 1 2 3 4 5 6 7 8 9 10 Horizontal Distance (km)

Figure 5.36. Composite graph showing N-S orientated water gap and wind gap longitudinal valley profiles. a)

1200 1000 ;5 800 ■ SRangeFront 600 r. Mid Channel 72 400 - c ■ Nrange 40 200 - 0 0

2 3 4 5 Water Gap b)

300

250

200

m) (

n • mean diff io

t 150 ■ max duff Eleva 100

50 -

wgl wg2 wnd1 wnd2 wg3 wg4 wg5 Water Gap Wind Gap

Figure 5.37. a) Plot of water gap channel width at the southern range front, mid channel and north of channel, b) graph plot of maximum and mean difference N-S water / wind gap relief.

200

400 350 300 250 200 - as 150 100 50 0 n

P c \ q, (1, e,§Z \64.9 e 0,, 0.9 09 042 c °' e 4 c•t, 4 4 4 *e, # A4‘ 44' 44' 44' hoc

Figure 5.37. c) Diagram showing upstream catchment area (km2) of outlet points positioned at the range font of the water and wind gaps. Catchments automatically generated using SRTM DEM.

201 5.5.8 Flaming Mountain Wind Gap Development The two primary wind gaps located to the east of water gap 2 in plan view are clearly geomorphologically similar to the adjacent water gap fluvial valley morphology. From the digital imagery and fieldwork these fluvial streams are no longer connected to the upstream water source and the original water gap fluvial morphology, even though degraded, is excellently preserved in the fold terrain. The formation of these primary wind gaps impacts upon the evolution growth and style of this blind thrust fold. The change in the dynamic balance between erosion processes and blind thrust fold growth resulted in the formation of the wind gaps. We know that at some point post-Q2 time the upstream water source was severed. What caused this? Two scenarios are considered. Firstly, in order for a water gap to continue incising erosional processes must defeat blind thrust uplift, in particular uplift due to displacement. Thus if there is a change in the upstream erosional source or tectonic displacement is relatively rapid then the water gap is defeated. We know through post Q2 to Q3 the palaeogeography was that of glacial out washing south from the Bogda Shan (Deng et al., 2000). Thus we know that there was a significant upstream erosional source. This implies that the wind gaps were formed due to water gap defeat by tectonic processes. Further inspection of the terrain adjacent to the wind gaps shows: • Aspect map shows drainage divide north of the main fold axis • Longitudinal profile shows a drainage divide near to the north edge of the north limb. • Incision by recent erosion is mapped north of the fold axis but south of the identified stream drainage divide. The key features indicate a dynamic erosional history and indicate periods of erosion. The fact that the modern day incision is clearly not related to the drainage divide indicates that this drainage divide is not formed by (1) the main fold axis or (2) recent incision into the terrain. An important fact is that the change in aspect of the wind gap fluvial terrace is the same as the divide mapped along the stream in the field. This suggests that another tectonic feature has deformed the preserved wind gap terraces and fluvial stream course. Structural mapping of the northern fold limb in this area indicates a secondary structure. A parallel set of pervasive lineaments, possibly joints orientated NW-SE, strike in the same direction as a slight change in limb dip along the northern section of Water Gap 3. Also the terrain on the northern front of the limb is linear and is degraded with badland type erosional indicative of the southern from of the main thrust fold. As a result I propose that the primary wind gaps were formed by a back thrust 202 striking WNW-ESE along the north part of the northern blind thrust limb. Displacement on the back thrust caused defeat of the original water gap with the drainage being diverted either side of the wind gap.

5.5.9 Water Gap / Wind Gap Transition Water gaps 3, 4 and 5 have similar fluvial morphologies and plan view shape. Each of the longitudinal profiles has linear shapes indicating a transition between mainly erosional processes and tectonic uplift. Also, plots of channel width indicate a decrease to the east towards the main fold centre. This in conjunction with an increase in regional slope to the east along the northern limb and outcrop of more consolidated N1 sandstones with interbedded anhydrite layers indicate that these water gaps are in a transitional state between continuing water gap development and the formation of a wind gaps. It is unclear whether a new wind gap may form in this region. It is likely that if the tectonic processes are able to defeat erosion then water gap 5 is currently the most likely candidate for a wind gap.

203 5.6 Terrain Morphology This section describes the change in topography between the water gaps. Traverse profiles show the west-east change in terrain shape. Five N-S swath profiles are used to show the change in terrain fold morphology from west to east.

5.6.1 Terrain 1: Western fold tip

The western fold axial tip terrain (Fig. 5.38a) approximately 3 km in plan view length extending eastwards at the westernmost termination of the blind thrust fold (E677991 N4766663), is an area where recent drainage is incising into the folded landscape. This area is important as it allows the mapping and interpretation of fluvial valley formation and terrain development at the fold tip during recent Q4 geological time. The morphology of water gap 1 is described in 5.5.1. There are two main orientations that join this water gap, marked in Figure 5.38. A series of well-developed fluvial valleys join the water gap to the north. There are two main valleys that have a sinuous form in plan view with a N-S orientation. From the DEM the maximum incision is 35 m. Further to the north east there are five valleys with a decrease in incision relief as shown by DEM profiles and dissection, shown by a change Corona image textural features. Drainage patterns 50 m to the west of these N-S fluvial streams are diverted towards the western fold termination. The orientation of the streams is SW-NE at approximately 014°. Four streams join the most western fluvial stream which is curved in plan view and appears to follow the contours of the folded terrain paralleling the nose of the emerging fold. This stream eventually joins the outlet of water gap 1. The morphology of these channels is shown by transverse profiles measured using the DEM. The profiles show the relative height position of the two sets of streams. The traverse profiles show a decrease in valley relief to the west.

204 8

8

t S Streams N- tD wes

0) to

O hy a rap og a)

c top 0 O in

crease De to a)

a_ • N a) 0_ o 0 2 0_ 0 0- . 0 o 2 a w CO 7, 0 (w) Ott t3 " 0 (/) § 8

4764000

CO 00089LP 000590 000490

Figure 5.38. Geomorphology of western fold tip terrain, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM.

205 5.6.2 Terrain 2: Minor wind gaps

The terrain between water gap 1 and 2 again has two sets of drainage patterns. The main difference between this terrain and the western landscape is that the N-S swath profile is asymmetric (Fig. 5.39). The morphology of the terrain is shown by swath profiles measured west-east and north-south. Aspect calculations on the DEM show the topographic divide of the terrain. In plan view none of the well-formed fluvial channel patterns are connected to upstream catchments. Fluvial streams from upstream hinterland bajada fan drainage catchments related to the Bogda Shan are diverted to the west into water gap 1 and to the east into water gap 2 forming a localised drainage divide to the north of this elevated terrain (Fig. 5.39). The two sets of plan view drainage patterns have similar orientations to those observed to the west. Two well developed fluvial streams are mapped in plan view with a N-S orientation. These drainages continue southwards and join water gap 2. The other set contains two dominant drainages orientated SW-NE with an orientation of 26°- 42° to the west. The terrain can be divided into two areas based on textural analysis of the Corona imagery, which shows the degree of dissection based on edge enhancements of the and drainages. The south part on the terrain in plan view is well dissected with gullies forming "Badland" type morphology. To the north of the range there is little evidence of erosion and in the Corona imagery there is less evidence of major dissection of the landscape, which was confirmed in the field. This indicates headward erosion of the southern area of the range front in the hinge zone. The N-S drainages in plan view show fluvial stream morphology. These are similar to those seen to the west with sinuous plan view patterns. In the field the streams have a cross sectional fluvial morphology (Fig. 5.40a). These features were two small to be mapped by the DEM. Transverse profiles were measured using a laser ranged levelled traverse (see Chapter 3). The stream has a concave transverse profile. There are localised alluvial fans that are normal to the stream course (Fig. 5.40). Longitudinal profiles generally show a convex profile shape with small scale irregularities that give the curve a non-smooth profile. Localise small erosional fans normal to channel length suggests that the profiles are recording oblique cross section profiles through localised fans that have filled the channel. The convex profile shapes and the fluvial channel morphology indicate that these channels are no longer active and have been preserved in the folded terrain. These drainages are interpreted as minor wind gaps.

206 Fi gen (—) gure 5 e rat 0 ed .39

682000 684000 usi I 7f); . G Swath Profile Convex - asymmetric topography n CR CD g N-S Streams N S eomor I th =t4 400

e A O I co I 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 ph STER % (14 dQ ol

o 1. gy D 5

EM O of I

th . • O 94 Profile e O Topographic Profiles t iva errai t1/2'N Minor wind gap DEM profile 1 Fluv.

n b acio VV — profiler 350

et 300

ween 250 FD. 0 0.5 1.5 2 2.5 3 3.5 4 Profile line distance (km) 0

Cr wat

er DEM profile 2 CD ,-f(4 ga soo W g 400 Decrease CD p 1 9. 300 — profile O 200 •#3-f 0

and 0 0.5 1.5 2 2.5 3 35 10 Profile line distance (km) 0 0 2 , k a) AL gre 678000 680000 6820•C 684000 yscal O

Fr e a) E Facing South

Preserved Fluvial Terraces

b) —•— Survey Points

mg -15 - Lu -20 -25 - -30 0 0.5 1 1.5 N Line Distance (km) S

2 15 - 10 * ************************ • 4. • ••• 5 •••••• 0 +•." Concave Shape -5 C>3 , • ▪ -10 • Cu -15 • Survey Points 0. -20 '• - 0 500 1000 1500 2000

N Line Distance (km)

Figure 5.40. a) Photograph of preserved fluvial channel, b) EDM levelled longitudinal profiles of channels.

208 O O O 10 Drainage Divide Gaps Minor Wind 3.76 (Degrees)

3.76 - 8.10 pe 10_ Slo El 8.10 - 14.63 II 14.63 - 23.52 111 23.526 - 50.41 680000 680000

4 Kilometres r VP to o o r--

-

, o o 0 I '7 0 0 2 N A .. 0 0

0 divide crosses drainage crosses drainage Fluvial terrace 680000 680000

Southeast (112.5-157.5) Southwest (202.5-247.5) North (337.5-360) North (0-22.5) Flat (-1) Aspect CI South (157.5-202.5) • ▪ 0 Northeast (22.5-67.5) 0 East (67.5-112.5) ▪ II Northwest (292.5-337.5) O 1.West(247.5-292.5)

-4- o 0 - 0 O 0 r---

Figure 5.40.c) as pect and d)sl opem apsof terrain 2 generated usingth e SRTM DEM. Watershed boundaries have beenadd ed tosh owth e locationof the drainage divides. 5.6.3 Terrain: Water Gap 3-4 The terrain between water gaps 3 and 4, centred on E690294 N4763224, (Fig. 5.41) is different to the landscapes already described to the west. Swath profiles show the N-S fold shape. Drainage patterns at the north show diversion of drainage either side of the terrain forming a triangular shape. Again, this shows a localised drainage divide. Generally the terrain is well dissected apart from the northern most area which forms a terrace platform.

