Paraglacial sediment storage quantification in the Turtmann Valley, Swiss Alps Dissertation zur Erlangung des Doktorgrades (Dr. rer. nat.) der Mathematisch-Naturwissenschaftlichen Fakultät der Rheinischen Friedrich-Wilhelms-Universität Bonn vorgelegt von Jan-Christoph Otto aus Lemgo Bonn 2006 Angefertigt mit Genehmigung der Mathematisch-Naturwissenschaftlichen Fakultät der Rheinischen Friedrich-Wilhelms-Universität Bonn 1. Referent: Prof. Dr. Richard Dikau 2. Referent: Prof. Dr. Lothar Schrott Tag der Promotion: 20.11.06 Diese Dissertation ist auf dem Hochschulschriftenserver der ULB Bonn http://hss.ulb.uni-bonn.de/diss_online elektronisch publiziert. CONTENTS CONTENTS I LIST OF FIGURES III LIST OF TABLES VII 1 PROBLEM STATEMENT AND MAIN OBJECTIVES 1 2 SCIENTIFIC FRAMEWORK 3 2.1 MOUNTAIN ENVIRONMENTS AS GEOMORPHOLOGICAL SYSTEMS 3 2.1.1 Time and space in mountain geosystems 8 2.2 THE SEDIMENT BUDGET APPROACH 11 2.2.1 Denudation rates and sediment yield 16 2.2.2 Sediment budget and storage quantification 17 2.3 EVOLUTION OF MOUNTAIN LANDSCAPE SYSTEMS 21 2.3.1 Uplift and erosion of mountains 21 2.3.2 The paraglacial sedimentation cycle 23 2.4 SEDIMENT STORAGE LANDFORMS 30 2.4.1 Talus slopes and talus cones 30 2.4.2 Block slopes 33 2.4.3 Rockglaciers 34 2.4.4 Moraines 36 2.4.5 Rock fall deposits 37 2.4.6 Alluvial deposits 38 3 METHODS FOR SEDIMENT STORAGE ANALYSIS 40 3.1 GEOMORPHOLOGICAL SYSTEM AND LAND SURFACE PATTERN ANALYSIS 40 3.2 LANDFORM CLASSIFICATION 42 3.2.1 Derivation of primary attributes 42 3.2.2 Derivation of secondary attributes 43 3.3 TOPOGRAPHICAL, DIGITAL IMAGERY AND GEOMORPHOLOGICAL BASE DATA 45 3.4 METHODS FOR SEDIMENT STORAGE QUANTIFICATION 46 3.4.1 Shallow subsurface geophysical investigations 46 3.4.1.1 Seismic refraction (SR) 47 3.4.1.2 2D- electrical resistivity tomography (ERT) 52 3.4.1.3 Ground penetrating radar (GPR) 54 3.4.1.4 Acquisition of geophysical data 56 3.4.2 Volume quantification using DTM analysis 59 3.4.2.1 Sediment thickness interpolation in the Hungerlitaelli 59 3.4.2.2 Volume quantification of the Turtmann Valley 61 3.4.3 Calculation of denudation rates and mass transfer 67 3.4.4 Uncertainties and error estimation of bedrock detection and volume estimation 68 I 3.4.4.1 Uncertainties of bedrock detection using geophysical methods 68 3.4.4.2 Error estimation in volume calculation 69 4 STUDY AREA 72 4.1 GEOMORPHOLOGY 73 4.2 GEOLOGY 75 4.3 CLIMATE 75 4.4 GLACIAL HISTORY AND PALEOCLIMATE 78 4.5 PREVIOUS WORK IN THE TURTMANN VALLEY 81 5 RESULTS 82 5.1 CHARACTERISTICS AND SPATIAL DISTRIBUTION OF SEDIMENT STORAGE LANDFORMS 82 5.1.1 Landform distribution within hanging valleys 88 5.2 GEOPHYSICAL SURVEYS 91 5.2.1 Detection of the regolith-bedrock boundary with seismic refraction surveying (SR) 91 5.2.2 Detection of the regolith-bedrock boundary using Electric Resistivity Tomography (2D-ER) 97 5.2.3 Detection of the regolith-bedrock boundary with ground penetrating radar (GPR) 101 5.3 SEDIMENT VOLUME QUANTIFICATION 105 5.3.1 Sediment volume of the Hungerlitaelli 105 5.3.2 Sediment volume of the Turtmann Valley 111 5.3.2.1 Subsystem hanging valleys 111 5.3.2.2 Subsystem main valley floor 114 5.3.2.3 Subsystem glacier forefield 116 5.3.2.4 Subsystem trough slopes and remaining areas 119 5.3.2.5 Total Sediment volume of the Turtmann Valley 120 5.4 MASS TRANSFER AND DENUDATION RATES 121 6 DISCUSSION 128 6.1 PARAGLACIAL LANDFORM EVOLUTION OF THE TURTMANN VALLEY 128 6.2 SEDIMENT STORAGE IN THE SEDIMENT FLUX SYSTEM OF THE TURTMANN VALLEY 132 6.2.1 Storage volumes and mass transfer 133 6.2.2 Denudation rates 135 7 CONCLUSION 138 8 SUMMARY 141 9 REFERENCES 144 10 APPENDIX A A. SEISMIC REFRACTION MODELLING RESULTS A B. 2D-RESISTIVITY INVERSION RESULTS P C. GROUND PENETRATING RADAR IMAGES V II LIST OF FIGURES Figure 2.1 Caine’s alpine sediment cascade model (Caine 1974) 5 Figure 2.2 Meso scale sediment flux model of the Turtmann Valley (Otto and Dikau 2004) 6 Figure 2.3 Mountain Zones by Fookes et al. (1985). Zone: 1 – High altitude glacial and periglacial, 2 – Free rock faces and associated slopes, 3 – Degraded middle slopes and ancient valleys floors, 4 – Active lower slopes, and 5 – Valley floors. 