5.6.4 Terrain: Water Gap 4-5 The terrain between water gap 4 and 5, centred on E693023 N4763174, (Figure 5.42) is similar to the landscape to the west. Swath profiles shown the N-S fold morphology. There is diverted drainage to the north forming a triangular plan view shape. This is a localised drainage divide. The main part of the landscape is well-dissected but the northern part is less dissected and forms an abandoned terrace platform.

210

IV

O a

>, N

E fL3 k (

co core 0 8 ld tance 0 fo

dis E e d 2 lin de 2 file o O Pro E co >, 'en

OI O

a) 0_

2 CL file N a)

Pro :E L.= eL O h

t as 2 tb CL 83388 8 < 8 o 2 2

Swa eL • LLI (w) UOPeA013 I-- 0 0 4768000 4766000 4764000 4762000

0 _ CV CV fC

80_

( ..., 0 o 0 o 03 -' CO CO .0 CD GO

00089LP 00099LV 000179LV 000Z9LV

Figure 5.41. Geomorphology of the terrain between water gap 3 and 4, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM.

211 w a

U)

CNI _c>, 0 0) 0 0_ 0 0 E k ( a)

E tance a) E 0 >, dis

In ine cO l I4) file 0 o

a) Pr

O a_C U)

ti=

file a_ a)

Pro LE 4=

h Q_ 2 t 0 0 0 0 0 0 0 0 In 0 0 0 0 0 0 2 `14 '41 tn

Swa z •o 2 0_ (w) uogenei3 (U) UORBAela

4768000 4766000 4764000 4762000 0 8 rn

0 0 0

O CO

O O O O O O O CO CO

00099LP 00049LV 000Z9L4 Figure 5.42. Geomorphology of the terrain between water gap 4 and 5, a) greyscale Corona image showing geomorphology in plan view, b) west — east topographic profiles generated using the ASTER DEM.

212 5.6.5 Terrain Morphology Comparison The minor or secondary wind gaps positioned to the west of Water Gap 2 are formed in a similar process to the primary wind gaps in that the streams are defeated and preserved in the growing uplifting hangingwall of the fold. Aspect and slope maps indicate that the drainage divide is just north of the fold axial trace. Longitudinal profiles measured in the field indicate convex up profiles. Plan view mapping of the wind gaps indicate similar fluvial morphology to recent north-south orientated streams located in the westernmost fold terrain. The same stream shapes are observed showing the same N-S sinuous plan view pattern. However, the longitudinal profile shows an irregular convex shape. These low amplitude fluctuations correspond to minor alluvial fans related to the preserved fluvial terraces. Degradation of these terraces have formed small alluvial fans normal to the terrace deposits within the old fluvial stream. The location of these preserved geomorphic features and the plan view connection to water gap 2 indicate an historical link. The direct comparison of the N-S active drainage at the present fold tip to the similar N-S fluvial channels in the folded terrain (Fig. 5.43) indicates a change in fold tip position.

The trace of the Q1 conglomerate outcrops along the southern range front shows a change in plan view orientated at water gap 2 (Fig. 5.43). The curvature to the north of the plan view trace of the Q1 ridge to the west of water gap 2 indicates the position of a fold closures in the vicinity of water gap 2.

The identification of preserved fluvial patterns within the fold terrain to the west of water gap 2 has a significant implication on the style of deformation in the westernmost 5 km of the fold. These secondary fluvial streams now preserved in the fold terrain clearly together with the change in Q I plan view outcrop pattern indicates the location of the fold tip position in the past. The new position of the fold nose tip 5 km to the west based on N-S fluvial patterns and incision into low elevated terrain together with the decrease in relief to 0 and the diversion of drainage around the tip indicates the present day position of the tip. This change in position indicates 5km westward lateral propagation of the blind thrust since Q1 (Lower Pleistocene). This has a major implication on the style of fold development in the western 20 kilometres.

213 Figure 5.43. Summary images showing the similarity between the plan view geometry of the minor wind gaps and drainage to the west.

214 5.6.6 N-S Terrain Swath Profiles The mapping of the water gaps and terrain geomorphology in the study area shows a marked change in transverse fold topography shown by N-S terrain profiles and west — east topography to the east. The topographic relief increases to the east with the N-S terrain shape changing from a gentle convex up curve to an asymmetric curve to a twin peaked asymmetric curve (Fig. 5.44). This indicates the change in fold geometry and related erosion in the more developed eastern fold shape with an eroded fold core, thus creating the twin peaked profile.

s‘C" Steeper north limb geo" E Water Gap 4 +

0 p \ Gx*-- Water Gap 2 + 3 S Pronounced Water Gap 1 Asymmetric Geometry

Asymmetric Geometry

Gentle Fold

Figure 5.44. Schematic diagram showing the change in topographic relief N-S profiles to the east.

215 5.7 Results Summary Classification of the western twenty kilometres of the Flaming Mountain has been carried using a combination of DEM spatial modelling and field observations. The resulting terrain classification has identified several key tectonic geomorphology features (Fig. 4.45). Water gap watersheds derived from the DEM based on outlets at the southern range front indicate a change in shape and area size to the west. There is a decrease in watershed area towards water gap 2 where there is a relatively large watershed area, which is linked to a large Bogda Shan mountain catchment. Drainage patterns also indicate focusing of drainage into water gap 2.

Two sets of wind gaps have been identified in the fold terrain study area. To the east of water gap two there are two wind gaps, which in the field have curved longitudinal stream terrace edges towards the north of the fold topography. These stream channels appear to cut across the fold terrain. DEM analyses indicate drainage divides in the northern portion of the terrain (Fig. 4.45) supported by the drainage network and the flow direction DEM product.

The secondary wind gaps are located to the west of water gap two. The DEM analyses again indicate drainage divides that cuts across stream channels that extend across the terrain. In the field these longitudinal stream profiles have north to south convex shapes. Comparison of these preserved stream channels have been made to similar north-south streams west of water gap 1 and are interpreted as fold tip area streams that due to folding have been uplifted and preserved into the fold. The range front has some interesting geomorphology. The plan view trace of the Q1 conglomerates form a defined ridge that decreases in area towards water gap two. The Q1 ridge then forms minor folding that the outlet of water gap two where it then forms a second ridge that decreases towards water gap one. This Q1 competent ridge locally forms reverse drainage especially in water gaps four and five.

Southern range front plan view triangular areas have been defined based on the water gap watersheds and the fold areas along the structure at the southern edge. The size of the triangular areas varies in size but generally decreases from water gap three to two. The increase in plan view triangular area from water gap five to four is interpreted as erosion within the hinge zone of the more competent N2 rocks. The plan view triangular areas decrease in size from water gap two to the tip of the fold. 216 These plan view triangular areas and Q1 areas (Fig. 5.45) are interpreted as two temporal geomorphic zones within the range front topography. Topography formed at the range front of the growing fold forms triangular zones between the competing watershed boundaries of the water gaps. The emerging Q1 ridge due to increased folding, indicate a second younger geomorphology area at the range front.

217

'Ma colmIsIng

el f nd a work work d i servati ob e e Th . ons sh er wat and d and s ed ge ge na rai work work net a pl s di s i r r ove yed a a

Fi . S . 5A5 gure y y ummar ph geomor p p ma gy o ol uct r nst co n usi ed eri d DEM g wat ved s s ed ersh N- 0 0 0 0 0 El around foldtip Deflected drainage Mountain front01 Mountain frontarea Watershed Stream Network 68001 680000 wind gaps Secondary Drainage focused into watergap2 Primary windgaps / 690000 / 690000 Erosion inhingezone Possible dfosedbasins 7 0

2

4 Kilometres A N — 8 0 5.8 Tectonic Geomorphology Model

The identification of characteristic geomorphic features along a blind thrust fold landscape allows an interpretation of the fault growth evolution (Burbank and Anderson, 1996; Mueller and Talling, 1997; Burbank et al., 1999). One distinctive feature is the identification of wind gaps along fold axial strike. Catchment and drainage patterns in the western tip of the Flaming Mountain fold indicate westward lateral propagation. The key geomorphic features are primary and secondary wind gaps, the decrease in water gap elevation to west, decease in topographic relief to west, a decrease in water gap relief to the west and diversion of drainage around the fold nose.

The geomorphic features identified and mapped at the western tip, the latest uplifted fan region can be used to establish latest displacement. The recognition of these drainages in the uplifted terrain to the east provides evidence of uplifted defeated drainage. Isolation of these drainages, due to the termination of source streams, results in the fossilisation in the uplifted terrain. Localised erosional processes modify the stream channel morphology with small fans infilling the morphology resulting in subdued morphology.

Established water gaps themselves act as foci for re-organised drainage. Continued displacement results in the re-organisation of drainage in the upstream Bajada fans with related formation of local drainage divides between the water gaps along the northern fold limb. The interaction of the uplifted topography and the local fan morphology results in the deflection and re-organisation of streams to the water gaps. Water gap 2 has a large upstream catchment when compared to the other water gaps. Contouring of the DEM indicates lower elevation north of water gap 2, Figure 5.5. This may result from deflected drainages from the west and east, and the morphology of the Bajada fan topography to the north of the fold.

As the fold develops transportation of sediment further south into the basin is hindered by the topography. Topographic profiles show a change in fan base level north and south of the centre part of the fold. Ponding of sediments against the northern fold limb modifies the angle of regional Bajada slope. Upstream source of the drainages in the Bajada either originate from the mountain front or drainage from range outlets that have a large catchments commonly with high elevated glacial terrains.

219

Bogda Shan range front to north 1. Drainage on Bajada fan slope

Flaming Mountain blind thrust anticline

V V • Drainage linked to Bogda Shan Mountain'., catchments form Bajada fans on north `.1 0 km 7 regional Turpan basin slope

2. Proto-fold topography

Blind Thrust segment

• N. \1111111411.1411"Imm"."01

V V V V• • V • 2 3 4 5 Formation of Water Gaps as drainage incises through growing fold topography

3. Secondary drainage incised through fold tip topography joining Water Gap 2

) c\s, Legend 2 3 4 Water Gap Drainage / • Water Gap Drainage diversion 2 3 4 5 17%.0..... Thrust Anticline Topographic development at fold tip results in drainage diversion Uplift `4 Topography

Figure 5.46. a) Conceptual sketch model of drainage and catchment reconstruction in the westernmost 20 km of the fold.

220 4.

V V V 2 3 4 5 Formation of Wind Gaps due to drainage 0 km 5 defeat at northern fold limb.

5. Abandonment of secondary drainage as fold topography grows westward. - _ • •

Westward lateral propagation of blind thrust

6. Minor Wind Gaps

Legend 2 3 4 Water Gap 10 - /0/4. Drainage / Water Gap ago//o Thrust I? 2 3 4 5 Anticline New westward position of blind thrust anticline tip. Formation of minor wind gaps west of water gap 2. 41111111111 Uplift Drainage development at fold tip. Topography

Figure 5.46. b) Conceptual sketch model of drainage and catchment reconstruction (continued).