8 Figure 2.4 Time and space scales in geomorphology (Brunsden 1996) 9 Figure 2.5 Qualitative sediment flux model of the Brändjitaelli hanging valley (Otto and Dikau 2004) 15 Figure 2.6 Cross profile through the Rhone Valley derived from seismic reflection surveying at Turtmann (Finckh and Frei 1990) 20 Figure 2.7 The paraglacial model by Church and Ryder (1978) 24 Figure 2.8 The paraglacial exhaustion model (Ballantyne 2002). Rate of sediment release (λ) is related to the proportion of sediment ‘available’ (S) at time (t) since deglaciation as l = ln(S )/- t . 25 t t Figure 2.9 The paraglacial sedimentation cycle modified by Church and Slaymaker (1989). The time scale spans approximately 10 ka. 26 Figure 2.10 Changing volume of sediment storage (Ballantyne 2003) 28 Figure 2.11 Episodic impacts on the sediment input within the paraglacial cycle of the Lillooet River, Canada (Jordan and Slaymaker 1991) 28 Figure 2.12 Model of paraglacial sediment yield for catchments of different size (Harbour and Warburton 1993) 29 Figure 2.13 Coalescing talus slopes at the entry to the Bortertaelli. 31 Figure 2.14 Different talus slope types (Ballantyne and Harris 1994). 32 Figure 2.15 A block slope exposed to the south in the Hungerlitaelli. 34 Figure 2.16 Active rock glacier in the Hungerlitaelli. 35 Figure 2.17 Lateral moraine deposits in the Pipjitaelli 37 Figure 2.18 Rock fall deposit in the Niggelingtaelli 38 Figure 2.19 Alluvial deposit have almost filled up a small lake the Niggelingtaelli 39 Figure 3.1 Toposequence for arctic-alpine environments, Greenland (from Huggett and Cheesmann 2002, originally by Stäblein 1984) 44 Figure 3.2 A – Principle of seismic wave refraction and reflection. B – Travel-time–distance plot (i – angle c of incidence, V – velocity layer 1, V – velocity layer 2, t – intercept time, X – crossover point). 51 1 2 i cross Figure 3.3 Configuration of the Wenner Array: A current is passed from electrode A to B. By measuring the potential between electrodes M and N the apparent resistivity ρ in layers 1 and 2 is determined. The distance a between the electrodes always remains constant, while the configuration is shifted along the spread. 53 Figure 3.4 Principle of GPR measurement. T – Transmitter of radar waves; R – Receiver; a – Offset between T and R. 55 Figure 3.5 Procedure steps of seismic refraction data analysis 57 Figure 3.6 Locations of geophysically derived (yellow) and modelled (blue) thickness locations used for the sediment thickness interpolation in the Hungerlitaelli. 60 III Figure 3.7 Principle of the SLBL method indicating intermediate steps of the procedure. At each step a point is replaced by the mean of its two neighbours minus the toleranceD DDD z. (from Jaboeydoff and Derron 2005) 64 Figure 3.8 The glacier forefield of the Turtmann Valley. 66 Figure 4.1 Location of the Turtmann Valley, Swiss Alps 72 Figure 4.2 The southern end of the Turtmann Valley terminated by the Turtmann glacier to the right and Brunegg glacier to the left. The peaks in the left background are Bishorn (4135 m) and Weisshorn (4504 m) 74 Figure 4.3 View from the Hungerlitaelli across the main trough into some western hanging valleys. The peak towards the left is Les Diablons (3609 m). 74 Figure 4.4 Geological cross section through the penninic nappes around the Turtmann Valley. The nappes are: 1–Houillère-Pontis, 2–Siviez-Mischabel, 3–Mont Fort, 4–Monte Rosa, 5–Zermatt-Sass Fee, 6– Tsaté, 7–Dent Blanche (from Laphart 2001) 75 Figure 4.5 Mean annual air temperature and monthly precipitation figure from the climate station in the Hungerlitaelli (Altitude 2770 m). 77 Figure 4.6 Younger Dryas extent in the Valais, Switzerland. (modified after Burri 1990, from: Schweizerische Gesellschaft für Ur- und Frühgeschichte 1993) 80 Figure 5.1 Land surface classification of the hanging valleys 83 Figure 5.2 A - Altitudinal distribution of classified storage land surface. B – Hypsometric curve of the hanging valley area. 84 Figure 5.3 Directional frequency distribution of mean aspect values for sediment storage landforms. (Colours correspond to Figure 5.2) 86 Figure 5.4 Different toposequences found in the Grüobtaelli. The roman numbers indicate the toposequence type (cf. Table 5.4) 88 Figure 5.5 Relative landform storage type area (%) per hanging valley. 