221 An evolution model is presented which summarises the proposed tectonic geomorphic development of the western 20 km of the Flaming Mountain during he Quaternary (Fig. 5.46 a & b).

5.8.1 Stage 1: Pre-Growth Bajada Fan Strata The Bajada fan slope to the west of the Flaming mountain was bisected by several drainage systems orientated north- south and sourced from the Bogda Shan range to the north. The deposition of Q1 conglomerates form extensive deposits of course grained basic rocks sourced from the carboniferous volcanics within the Bogda Shan mountain. These laterally extensive Q1 conglomerates, formed by on large Bajada fans on the north regional slope form a pre fold growth sediment deposit, which is used as an important fold growth marker.

5.8.2 Stage 2: Initial Fold Growth, Water Gap Development An approximately west — east blind thrust segment to the west of the Flaming Mountain forms with deformation results in fold topographic which decreases to the west. A series of water gaps are formed where Bajada fan drainage channels are able to continue to incise across the growing fold. The positions of the present day water gaps are marked.

The western end of the Flaming Mountain has undergone compressional deformation since Q1 times (Lower Pleistocene). It is unclear about the structural style of fault progpagtion on a blind thrust segment which is approximately 20km in length oriented west — east in plan view. The plan view outcrop of Q1 conglomerates indicates a continuous ridge extending westwards along the southern front of the fold. The plan view outcrop pattern curves northwards into water gap 2. This outcrop pattern together with bedding data suggests a fold closure near to water gap 2 for this Q1 layer. This together with outcrop patterns of N2 indicates a fault segment that has developed by a vertical displacement gradient following an initial lateral growth.

Outcrop patterns of Q2 including the absent of these conglomerates and sandstones on the southern limb are interpreted as growth strata on the developing fold with sediment ponding against the back rotated northern limb. The sediment packages within the Q2 unit indicate alternating decimetre to metre thick alternating layers of sandstones and course conglomerates. These have been interpreted by Deng et al as changes in climatic conditions during Upper Pleistocene outwash from retreating glaciers in the Bogda Shan 222 range. However, observations of the relationship been Q4 active channels and Q4 co- lluvial fans along the water gap channels suggest that the Q2 sediments are derived from two sedimentary sources. The conglomerates may simply indicate emphemeral outwash deposits from the Bogda Shan that are channelled into the water gaps, while the interbedded sandstone layers indicate localised erosion from the growing fold topography, as observed by modern day erosion patterns within the water gap.

5.8.3 Stage 3:Syn-Fold Growth Continued fold growth results in drainage being diverted around the fold nose. Secondary N-S oriented drainage is formed in the low relief fold tip topography west of water gap 2 where streams are able to incise the terrain and join water gap 2.

5.8.4 Stage 4: Wind Gap Development During Q2 time i.e. late upper Pleistocene or possibly post Q2 / early Q3 time based on stratigraphic relationships, channels and related terraces formed cutting through Q1 and Q2 deposits within 2 water gaps to the east of water gap 2 undergo a change in sediment supply. The upstream source to these water gaps is severed and these water gaps become preserved into the growing fold forming wind gaps.

It is unclear from the observations what has caused this change to form wind gaps. It maybe a combination of the following: • a decrease in stream power due to a shallowing of regional slope angles resulting from fan ponding • the result of small upstream Bogda Shan catchments that are unable to keep up with fold growth • increase stream power from water gap2 due to connection to large upstream Bogda catchment resulting in localised change in slope topography resulting in capture of neighbouring drainages due to increased incision.

Observation of the topography east of water gap 2 together with the structural mapping along water gap3 indicates the development of a kink band geometry with a minor fold on the north limb verging northwards. This change in fold geometry from a simple asymmetric geometry to a kink band geometry may have been enough to tip the balance

223 resulting in the diversion of stream along the trace of the northern outcrop of the north limb result in the wind gap.

5.8.5 Stage 5: Lateral Propagation and Minor Wind Gap Development Westward lateral propagation of the blind thrust at depth results in a change in the fold tip position. The fold growth extends westwards with a differential increase in topography. As a result drainage re-organisation occurs in the western part of the fold. Defeat of upstream Bajada Fan drainage results in the formation of minor wind gaps to the west of water gap 2. The formation of the minor wind gaps must had occurred post Q2 onwards as these original streams had incised through low relief Q2 deposits. This indicates that the change in the fold tip position occurred post Q2.

5.8.6 Stage 6: New Fold Tip Position The new fold tip is located approximately five kilometres to the west of the previous position at water gap 2. Drainage is diverted around the new position of the fold tip. Also, secondary drainage incises through the low relief joining water gap 1. This final stage summarises the plan view map of the current geomorphology at the western end of the Flaming Mountain.

224 5.9 Discussion

The observations and interpretations of the fold geomorphology has implications on the style of fault segment evolution and the significance of erosion in influencing the fold growth. Numerical model simulations of coupled fold growth and erosional processes compared to the above observations allow a further understanding of the mechanisms of fold development in the western end of the Flaming Mountain.

5.9.1 Implications for Fault Segment Evolution As discussed in Chapter 2, fault segment evolution depends on the rate and style of lateral propagation and displacement. Two end members of blind thrust segment development depends on the initial rates of lateral propagation and displacement. As a result, two end member blind thrust fold segment scenarios are considered (Azor et al, 2002). The first scenario involves the gradual lateral propagation of a blind thrust segment with an increase in segment length and differential displacement along strike. The second involves an initial lateral propagation of the fault with relatively little differential displacement with the segment reaching its final segment length. The fault then evolved by differential displacement or a lateral decrease in slip.

Geomorphic features evolving in these two scenarios will have different features (Azor et al 2002, Burbank et al, 2000). Lateral propagation of a blind thrust fold growing in length with differential displacement along strike produces characteristic landscape features including decrease in topographic relief in direction of propagation, decrease in wind gap relief, decrease in drainage density, drainage diversion around the fold nose, deformation of stratigraphically younger deposits in fold direction, and a decrease in rotation and inclination of the fold limbs (Azor et al, 2002). The western twenty kilometres of the Flaming Mountain terrain demonstrates all these key geomorphic characteristics.

Structurally the main difference between the two end member fault growth scenarios is the relative change in geological strike along the fold back limbs, in this case the northern fold limb, towards the tip. In a blind thrust fold that has reached its final segment length with little displacement, sedimentary deposition with an increasing displacement gradient will result in outcrop terminations near to the fold tip. Also the angle of the outcrop layer terminations i.e. backlimb rotations, decrease to the tip.

225 In a gradual lateral propagation with displacement the back limb rotation and outcrop pattern with be more elongate along fold limb strike. This is observed from published geological maps and fieldwork supporting the lateral propagation of the blind trust fold. The geologic map of the western end of the Flaming Mountain (Fig. 5.12) clearly indicates a gradual change in bedding along strike implying a westward lateral propagation resulting from lateral blind thrust growth.

5.9.2 Comparison with Numerical Landscape Model Simulations The observed tectonic geomorphology is compared to the cascade numerical model modified by Van der Beek et al, 2002. Comparison of the mapping and interpretation of the Flaming Mountain landscape to the various model simulations outlined in chapter 2 indicate similarities in geomorphic features. The Turpan fold belt terrain and drainage patterns are comparable to drainage patterns and morphology generated by the numerical model with a flat detachment.

The cause of water gap defeat depends on the relative rate of erosion and tectonism during fold growth. Assuming that erosion rates over the Quaternary period are constant the relative rate of propagation will impact on water gap formation. Slow propagation rates allow relatively more time for erosion to continually incised fluvial channel across the fold topography. Comparison of theses key geomorphic features indicate the evolution of the Flaming Mountain blind thrust growth.

• Fast lateral propagation rate • Flat detachment at depth

The only published seismic line interpretation (Nishidai et al., 1992) supports a flat detachment. Wind gap formation within latest Quaternary conglomerates and sandstones indicate fast propagation rates, which are also supported by deformed terrace dating, Deng et al., 2000.

Numerical model simulations of growing fold on regional slopes (Simpson 2003) indicate that the angle of the regional slope prior to deformation has an important influence on the development of growing fold. Two end member results of these model runs indicate important interaction between erosion and deformation. Initial slopes of 0 degrees, which are then subjected to a deformation field result in a series of parallel

226 cylindrical folds perpendicular to the deformation field. However, slopes with 2 degrees have very different fold geometries. The erosion from these related steeper regional slopes result in the formation of water gaps across the growing fold. These maintained water gaps remove sediments away from the growing fold topography thus changing the rheology of the structure. This changes the stress patterns within the fold area resulting the termination of fold tip and the en echelon plan view geometry of multiple fold segments. These simulations and observations suggest that the relatively steep regional slopes within the northern part of the basin are over time dramtically influencing the fold growth and development of the Flaming Mountain structure. Evidence for lateral propagation in the western 5 km of fold

• Minor wind gaps indicate past fold position • Q1 plan view geometry indicates change in fold geometry • Decrease in relief from minor wind gaps to present day drainage patterns.

Wind gaps suggest that fold growth is greater than erosion but this depends on the initial regional slope angle and the stream power of Bajada slope drainage. An important factor is the source of the drainage. Large Bogda Shan catchments are able to supply more material than small range catchments. The abundance of large glaciated catchments outlets along the range font, with relief changes of approximately 2500 m indicate upstream sources with relatively high erosive stream power that favours active erosion of the growing fold. This indicates that even in a fast propagating fold the drainage is able to continue eroding through the evolving fold topography. Is the lateral propagation in the western 5 km the result in the change in rheology of the growing fold due to the removal of relatively more sediments (fold material) as a result relatively steep (2 degree) upstream Bajada slopes. The change in rock rheology results in a change in structural deformation. This may explain the en echelon plan view pattern of the folds in the Turpan basin.

5.9.3 Fold Growth by Segmentation The conceptual model indicates that there has been lateral propagation of the fold tip during the Quaternary. Champel et al., (2002) carried out geomorphic observations and numerical simulations of fold development on the Dundwa Ridge, Siwalik, Nepal. The along strike maximum elevation of the ridge shows a stepped change in elevation of approximately 200m towards the fold tip. They carried out numerical simulations where

227 a propagating fold was able to develop over a period of 50k years. At this time the model was modified where a new thrust segment nucleation was allowed to develop in front of the fold tip of the established fold. The new segment developed in both strike directions where the one segment tip interfered with the established fold tip. Continued growth resulted in the formation of wind gaps at the location where the two segment met and the formation of a hinterland closed alluvial basin. The drainage patterns and geomorphology at 75 k years shows a remarkably similarity to the current landscape at the Flaming Mountain tip.

The lateral propagation growth of the Flaming Mountain western tip is not interpretated as a blind trust laterally propagating. The propagation is the result of the nucleation of a new segment west of the present day position of water gap 2 where this new segment has laterally propagated west and eastwards.