90 Figure 5.6 Location of seismic profiles (SR) and sediment storage landforms in the Hungerlitaelli. (For a description of landform colours please refer to Figure 5.1). 92 Figure 5.7 Sounding SR04_2: Model of refractor locations and velocity distribution (A), travel-times (B) and cross-section of refractor layers (C). The seismic modelling includes the location of the refractor surfaces calculated with the network raytracing method and of the velocity distribution derived from the tomography modelling. The numbers give the velocities (in m s-1) of the modelled layers using the network raytracing method. Diagram B shows the observed (black lines) and modelled (coloured lines) travel-times of this sounding. The colour scale on the right refers to the modelled velocity distribution derived from the tomography modelling. The lower diagram (C) depicts a cross-section through the talus slope indicating the location of the two observed refractor surfaces. 96 Figure 5.8 Location of the electric resistivity profile (2D-ER) and sediment storage landforms in the Hungerlitaelli. (For a description of landform colours please refer to Figure 5.1). 97 Figure 5.9 Combined inversion of ER profiles ER04_5q and ER04_5q2. Bedrock boundary is indicated by the white dashed line. 99 Figure 5.10 Inversion of profile ER05_6 located in the centre of the Hungerlitaelli. A strong resistivity change is observed at two locations that is attributed to the groundwater situation assumed. 101 IV Figure 5.11 Location of GPR-profiles and sediment storage landforms in the Hungerlitaelli (For a description of landform colours please refer to Figure 5.1). 102 Figure 5.12 Radargram of survey GPR04_6 in the forefield of the Rothorn glacier, upper Hungerlitaelli. Internal reflections are marked in red. The upper image shows the recorded data without including the topography, the lower image includes the topography. 103 Figure 5.13 Interpolated regolith thickness in the Hungerlitaelli. Geophysical data is indicated in yellow. Blue lines indicate the transects used for the interpolation. The interpolation was done with the TOPOGRID algorithm in ArcGIS 9.1. 105 Figure 5.14 Bedrock transects through the Hungerlitaelli. The dark line represents the land surface, the grey line is the interpolated bedrock surface based on the squares. The gray diamonds represent bedrock surface information derived from geophysics, the black squares show points of assumed depth. Transect A – Cross profile through the Rothorn cirque (vertical exaggeration: 3.75:1), Transect B – Longitudinal profile along the central thalweg of the Hungerlitaelli starting below the Rothorn glacier and terminating at the valley entry (vertical exaggeration: 4.2:1). 107 Figure 5.15 Boxplot of storage landform sediment thickness derived from the interpolation in the Hungerlitaelli. The single marks represent extreme values that lie outside a range of more than 1.5 box length away from the upper quartile. 109 Figure 5.16 Location of the sediment storage subsystems and sediment source areas 111 Figure 5.17 Comparison of volume distribution between scenario I (A) and scenario II (B) in all hanging valleys. Main differences between scenario I and II result from correction of rock glacier thicknesses. 114 Figure 5.18 A – 3-dimensional shaded relief image (DTM 5m) of the modelled glacial trough base. The valley floor part of the DSM has been lowered using the SLBL procedure. The curvature of the modelled bedrock surface corresponds to the mean trough slope profile curvature. B – Depth of the modelled valley fill. Bright colours represent deeper areas, dark colours shallower parts. C – Close- up of the modelled trough surface showing the deeper surface (dark colours) in the wider valley parts (foreground) and a decrease of depth (bright colours) at the narrow locations (background). 115 Figure 5.19 Cross-profiles through the valley floor with modelled bedrock surface (gray line). A – Profile crossing a narrow valley floor part. B – Profile located across a wider part of the valley floor. 