228 5.10 Conclusion and Summary Analysis and interpretation of the digital imagery and fieldwork has shown differences in geomorphology along the western twenty kilometres of the Flaming Mountain blind thrust anticline. The dynamic link between erosional processes and the relative rate of blind thrust fold displacement produce key geomorphic features summarised below

• Fold topographic relief decreases to west. • Drainage diverted to west around fold nose. • Water gap elevations decrease to west. • Identification of major wind gaps resulting from terminated water gaps. • Identification of minor wind gaps formed by preservation of secondary water gap drainage within the folded terrain. • Local drainage divides along northern fold limb indicating localised adjustment of drainage to fold growth. • Drainage deflections focused into water gap 2.

Image interpretation and targeted fieldwork has identified five water gaps and two wind gaps in the western twenty kilometres of the Flaming Mountain fold. Water gaps 1 and 2 have continuous water flow and have lower channel elevations. To the east, water gaps 3-5, have no continuous water flow, higher valley floor elevations and decrease in valley floor width, resulting in the interpretation of the furthest east water gap as a marginal wind gap. The formation of the two wind gaps can be explained by two processes. Fold development interactions with localised Bajada fan morphology may result in the defeat of water gaps upstream resulting in the isolation of the fluvial channels and the formation of wind gaps. The more favoured interpretation based on the geomorphology and pervasive lineaments, is water gap defeat along the northern limb due to the development of kink bands in the growing fold. The evolution of key geomorphic features, such as water gaps and wind gaps is a process of the interaction of erosion and thrust hangingwall uplift during the Quaternary. The results indicate westward propagation of the Flaming Mountain implying lateral propagation of thrust segments at depth.

229 Chapter 6: Conclusions

6.1 Summary of research conclusions The evolution of extensional and foreland basins is controlled by the interplay between erosional processes, tectonic deformation and climate (Burbank 2000). This study indicates that the identification of key tectonic geomorphic features in extensional and foreland basins allows an interpretation of the along strike spatial and temporal tectonic development. Numerical models of landscape evolution (Densmore et al 2003, Van de Beek et a12003, and Simpson 2004) are predictions of landscape morphology based on current tectonic and erosional models. Iterative numerical simulation of landscape development, based on initial spatial and temporal variable parameters, result in a range of drainage patterns and geomorphic features. The user is able to observe the effect of fixing initial model run parameters by making one or more of the variables constant. This results in a range of simulated end member scenario landscape morphologies resulting from temporal and spatial development within a region of deformation. Similar end member numerical model runs resulting in distinctive drainage patterns and geomorphic features (Densmore et al 2003, Van de Beek et al 2003, and Simpson 2004) have been observed in the two study areas.

6.1.1 Major contributions of this research The comparison of the digital and field observations in both study areas to published numerical models (Densmore et al 2003, Van de Beek et al 2003, and Simpson 2004), show similarities to specific end member first order primary geomorphic features model results. These comparisons allow an interpretation of the temporal and spatial evolution of structures in these tectonically active regions.

Star Valley Star Valley is in a region of active tectonic deformation related to the Cenozoic extension in the Yellowstone hotspot region. The stunning scenery of this valley is dominated to the east by the dramatic change in mountain range morphology related to the Star Valley faults. The integrated digital image analysis and field observations have identified key geomorphic features that have enabled an interpretation of the temporal and spatial evolution of the Northern and Southern fault segments and the zone of overlap, the Grover Relay.

230 The major contributions of the research in Star Valley are:

• Identification of catchment capture in the Grover relay • Identification of a "new to science" major Quaternary normal fault cross-cutting across the Grover relay. • Northward lateral propagation of the Southern Star Valley fault segment since the deposition of the Salt Lake Formation forming a fault tip monocline.

The most significant result is the identification of catchment capture in the Grover relay. Comparing the geomorphic mapping and field observations to numerical simulation landscape models (Densmore et al, 2003) indicates the spatial and temporal style and growth of the Star Valley normal fault segments. The fault segments are interpreted to have propagated and reached an overlap position at an early stage with relatively little vertical displacement. The main growth of the footwall has evolved predominantly by a vertical displacement gradient. However, since the deposition of the early rift phase Salt Lake Formation conglomerates there has been northward lateral propagation of the front fault indicating possible change and complexity to the temporal fault development. The new discovered Grover Relay fault has important implications on the geohazard potential for the Star Valley area and indicates that the relay is in a transitional phase and is nearly breached. Comparison of the relay fault scarps with published scarp measurements (Warren 1992 and Piety et al, 1992), indicate similar geomorphology suggesting latest Holocene displacements.

Flaming Mountain The Turpan basin has been identified as a new case study natural laboratory for tectonic geomorphology evolution in foreland basins. The excellent preserved geomorphic features have been mapped at various scales using both the digital imagery and field observations. The geometry of the Turpan fold belt is that of a series of en-echelon blind thrust anticlines in the process of overlap and coalescence. The preservation of single fold structures makes it ideal to study the tectonic geomorphic features in isolation. Fold structures to the north of the Tien Shan and Tarim basin (Molnar et al 1994, Suppe et al 2004) have complicated fold geometries often with complex overlapped and merged thrust anticline segments.

231 The major contributions of the Flaming Mountain research are: • The identification of the water gap and wind gaps in the western end indicates a dynamic landscape evolution during the Quaternary. • 5km lateral propagation of the western end of the Flaming Mountain during the Quaternary (since Q2 times). • The formation of water gaps in proximity to large upstream catchments in the Bogda Shan Mountain plays an important part in thrust anticline development.

The topographic development above a Quaternary blind thrust anticline segment has been interpreted by geomorphology and geological mapping, using digital imagery verified by field observations. The identification of water gaps, wind gaps and drainage diversion have allowed an interpretation of the Quaternary evolution in the western end of the Flaming Mountain enabling an understanding of the growth style evolution of the thrust fault anticline segment. The identification of secondary wind gaps and range front fold geometry indicates the past position of the fold nose. This together with relative fluvial incision indicates that the segment has laterally propagated five kilometres to the west since the Upper Pleistocene (Q2). It is difficult to establish the fault segment deformation style in a blind thrust as we are only observing the surface topographic expression of this structure. The continuous range front geometry together with the erosion further to the east makes it difficult to distinguish between the segment growing by gradual lateral propagation or a vertical displacement gradient. The fluvial incised topography on the north part of the fold limb along water gap 5 may indicate past fold nose position of the Flaming Mountain prior to Q1 times.

6.2 Implications for research in academia and industry Climate Effects The complex temporal and spatial interactions betweens climate, erosion and tectonic deformation result in terrain features being dominated by one or more of these main mechanisms depending on the temporal evolution and the evolutionary stage of the landscape. The two regions chosen have an historical record of active tectonics. It is clear that even in a region of active tectonics it is not simply the tectonics that is influencing the geomorphic development (Simpson 2004). Climate has an important role both in the short and long term. The constant change in hillslope erosion and removal of material from the tectonic structures redefines the rheology and shape of the

232 hangingwall and footwall, thus modifying the deformation style and growth of the evolving structure. In these regions, where the tectonic imprint is observable, can the importance of climate be assessed? The two study areas both show that proximity and size of sediment supply from upstream catchment is important.

Petroleum Basin Analysis Understanding sediment dispersal patterns within extensional rift and foreland basins is important in identifying potential petroleum source rocks, traps and reservoirs. Recognition of key geomorphic features allows predictions of regional sedimentary fluxes to basins along the strike of tectonic structures. In extensional basin fault arrays, regions of growth and interaction of multiple fault segments, catchment capture resulting from displacement gradients on normal fault segments in relay zones indicate a change in the focusing of drainage down along the relay ramp. The capture allows the formation of relatively large fans on the front fault. In large fault arrays with overlap segments it is important to look along the front segment related to the relay as these areas may form important source rock / petroleum reservoirs.

In foreland basins the development of blind thrust anticlines interacts and alters existing antecedent drainage from upstream sources such as mountain ranges. The formation of water gaps, wind gaps and drainage diversion results in a change of sediment input to a foreland basin. The research results can be applied to new exploration areas and to modify existing petroleum basin models, which will improve the mapping and interpretation of sediment dispersal pathways into basins.

233 Chapter 7: Further Work The excellent preservation of key geomorphic features in both study areas makes it ideal for further mapping. This research presents models of formation indicating the important of tectonic deformation on landscape evolution. However, the active tectonics allows the identification of key tectonic related features. This predominant overprint now allows the identification and interpretation of climate responses on a larger time scale.

7.1 Star Valley The Bridger National Forest covering the eastern part of the Star Valley has inhibited the mapping of small scale structures along the range front and in the Grover Relay. Further work would initially focus on improving the mapping of small scale geomorphic features in targeted areas. Further work is listed in the following:

• Map the Range front and Relay using airborne Laser altimetry. • Date ash layers in Salt Lake Formation to constraint SLF locally. • To carry out seismic geophysical survey across the Monocline to map the cross- sectional geometry. • To compare 10m DEM results to similar relays.

Laser Altimetry The geomorphology of the relay and the mountain range front along the eastern side of Star Valley has now been mapped using the 10m DEM. However, subtle features such as metre scale fault scarps within the forested regions are difficult to image using photogrammetric methods. The 10m DEM has been corrected for the tree canopy producing a "bald earth", treeless elevation model. This resulting elevation model gives a smoothed surface where fine scale features are lost. Laser altimetry is able to map these features by penetrating through the tree canopy and imaging the ground directly. A targeted survey mapping the mountain front along the north and south Star Valley segments, the monocline and the relay would aim to identify:

• Subtle metre scale geomorphic fault features along the mountain range front • Geomorphology of the relay fault at 1:1000 scale • Map the monocline for subtle geomorphic features 234 Post processing of the raw elevation values would need to be carried out. The laser scanning of the surface will measure heights of all landcover features producing a very large dataset of spot heights. This "point cloud" dataset includes height data related to trees and bushes in the study area. This data can be removed leaving a "bald earth" elevation model or digital terrain model representing the surface with an irregular network of heights in the non-vegetated areas. In the relay fault area the tree canopy density is not continuous so a large density of spot heights across the area would be possible. Problem areas would be along the fault scarp itself as there are many dead trees and bushes populating the scarp. The footwall wall area is relatively devoid of vegetation especially between the ridges where there are grass clearings. The hangingwall fault scarp along strike have densely spaced common bushes and trees which gradually change into more open grass areas and openly spaced trees to the north of the fault in the slip drainage areas. The laser altimetry survey would map the geomorgphology of the three targeted areas at an improved scale (approximately 1:1000). The following lists the main improvements in terrain mapping using this method.

• An updated map of the mountain range front showing scarp locations and identification of possible new fault scarps especially along the northern segment. • An updated absolute map of the relay with accurate mapping of relay fault and mapping of possible related minor structures within relay. • Updated map of geomorphology and possible fault scraps in the monocline area.

These new improved resolution elevation maps would allow further three dimensional analyses. In particular, the improved absolute mapping of the relay fault would allow three dimensional restoration of the fault along strike based on fault geomorphology and displaced fluvial terraces.

Dating of ash layers in Salt Lake formation. The identification of ash layers in new road cut exposures in the monocline area and exposures along the north side of Willow Creek allow in-situ dating of the Salt Lake Formation. Ar/Ar dating can be used to date volcanic Sanidine feldspar crystals to obtain an absolute age for the deposition of the ash layer. This would constrain the local age for the Salt Lake Formation. We would expect that stratigraphically the ash layers should be older lower down in the Salt Lake deposit sequence. However, if these ash 235 layers were deposited in the growing monocline in terraces that have become preserved in the fold then we would expect to see a reverse stratigraphic age with the ash layers being older in age. No exposures of ash layer were found in these uplifted terraces and either drill coring or back hoeing would be needed to expose stratigraphic sections across these terrace areas. Seismic mapping of the Front Fault Tip Fold The three dimensional form of the front fault tip fold could be mapped by using seismic profiling normal to the monocline fold axis. The data could be used to construct a structural model using computer modelling software.

Comparison to a similar relay (Pleasant Valley) The Pearce and Tobin fault segment in Pleasant Valley, Nevada, is the type locality for the lateral propagation fault model proposed by Jackson et al, 1994. An analysis of the 10m DEM of the fault segments and the relay would give a similar set of data already outlined and presented in Chapter 3. Comparison of the catchment morphology on the Pearce fault to the southern Star Valley segment would be used to establish if these two relays are similar and that catchment capture development has occurred in Pleasant Valley. Preliminary inspection the 10m DEM of the Pleasant Valley indicates similar footwall catchment morphology shape to the Star Valley (Fig. 5.1)

236 Fi sn gure 6 5ut r .1 . P

m0I Similar drainage patterns and ers catchment morphology to ,

pecti Grover Relay ve vi ew of

th e P ea rce and T obi n f aul t se gm Elevation ent MI 2977m s i 10 n Pl

easant 111 1352m V all e y 7.2 Flaming Mountain

The non-vegetated terrain in the western end of the Flaming Mountain makes this an ideal area for improving the mapping scale of the elevation model. Further work on the Flaming Mountain Thrust Anticline would focus on:

• Improved DEM for digital analyses to improve landscape feature comparison. • Dating of fan deposits around the fold nose and in the fold topography to indicate the change in fan age along fold strike. • To improve the Q1-Q3 mapping in the western end of the fold based on dating. • Obtain 3D seismic profile line to construct a 3D model of the structure.

DEM Resolution improvement The elevation mapping of the western part of the fold could be improved with high resolution satellite stereo imagery using Ikonos, Spot 5 stereo or ALOS imagery). The current political sensitively in the area would not permit the acquisition of stereo aerial photograph or an airborne laser altimetry survey.

Quaternary Stratigraphy Dating To use cosmogenic isotope dating (Burbank 2000) to map the age of fan deposits preserved in the fold topography. We would expect the ages to decrease more gradually and in a stepped like pattern towards the fold tip in the lateral propagation section in the western five kilometres. (Schraer et al, 2004)

Structural Modelling Acquisition of 3D seismic integrated with the elevation model for constructing a 4D structural model using a structural geology computer model (3D Move) or a tailored geomorphology computer simulation model (Van der Beek, Simpson). The 3D seismics would be essential to constraint the fault geometry at depth and to map the 3D fold geometry along strike. If the resolution is fine enough it would be important to relate the Q1-3 boundaries along fold axial strike.

238 650000 655000

650000 0 0.5 1 655000 A ...- Thrust at depth Kilometres Anticline

E

Survey Point' -5 Elevation Decreases -10 E c -15 -20

III -25

0.5 1.5 2 0 Elevation Decreases —is—Survey Point] -5

E -10

I -15

-20 Profile 2 25 0 0.2 OA 0.6 08 1.2 1.4 1.6 1.8 2

- Survey Point

-8 Profile 3 0 0.5 1 1.5 Line Distance (km)

Figure 6.2. a) Corona greyscale image showing fold morphology and diverted drainage , b) west — east laser ranged profiles of the north fold segment.

239 Salt Mountain Further tectonic geomorph features have been identified in the basin during fieldwork in 2002 and not presented here. The Salt Mountain fold to the west of the Flaming Mountain is a series of thrust anticline segments in overlap and coalescence (Fig. 5.1 (a)). Mapping has indicated that the position of the fold nose mapped by Deng et al 2000 is 2km further to the east and is defined by a series fluvial terraces formed by drainage incising through the latest fold topography.

The Deng et al, 2000 map indicates that this fan surface is an undefined Q3-Q4 in age. The west — east profile traverses indicates that the terraces in the overlap region has been deformed (Fig. 5.2(b)). The top and middle profiles are cross-sections through the fold tip topography showing a decrease in stream channel elevations to the east with an 10-15m elevation drop to the east. These elevation changes together with diverted plan view drainage patterns indicate the new position of the fold nose. The bottom profile is oriented south of the fold showing a relatively flat topography with the fluvial streams occurring at the same elevation. If the surface is Q4 in age then this indicates recent deformation, further indicating the geohazard potential in the Turpan basin. More fieldwork is required to systematically date the fan deposits in the overlap zone using cosmogenic isotope dating (Burbank, 2000).

Turpan Basin The Turpan basin, an enclosed intermontane foreland basin, is an ideal region for looking at the effects of tectonic geomorphology in a range of linked erosional environments from high glaciated terrains in the Bogda Shan Mountain to the arid desert conditions of the central lake area which is below sea level. To fully explore the Cenozoic evolution of the Turpan basin an integrated digital image analyses and field programme is proposed to map tectonic geomorphic features across the basin not just related to the thrust anticlines.

To map and interpret the basin / mountain range interaction an accurate regional elevation model is require. The SRTM mapping of terrain using interferometric radar resulted in DEM errors in mountainous regions. The SRTM DEM is accurate in low to intermediate relief areas but less accurate in high steep-sided valleys caused by lay-over radar effects. AS [ER stereo imagery which is better than SRTM in areas of high relief contrast could be used for elevation generation in these mountainous areas. The

240 development of a methodology for merging these two datasets would be essential for obtaining an accurate regional 90m elevation model for the basin surface analyses. This regional DEM would be used to firstly map the regional drainage and geomorphology of the Turpan Basin. The integration of the DEM with other digital data would be used to:

• Test regional models of erosional process mechansims located to fold topography and the influence this has on structural deformation on a local and regional scale within foreland basins. • Develop a regional tectonic geomorphology model of the Turpan Basin by modifying existing foreland basin numerical models using local Turpan Basin structural and geomorphology data.

241 Chapter 8: Bibliography

Abrams M., 2000, The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER): data products for the high spatial resolution imager on NASA's Terra platform, Int. j. Remote Sensing, vol. 21, No. 5, p. 847-859.

Allen M.B., Windley B.F., and Zhang C., 1994, Cenozoic tectonic in the Urumqi-Korla region of the Chinese Tien Shan, Geol Rundsch, 83, p. 406-416.

Allen P.A., and Densmore A.L., 2000, Sediment flux from an uplifting fault block. Basin Research, 12, p. 367-380.

Anders M.H., and Schlishe R.W., 1994, Overlapping Faults, Intrabasin Highs, and the Growth of Normal Faults. The Journal of Geology, v. 102, p. 165-180.

Azor, A., Keller,E., and Yeats, R.S., Geomorphic indicators of active fold growth: South Mountain- Oak Ridge anticline, Ventura basin, southern California, GSA Bulletin, June 2002, v. 114, no. 6, p. 745-753.

Bucknam, R. C., and Anderson, R. E., 1979, Estimation of fault-scarp ages from a scarp-height-slope-angle relationship, Geology, vol. 7, p. 11-14.

Braun, J., and M. Sambridge, 1997, Modelling landscape evolution on geological time scales: A new method based on irregular spatial discretization, Basin Res., 9, p. 27— 52.

Bruhn, R.L., Gibler, P.R., and Parry, W.T., 1987, Rupture characteristics of normal faults: and example from the Wasatch fault zone, Utah, Continental Extensional Tectonics, Geological Society Special Publication No. 28, p. 337-353.

Burbank D.W. and Anderson R. S., 2001, Tectonic Geomorphology, 294 pages.

Burbank, D.W. and Pinter, N., 1999, Landscape evolution: the interactions of tectonics and surface processes, Basin Research, 11, p. 1-6.

242 Burbank D.W., McLean J.K., Bullen M., Abdrakhmatov K.Y., and Miller M.M., 1999, Partitioning of intermontane basins by thrust-related folding, Tien Shan, Kyrgyzstan. Basin Research, 11, p. 75-92.

Burbank, D., A. Meigs, and N. Brozovic', 1996, Interactions of growing folds and coeval depositional systems, Basin Research, 8, p. 199-223.

Cartwright, J.A., Trudgill, B.D., Mansfield, C.S., 1995. Fault growth by segment link age: an explanation for scatter in maximum dis-placement and trace length data from the Canyonlands Grabens of SE Utah. Journal of Structural Geology, 17, p. 1319-1326.

Champel, B., van der Beek, P., Mugnier, J., and Leturmy, P, 2002, Growth and lateral propagation of fault-related folds in the Siwaliks of westrn Nepal: Rates, mechanisms, and geomorphic signature, Journal of Geophysical Research, vol. 107, no. B6, doi:101029/2001JB000578.

Chester, J, 2003, Mechanical stratigraphy and fault—fold interaction, Absaroka thrust sheet, Salt River Range, Wyoming, Journal of Structural Geology, 25, p. 1171-1192

Crider J.G., Oblique slip and the geometry of normal fault linkage: mechanics and a case study from the Basin and Range in Oregon, Journal of Structural Geology, 23, p.1997-2009.

Commin, D, Gupta, S and Cartwright, J, 2005, Deformed streams reveal growth and linkage of a normal fault array in the Canyonlands graben, Utah, Geology, v. 33, no. 8, p. 645-648.

Cowie, P.A., Gupta, S., and Dawers N.H., 2000, Implications of fault array evolution for synrift depocentre development: Insights from numerical fault growth model, Basin Research, 12, p. 241-262.

Cunningham, W.D., Windley, B.F., Dorjnamjaa, D., Badamgarov J. and Saandar M., 1996, Late Cenozoic transpression in southwestern Mongolia and the Gobi Altai-Tien Shan connection, Earth and Planetary Science Letters, 140, p. 67-81.

243 Davis, A. M. & Liu, J. G., 2001, ASTER DEM Generation for Geomorphic Analysis in the Bogda Shan Mountains, Xinjiang, China., Proceedings of the First Annual Conference of the Remote Sensing & Photogrammetry Society, 12-14 September 2001, p. 170-183.

Davis, A.M., Mason P.J. and Moore J,McM., 2000, Three Dimensional Visualisation of Landslides in SE Spain using Digital ElevationModel(DEM)and Orthophotography. In: Landslides in Research, Theory & Practice. Eds. E.Bromhead, N.Dixon & M-L.Ibsen. Proc.8th Intl Symposium on Landslides. Cardiff 26-30 June, Thomas Telford, London, Vol 1 403-409.

Davis, A.M., J.G. Liu and M.A.N. Brown, 2003, Drainage Development Related to Blind Thrust Evolution Along the Flaming Mountain, Turpan Basin, Xinjiang, NW China, IEEE International Geoscience and Remote Sensing Symposium Proceedings IGARSS, Vol 5, p. 3326 — 3328.

Densmore, A.L., Dawers, N.H., Gupta, S., Allen, P.A., and Gilpin, R., Fault growth and landscape evolution at extensional relay zones, Journal of Geophysical Research, 108 (B5), 2273, doi:10.1029/2001JB001741.

Densmore, A.L., Dawers, N.H., Gupta, S., Guidon, R., and Goldin, T., 2004, Footwall topographic development during continental extension, J. Geophys. Res., 109, F03001, doi:10.1029/2003JF000115.

Densmore, A. L., M. A. Ellis, and R. S. Anderson, 1998, Landsliding and the evolution of normal-fault-bounded mountains, J. Geophys. Res., 103, p. 15,203-15,219.

Deng, Q., Feng, X., Zhang, P., Xu, W., Yang, X., Peng, S., and Li J., 2000, Neo- Tectonics of the Tianshan Mountains, Seismology Publication House, Beijing, China. (in Chinese) , p. 1- 400.

Eliet, P.P., and Gawthorpe, R.L., 1995, Drainage development and sediment supply within rifts, examples from the Sperchios basin, central Greece, Journal of the Geological Society, London, Vol. 152, p. 883-893.

244 Ellis, MA., Densmore, A.L., and Anderson, R.S., 1999, Development of mountainous topography in the Basin Ranges, USA, Basin Research, 11, p. 21-41.

Formento-Trigilio, M.L., and Pazzaglia, F.J., 1998, Tectonic Geomorphology of the Sierra Nacimento: Traditional and New Techniques in Assessing Long-Term Landscape Evolution in the Southern Rocky Mountains, The Journal of Geology, vol. 106, p. 433- 453.

Fu, B., Lin, A., Kano, K., Maruyama, T., and Guo, J., 2003, Quaternary folding of the eastern Tian Shan, northwest China, Tectonophysics, 369, p. 79-101.

Goldworthy, M., and Jackson, J., 2000, Active normal fault evolution in Greece revealed by geomorphology and drainage patterns, Journal of the Geological Society, London, Vol. 157, p. 967-981.

Gupta, A. & Scholz, C.H., 2000, A model of normal fault interaction based on observations and theory, Journal of Structural Geology, 22, p. 865-880.

Gupta, S., 1997, Himalayan drainage patterns and the origin of fluvial megafans in the Ganges foreland basin, Geology, 25, p. 11— 14.

Gupta, S., and Cowie, P., 2000, Processes and controls in the stratigraphic development of extensional basins, Basin Research, 12, p. 185-194

Harbor, D.J., 1997, Landscape evolution at the margin of the Basin and Range. Geology, v.25, no. 12, p. 1111-1114.

Holmes, K.W., Chadwick, 0. A., and Kyriakidis, P.C., 2000, Error in a USGS 30- metere digital elevation models and its impact on terrain modelling, Journal of Hydrology, vol. 233, p. 154-173.

Huggins, P., Watterson, J., Walsh, J. J., and Childs, C., 1995, Relay zone geometry and displacement transfer between normal faults recorded in coal-mine plans, Journal of Structural Geology, vol. 17, No. 12, p. 1741 — 1755.

245 Jackson, J.A., and Leeder, M., 1994, Drainage systems and the development of normal faults: an example from Pleasant Valley, Nevada. Journal of Structural Geology, vol. 16, no. 8, p. 1041-1059.

Jenson S. K., and Domingue J.0., 1988, Extracting Topographic Structure from Digital Elevation Data for Geographic Information System Analysis, Photogrammetric Engineering and Remote Sensing, 54 (11): p. 1593-1600.

Jun, G., Maosong, L., Xuchang, X., Yaoqing, T., and Guoqi, H., 1998, Paleozoic tectonic evolution of the Tianshan Orogen, northwestern China, Tectonophysics, 287, p. 213-231.

Keller, E.A., and Pinter, N. 2002. Active Tectonics, 2nd edition, 362 pages.

Kenward, T., Lettenmaier, D. P., Wood, E. F. and Fielding, E., 2000, Effects of Digital Elevation Model Accuracy on Hydrologic Predictions, Remote Sensing of Environment, vol. 74, p. 432-444.

Lageson D. R., and Spearing D R., 2000, Roadside Geology of Wyoming, 271 pages.

Leeder M.R., and Gawthorpe R.L., 1987, Sedimentary models for extensional tilt- block/half-graben basins. Continental Extensional Tectonics, Geological Society Special Publication No. 28, p. 139-152.

Lin, A., Fu, B., Kano, K., Maruyama, T., and Guo, J., 2002, Late Quaternary right- lateral displacement along active faults in the Yanqi Basin, southeastern Tian Shan, northwest China, Tectonophysics, 354, p. 157-178.

Liu, J. G., 1991, Balanced contrast enhancement technique and its application in image colour composition, Int. j. Remote Sensing, vol. 12, no. 10, p. 2133-2151.

Liu, J.G., and McM. Moore, J., 1996, Direct decorrelation stretch technique for RGB colour composition, Int. j. Remote Sensing, vol. 55, no.2, p. 183-189

246 Liu, J. G., Mason, P. J., Clerici, N., Chen, S., Davis A. M., Miao, F., Deng H. and Liang L., 2004, "Landslide hazard assessment in the Three Gorges area of the Yangtze river using ASTER imagery: Zigui—Badong", Geomorphology, Volume 61, Issues 1-2, p. 171-187.

Machette, M.N., Personius, S.F., Nelson, A.R., Schwartz, D.P., and Lund, W.R., 1991, The Wasatch fault zone, Utah — segmentation and history of Holocene earthquakes, Journal of Structural Geology, Vol. 13, No.2, p.137 — 149.

Mason, P.J., Rosenbaum, M.S., and McM. Moore, J., 1998, Digital image texture analysis for landslide hazard mapping, Geohazards in Engineering Geology, Geological Society, London, Engineering Geology Special Publications, 15, p. 297-305.

Molnar, P., 1989, The Geologic Evolution of the Tibetan Plateau, American Scientist, Vol. 77, p. 350-360.

Molnar, P., Brown, E.T., Burchfiel, B.C., Deng, Q., Feng, X., Li, J., Raisbeck, G.M., Shi, J., Wu, Z., Yiou, F., Huichuan, Y., 1994. Quaternary climate change and the formation of river terraces across growing anticlines on the north flank of the Tien Shan, China. J. Geol. 102, p. 583— 602.

Moreword, N.C., & Roberts, G.P., 2002, Surface observations of active normal fault propagation: implications for growth, Journal of the Geological Society, London, Vol. 159, p. 263-272.

Mueller, K and Suppe, J, 1997, Growth of Wheeler Ridge anticline, California: geomorphic evidence for fault-bend folding behaviour during earthquakes, Journal of Structural Geology, nos 3-4, p. 383 - 396

Mueller, K, and Tailing, P, 1997, Geomorphic evidence for tear faults accommodating lateral propagation of an active fault-bend fold, Wheeler Ridge, California, Journal of Structural Geology, vol. 19, nos 3-4, p. 397 — 411.

247 Nicol, A., Gillespie, P. A., Childs, C., and Walsh, J. J., Relay zones between mesoscopic thrust faults in layered sedimentary sequences, Journal of Structural Geology, vol. 24, p. 709-727.

Nishidai T., and Berry J.L., 1991, Geological Interpretation and Hydrocarbon Potential of the Turpan Basin (NW China) from Satellite Imagery. Proceedings of the Eighth Thematic ERIM Conference on Geologic Remote Sensing, Denver, Colorado, USA, p. 373-389.

Peacock, D. C. P., and Sanderson, D. J., 1991, Displacements, segment linkage and relay ramps in normal fault zones, Journal of Structural Geology, vol. 13, p. 721-733.

Piety L.A., Sullivan J.T., and Anders M.H., 1992, Segmentation and paleoseismicity of the Grand Valley fault, southeastern Idaho and western Wyoming. Geological Society of America, Memoir 179.

Piety L.A., Wood C. K., Gilbert J. D., Sullivan J.T., and Anders M.H., 1986, Seismotectonic study for Palisades Dam and Reservoir, Palisades Project, Idaho: Denver, Colorado, and Boise, Idaho, U.S. Bureau of Reclamation, Engineering and Research Center and Pacific Northwest Regional Office, Seismotectonic Report 86-3, 198 p.

Schlische R. W., and Anders M., 1996, Stratigraphic effects and tectonic implications of the growth of normal faults and extensional basins, GSA Special Paper 303.

Schwartz, D.P., and Coppersmith, K.J., 1984, Fault Behaviour and Characteristic Earthquakes: Examples From the Wasatch and San Andreas Fault Zones, Journal of Geophysical Research, Vol. 89, No. B7, p. 5681-5698.

Scholtz, C. H., and Cowie, P. A., 1990, Determination of geologic strain from fault slip data, Geology, vol. 26, p. 967-970.

Shao L., Stattegger K., and Garbe-Schoenberg C-D, 2001, Sandstone petrology and geochemistry of the Turpan basin (NW China): Implications for the tectonic evolution of a continental basin. Journal of Sedimentary Research, 71, p. 37-49.

248 Simpson, G., Dynamic interactions between erosion, deposition, and three-dimensional deformation in compressional fold belt settings, Journal of Geophysical Research, vol. 109, doi:10.1029/2003JF000111.

Suppe, J., and Medwedeff, D. A., 1990, Geometry and kinematics of fault-propagation folding, Eclog. Geol. Helv., vol. 83, p. 409-454.

Tapponnier, P., Peltzer, G., and Armijo, R, 1986, On the mechanics of the collision between India and Asia, Geol. Soc. London Spac. Publ. 19, p. 115-157.

Tarboton, D. G., Bras, R. L., and Rodriguez-Iturbe, I., 1991, On the Extraction of Channel Networks from Digital Elevation Data, Hydrological Processes, 5: p. 81-100. van der Beek, P., Champel, B., and Mugnier, J. L., 2002, Control of detachment dip on drainage development in regions of active fault propagation folding, Geology, 30, p. 471— 474.

Warren G.A., 1992, Quaternary Geology and Neotectonics of Southern Star Valley and the Southwest Flank of the Salt River Range Western Wyoming, unpublished masters thesis, Utah State University, Logan, Utah, 96 pages.

Wallace R. E., 1978, Geometry and rates of change of fault-generated range fronts, nourth-central Nevada, Journal of Research, U.S. Geological Survey, Vol. 6, No. 5, p 637-650.

Zelilidis, A., 2000, Drainage evolution in a rifted basin, Corinth graben, Greece, Geomorphology, 35, pp. 69-85.

249 Website URL addresses:

NIMA GPS accuracy information: http://www.nima.mil/portal/siteinga01/

SRTM elevation accuracy: http://wwvv2.jpl.nasa.gov/srtm/datafinaldescriptions.html

Wyoming Geological Survey http://www.wsgs.uwyo.edu/

Wyoming GIS data http://www.sdvc.uwyo.edu/clearinghouse/

250 Appendix 1: Star Valley Data Front Segment Stream Longitudinal Profiles (Willow Creek to Swift Creek). Graphs oriented east-west due to extraction from Rivertools software. In Chapter 4 the profiles have been oriented west-east for presentation.

Location

2150 2100 —DEM Protle E 2050 o 2000

>o 1950 1900 1850

0 0.2 0.4 0.6 0.8 Line Distance (km)

2

2150 DEM Profile 2100 E SwF,- c 2050

2000

1950 -

1900

0 0.2 0.4 0.6 0.8 1 1 2 Line Distance (km)

251 3

2150 —DEM Profile ..... 2100 - E c 2050 -

i 2000 - m al 1950 -

1900 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 6 Line Distance (km)

2200 ...... 2150 - —DEM Profile ..g... 2100 - c Si 2050 - W it 2000 - w 1950 - 1900 0 0.2 0.4 0.6 0.8 1 1.2 1.4 16 Line Distance (km)

2250 2200 - —DEM Profile E 2150 - c 2100 - 4g, 2050 - 2000 - m ITI 1950 - 1900 - 1850 0 0.5 1 1.5 2 2.5 3 3 5 Line Distance (km)

252 7

2200 2150 _.„,______,._...... „...... s ,„,..,...___._...... * :DEM Profile 1 2100 oc 2050 3 2000 - 2 1950 w 1900 1850 0 0.5 1 1.5 2 2.5 3 35 Line Distance (km)

8

2300 2250 —DEM Prate .t... 2200 - .E. 2150 g 2100 - '5 2050 - it') 2000 - W 1950 - 1900 - 1850 , . , 0 0.5 1 1.5 2 2 5 Line Distance (km)

9

10

t 2000 - M. 1950 - 1900 - 1850

253

11

2250 2200 —DEM Profile 2150 2100 - g 2050 2000 1950 - 1900 1850 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 6 Line Distance (km)

12

2250 2200 - —DEM Profile 2150 2100 - ...7-9 2050 - (13 r, 2000 - w 1950 - 1900 - 1850 0 0.5 1 1.5 2 2.5 3 35 tine Distance (km)

13

2200 2150 -` —DEM Profile 1 2100- c 2050 I 2000 iu-2 1950 - 1900 - 1850 - . , . , . 0 0.5 1 1.5 2 2.5 3 Line Distance (km)

14

2200 2150 - —DEM Profile -i- 2100 - co 2050 :o as 2000 - i.02 1950 - 1900 - 1850 0 0.5 1 1.5 2 2.5 3 3 5 Line Distance (kin)

254 15

2250 2200 —DEM Profile .g. 2150 -- 2100 - c 2. 2050 - N ri 2000 111 1950 1900 1850 , 0 1 2 3 4 5 Line Distance (krn)

16

,nnn —DEM Profile 8 E

(m) 8 ion 8 t a N 8 E 8 Elev 8 3 low 0 1 2 3 4 5 6 7 Line Distance (km)

17

.10/111 —DEM Profile N 8 E

)

(m !;. 8 n io t

N Eleva 0 8

tout,O 3 0 1 2 3 4 5 6 7 Line Distance (km)

18

2300 —DEM Profile 2200 -

c 2100 - 0 ={ 2000 - o IL I 1900 ------_,-----_-. 1800 0 1 2 3 4 5 6 7 Line Distance (km)

255

19

20

2800 —DEM Profile 2600 - E c 2400 - 0 wa V. 2200 - o w 2000 - 1800 0 1 2 3 4 5 6 Line Distance (km)

21

2250 2200 —DEM Profile i 2150 - E 2100 - 2050 - 2000 - 43 a; 1950 - 1900 - 1850 0 0.5 1 1.5 2 Line Distance (km)

22

2200 2150 —DEM Profile E 2100 oc 2050 wiv 2000 92 1950 im 1900 1850

0 0.5 1 15 Line Distance (km)

256 23

2150 2100 - —DEM Profile 2050 - c a 2000 - R m 1950 - W 1900 - 1850 0 0.5 1 Line Distance (km)

24

2800 —DEM Profile 2600 - -E = 2400 - 0 (>,' 2200 - m III 2000 -

1800 0 0.5 1 1.5 2 2.5 3 3.5 4 Line Distance (km)

25

2800 —DEM Profile 2600 - E c 2400

2200 - m 2000

1800 0 0.5 1 1.5 2 2.5 3 3.5 4 Line Distance (km)

26

2200 2150 - —DEM Profile j 2100 - = 2050 - i 2000 -

1950 - ILI2 1900 - ----...... 1850 , . , , , 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 6 Line Distance (km)

257

27

2700 2600 —DEM Profile 2500 - .1.. 2400 - g 2300 1 2200 el; 2100 -

ILI 2000 - ...... 1••••••••• 1900 - ••••••••••••• 1800 , 0 0.5 1 1.5 2 2.5 3 Line Distance (km)

28

2600 2500 - —DEM Profile ,g, 2400 - 2300 -4-2. 2200 - m 2100 - w 2000 - 1900 - 1800 , . , , . 0 0.5 1 1.5 2 2.5 3 35 Line Distance (km)

29

2250 2200 —DEM Profile i 21_ 50 c 21°° sz° 2050 E 2000 in 1950 1900 - 1850 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 6 Line Distance (km)

30

2600 2500 —DEM Profile -E. 2400 2300 - ° 2200 - r, 2100 2000 - 1900 - 1800

0 0.5 1 1.5 2 2.5 3 Line Distance (km)

258

31

2200 2150 —DEM Profile ? 2100 - c 2050 - 2000 - 1950 - 1900- 1850 , 0 0.5 1 Line Distance (kin)

32

2250 2200 - —DEM Profile — 2150 - E = 2100 - 2050 - II oi 2000 - 11.1 1950 - 1900 - 1850 , , . 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 6 Line Distance (km)

33

2600 2500 —DEM Profile I 2400 --c " 2300 ct0 2200 as a 2100 - m ii) 2000 - 1900- 1800

0 0.5 1 1.5 2 Line Distance (km)

34

2500 2400 - —DEM Profile 1 2300 - c 2200 - •ci itg 2100 - ui2 2000 - 1900 1800 0 0.5 1 1.5 2 Line Distance (km)

259 35

2300 —DEM Profile ..„ 2200 - E c 2100 - 0 1 2000 m W 1900 1800 0 0.2 0.4 0.6 0.8 1 Line Distance (km)

36

2300 2250 —DEM Profile 2200 I_ 2150 ;c 2100 2050 - _m 2000 I" 1950 - 1900 1850 , 0 0.2 0.4 0.6 0.8 1 Line Distance (km)

260 Front Fault Swath Profiles Swath Profiles Difference Profiles West East West East Canyon 1

31. 201

303

00

05 0 A 0S 02 0 2 n00003.131.0003.32

Canyon 2

1-00.109

300

303

120

03 01 OS 000.001030.21.1 0303030.111.0001/20)

Canyon 3

202

01 as O. 07 as Os 05 OA 03 1.5.0.101.2.0•10.1 0•020.1120.202..

Canyon 4

20)

2100

.03

SS .6 00 11.11.1.11101111110111.01 a

261 Canyon 5

710 620

2100101041011892291001 1.1.4.0.1010.0111•11 Canyon 6

122

326

I 1100

24294600622240290 1.21920.40200202

Canyon 7

300

nse 293

180

100

11 as 06 06 601492416622460123 11013421•1131220202.

Canyon 8

103

105 nee

*al /OD

28 0.6 44 11212001•2122911141 110..101119...01*

262

Canyon 9

1--004.000

.30 100

2000

1.1500•1010.00•03i1 01.0.00110101000

Canyon 10

440

200 aaeo

i also

2rea 103

2009

15 Nonzent0100antots. 0..100.101100.0 0.0

Canyon 11

25 2 .0•00011100son0.02 witpeal 00•001.0a0,

Canyon 12

Iwo OS •••••••1010001•0•10.0 10.0000101 MIMS 10.1.

263 Canyon 13

310

20o

OA OA •• 02 ao 00.00.•,...... 10.40011010•10•11M Canyon 14

10.0.00000.A.0•0 .11•419•00

Canyon 15

203

200

10 0

1.6

SOO OB OS as u a a a 1011...10•1411/0 11.1.0.1011.1mollni

Canyon 16

1+—roavoti

ad OA ILO ae as 0.2 0 1000001110100.0,06101 mansont0lAsmoo0.11

264

Canyon 17

150

11 5 .3 .9xenta M190 199100.0.1111•1•9090910 Canyon 18

X0

010

1.5 5.95s9.1.5.19999.51 101.9.99.191.59.110119

Canyon 19

TOO

2600

3505

2X0

1.6 00 U OS OA 01 0 1900/9000.1101/99.50111m1 11.0199901091910.091.5.1

Canyon 20

361

2009

13 OA

110090991916101.00.9199 MOOR. POMO. 1061

265

Canyon 21

SAO

SSA

2000

logo" 12 22 24 .15224.50100.4•11051 oss.ssiossassallis0

Canyon 22

210

100

0.2 P0 OS as OS as as

1.11..romososea 55.1.2.55.54.00.5 Canyon 23

250

2000

2003

PO 0.3 P1 0.7 OS 04 0.2 Pt 01 11251•2.101.110.401 11.5010101.050.1

266 Grover Relay Fault Scarp Profiles (see chapter 4 for locations).

O ts-

0 O U) LU ) (m) (m e e c tan tanc dis l dis l ta ta izon izon r Ho

0 O Hor cv)

O O r

Li) LU ••r Lr) (w) eoueislp leo!PeA (w) aouelsp leD!1-1aA T4 FT4 F

1 -

2 - le file fi Pro Pro

267 Profile 3 - FT4

m) ( 15 ce tan is l d ica t 5 Ver

10 30 50

Horizontal distance (m) Profile 4 - FT4

10 30 50 70 Horizontal distance (m) Profile 5 - TF1

25 E

U

15

5

10 30 50 70 90

Horizontal distance (m) Profile 6 - FT3 E.- 15; C8

5

10 30 50 70 Horizontal distance (m) Profile 7 - FT3

(m) ce tan 10. l dis ica t 5 • . • " ''''' • Ver

10 30 50

Horizontal distance (m) Profile 8 - FT3

10 30 50 70 Horizontal distance (m)

Profile 9 - TF2

m) ( 30 nce ta is l d tica

r 10 Ve

20 60 100 140

Horizontal distance (m) Profile 10 - FT2

20 60 100 140 Horizontal distance (m)

Profile 11 - TF1

L ...- 100 200 300 400

Horizontal distance (m) Profile 12 - FT1

. • •

10 30 50 70

Horizontal distance (m) Profile 3 - FT4

(m) 15 nce ta is l d ica t

r 5 Ve

10 30 50

Horizontal distance (m) Profile 4 - FT4

15 a) C co U)

5

ci) 20 60 100 70 Horizontal distance (m) Grover Relay Fault Measurements CHD = Cumulative horizontal distance CVD = Cumulative vertical distance

Profilel Profile2 Profile3 CHD Height CHD CVD CHD CVD 0 0 0 0 0 0 4.4 0.2 12.05 1.08 5.5 0.15 10.79 0.62 21.87 1.83 11.05 0.92 18.98 1.26 28.86 2.8 16.8 1.47 23.99 1.7 32.95 3.45 22.39 2.23 28.06 2.01 35.58 4.3 27.06 2.87 29.72 2.17 38.07 5.42 29.79 3.34 31.75 2.3 39.82 6.22 31.49 4.06 33.73 2.58 41.53 6.8 33.81 4.78 35.4 2.98 43.81 7.33 37.2 5.43 36.84 3.41 46.09 7.74 41.8 6.08 38.56 3.91 48.7 8.15 46.42 6.53 40.22 4.44 55.69 8.69 51.75 6.85 41.83 4.8 66.93 9.72 57.1 7.41 43.42 5 72.35 10.12 60.88 7.71 45.4 5.09 66.45 8.11 50.69 5.57 74.58 8.81 55.53 5.87 58.83 5.93 64.84 6.29

Profile 4 Profile5 Profile6 CHD CVD CHD CVD CHD CVD 0 0 0 0 0 0 5.87 0.77 4.59 0.47 8.12 0.64 12.1 1.39 8.54 0.95 13.58 1.16 21.29 2.21 13.49 1.73 17.86 1.64 24.5 2.8 17.75 2.46 22.15 2.25 27.73 3.43 21.07 3.13 24.59 2.74 30.14 4 24.53 3.87 27.34 3.43 32.57 4.67 27.73 4.92 30.92 4.62 35.31 5.4 30.84 6.52 33.69 5.31 37.37 5.85 32.68 7.61 36.79 5.75 39.57 6.24 34.88 8.39 41.46 6.29 43.1 6.72 37.8 9.54 47.5 6.77 47.97 7.18 40.78 10.66 55.28 7.17 53.48 7.7 41.42 10.86 62.95 7.49 60.12 8.4 43.43 11.52 71.78 8.1 66.37 9.05 45.42 11.92 82.14 8.88 72.32 9.84 48.44 12.74 51.75 13.81 55.21 14.98

274 59.36 16.02 62.52 16.94 65.29 17.79 67.32 18.4 69.05 18.96 71.53 19.67 75.93 20.94 79.72 21.78 84.72 22.78 89.43 23.42 94.85 24.01 101.37 24.61 106.84 25 114.73 25.44

Profile 7 Profile8 CHD CVD CHD CVD 0 0 0 0 4.81 0.35 12.09 1.02 10.15 0.9 17.96 1.65 14.86 1.45 23.52 2.25 19.62 1.91 26.87 2.75 25.15 2.48 29.76 3.28 29.45 2.96 32.28 3.66 32.82 3.4 34.58 4.18 36.02 3.73 36.73 4.74 38.15 4,07 39.26 5.34 39.5 4.26 41.95 5.93 41.39 4.61 43.82 6.17 42.88 4.86 45.57 6.45 44.24 5.02 48.88 6.87 45.76 5.23 52.01 7.21 47.19 5.34 56.59 7.77 48.75 5.43 60.78 8.24 50.71 5.57 66.72 8.98 54.65 5.92 72.01 9.75 59.3 6.37 77.02 10.32 66.6 7.01 82.01 10.85 71.39 7.32 86.9 11.42 82.18 8.08 92.35 11.96

Profile 9 Profile10 Profile11 CHD CVD CHD CVD CHD CVD 0 0 0 0 0 0 6.24 0.27 9.1 0.45 10.26343 0.799747 14.51 0.47 18.16 0.91 20.52686 1.102579 22.07 0.74 29.34 1.89 30.79028 1.405411 27.39 0.73 34.25 2.26 41.05371 1.708243 31.73 0.74 39.96 2.92 51.31714 2.011075 36.79 1.1 43.89 3.59 61.58057 2.952705 40.17 1.56 47.87 4.44 71.844 3.616739 43.73 2.44 51.59 5.1 82.10743 4.280726 45.9 2.9 54.8 5.83 92.37085 5.444737 48.04 3.6 57.89 6.75 102.6343 6.05047 49.9 4.3 60.41 7.55 112.8977 6.766308 275 52.03 5.13 62.22 8.42 123.1611 8.1987 54.16 6.33 64.46 9.68 133.4246 8.867462 56.09 7.13 67.22 11.16 143.688 9.363078 58.04 8.23 69.36 12.37 153.9514 9.825681 60.14 9.48 72.5 13.96 164.2149 10.34223 62.42 11.04 74.67 15.26 174.4783 10.80898 64.81 12.35 77.21 16.36 184.7417 12.91882 67.13 13.69 79.16 17.15 195.0034 14.7011 69.45 15.5 81.62 18.09 205.2668 17.22347 71.73 16.75 84.17 19.06 215.5303 19.23552 72.92 17.47 86.5 19.86 225.7937 21.62736 74.54 18.21 89.56 20.62 236.0562 24.41682 75.8 18.61 92.25 21.2 246.3197 27.09278 76.92 19.38 95.47 21.74 256.5831 30.07055 78.19 19.88 99.16 22.29 266.8465 32.57972 79.92 20.6 105.01 22.93 277.1099 36.55998 81.45 21.39 110.52 23.53 287.3733 40.76207 82.93 21.99 118.93 24.47 297.6367 45.34934 84.38 22.69 126.35 25.34 307.9001 49.83095 85.66 23.27 132.66 25.92 318.1636 53.92816 87.31 24.09 138.66 26.63 328.427 58.06955 88.84 24.63 338.6904 61.73589 90.27 25.11 348.9539 64.8844 91.87 25.89 359.2173 68.02095 93.18 26.49 369.4807 70.18406 94.47 27.04 379.736 71.86333 96.54 27.67 389.9994 72.70797 98.75 28.64 400.2611 73.0999 100.19 29.22 410.5245 73.30585 102.06 29.76 420.7879 73.4302 103.28 30.21 431.0513 73.6832 105.03 30.73 441.3148 74 106.8 31.24 451.5782 73.52796 109.37 31.85 461.8414 73.90647 112.81 32.55 472.1048 73.46751 115.11 32.99 482.3682 73.77722 118.61 33.53 492.6316 74.80781 122.25 34.16 502.8951 76.07984 127.29 34.59 133.44 35.27 140.77 35.97 150.97 36.86

Profile12 Profile13 Profile14 CHD CVD CHD CVD CHD CVD 0 0 0 0 0 0 4.34 0.53 6.31 1.45 4.53 0.09 10.46 1.08 8.56 2.05 9.82 0.33 12.87 1.35 10.98 2.86 15.31 0.9 17.86 1.68 12.83 3.35 22.7 1.85 20.99 1.89 14.68 3.94 26.99 2.55 22.6 2.09 17.58 4.69 33.29 3.24 24.91 2.22 21.43 5.74 37.41 4 26.85 2.55 23.86 6.56 41.06 4.94 29.42 3.03 26.35 8.12 42.47 5.12 32.02 3.6 28.06 8.13 44.51 5.69 276 33.43 4.02 29.72 8.94 46.81 6.69 34.86 4.66 31.43 9.86 49.45 7.39 36.26 5.44 33.05 10.78 51.33 7.9 37.39 6.24 34.38 11.58 55.9 8.66 38.51 7.14 35.87 12.14 60.44 9.57 40.03 8.34 37.78 12.8 63.19 10.08 41.52 9.45 40.44 13.46 65.19 10.41 42.98 10.4 42.98 13.87 69.58 11.02 44.03 11.22 46.12 14.43 78.19 12.26 45.24 12.04 53.1 15.22 85.81 13.5 46.34 12.75 90.54 14.07 47.61 13.62 96.53 14.97 48.84 14.29 104.04 16.14 50.4 14.88 107.86 16.75 52.03 15.41 113.44 17.77 59.6 16.89 121.65 19.85 69.15 18.57 125.2 20.98 127.45 21.7 129.74 22.46 133.44 23.66 137.37 25 139.38 25,81 142.12 26.64 144.81 27.75 147.47 28.67 150.72 29.57 153.22 30.2 158.94 31.72 161.75 32.53 165.57 33.66 168.1 34.31 171.25 34.94 176.83 36.37 180.55 37.24 181.94 37.35 183.42 37.63 185.2 37.95 186.85 38.19 188.37 38.68 191.84 39.82

277

Appendix 2: Turpan Fold Data N-S Swath Profile Traverses along the Flaming Mountain.

1

5 10 15 20 25 Swath line distance (km)

2wg1

900 - E 700 -t- ----.._,_„..__,___ —Max g 500 - — Min ss ------,,,_,______0$ 300 > e 100 ------W -100 0 5 10 15 20 25 Swath line distance (km)

3wg2

900 E 700 c500 to 300 0 100 -100

5 10 15 20 25 Swath line distance (km)

4-wind gaps

900 —Max E 700 —Min oc 500 - on m 330 - w 100 - -100 0 5 10 15 20 25 Swath line profile (km)

5

900 700 — Max o 500 - —Min 42.0 300- .' 100 - W -100

0 5 10 15 20 25 Swath line distance (km)

278

6 wg4

5 10 15 20 25 Swath line distance (km)

7 wg5

900 700 g 500 Max = 300 —Min o 100 IZ -100

5 10 15 20 25 Swath line distance (km)

8

• 1900 1400 - c - max wa• 900 — min a 400 w -100

O 5 10 15 20 25 30 35 40 45 Swath line distance (kin)

9 water gap 6

1400 —max 900 —min > 400 al -100

0 5 10 15 20 25 30 35 40 45 Swath line distance (km)

10 water gap 7

1400 E — max 900 0 — min ea• 400 a w -100

O 5 10 15 20 25 30 35 40 45 Swath line distance (km)

279

11

900 E 700 — max g 500 — mmR 300 ▪ 100 - W -100 5 10 15 20 25 30 35 40 45 Swath line distance (km)

12

800 600 e, aoo - —max is t 200 - —min El 0

5 10 15 20 25 Swath line distance (kin)

13

800 600 — max S 400 - — min a rs 200 • 0

5 10 15 20 25 Swath line distance (km)

14

800 —max 600 400 — min a d 200 al 0

0 5000 10000 15000 20000 25000 Swath tine distance (kin)

280