116 Figure 5.20 A – Cross profile through the lowest part of the glacier forefield in close proximity to the dam. Black dots represent the inserted assumed interpolation points. See text for details. B – Longitudinal profile through the glacier forefield. Black dots mark interpolation points at crossings with the cross profiles. 118 Figure 5.21 Interpolation of the Turtmann glacier forefield sediment thickness. The blue dots represent the interpolation points used in the surface modelling. The glacier area was removed afterwards before the sediment volume is calculated. 119 Figure 6.1 Model of paraglacial landform succession based on the formation of glacier derived rock glaciers in the hanging valleys in three time steps. 130 V Figure 6.2 Sediment storage and Post Glacial subsystem coupling in the Turtmann Valley sediment flux system. Coupling between glacier forefield and main valley floor does not regard the construction of the dam (A = area and V = volume) 133 VI LIST OF TABLES Table 2.1 Mountain geomorphic systems and appropriate approaches to measurement (Slaymaker 1991) 4 Table 2.2 Mean sediment thickness values from preceding studies. 32 Table 2.3 Compilation of rock glacier thickness from literature. 36 Table 3.1 Primary and secondary landform attributes (Dikau 1989) 42 Table 3.2 Methods and previous studies of storage volume quantification 46 Table 3.3 Geophysical properties of chosen subsurface material (different sources). 48 Table 3.4 Electrical properties of different material (Dielectric constant, conductivity and radar wave velocity) (different sources) 56 Table 5.1 Sediment storage size and altitudinal distribution 82 Table 5.2 Geomorphometric parameters of storage landforms. 85 Table 5.3 Mean minimum and maximum distance of storage landforms to ridges and drainage ways. 87 Table 5.4 Landform toposequence mapped in the Turtmann valley. The gray shaded sequence parts represent a landform coupling in a coarse debris sediment cascade. 88 Table 5.5 Geometric characteristics of the hanging valleys in the Turtmann Valley 89 Table 5.6 P-wave velocities and refractor depths of seismic profiles in the Hungerlitaelli. 95 Table 5.7 2D-ER soundings in the Hungerlitaelli. 98 Table 5.8 Ground penetrating radar profiles and detected bedrock surfaces in the Hungerlitaelli 104 Table 5.9 Area and volume distribution of sediment storage landforms in the Hungerlitaelli. Rock glacier volumes are calculated assuming an ice content of 50% for active, and 30 % for inactive rock glaciers. Mean depth of moraine deposits, inactive and relict rock glaciers include uncorrected values in brackets (see text). 110 Table 5.10 Modelled sediment storage volumes in the Turtmann hanging valleys. Volumes for active and inactive rock glaciers consider debris contents of 30 % and 50 %, respectively. 113 Table 5.11 Modelled sediment volume distribution and volume–area ratio for different subsystems of the Turtmann Valley. For a description of the two scenarios refer to chapter 5.3.2.1. 120 Table 5.12 Mass transfer within the different subsystems of the Turtmann Valley. 122 Table 5.13 Mass transfer of the different storage types within the hanging valleys. 123 Table 5.14 Denudation rates for different subsystems of the Turtmann Valley. 125 Table 5.15 Denudation rates of single landforms: A – talus slopes, B – talus cones, C – block slopes, D – talus–derived active rock glaciers based on volumes of this study, and E – talus–derived active rock glaciers based on volumes of Nyenhuis (2005). 126 Table 6.1 Comparison of alpine denudation rates. 136 Table 6.2 Comparison of denudation rates and rock wall retreat rates in alpine and arctic environments. 137 VII 1. Problem statement and main objectives 1 Problem statement and main objectives Sediment flux plays a central role within the evolution of land surfaces and the Earth’s biogeochemical system. A sediment budget tries to quantify sediment fluxes on various scales. Sources, sinks and storages of sediment are the major components of a sediment budget. The quantification of the magnitude and time-scale of sediment storage flux is still the weakest part of sediment budget studies. However, it is considered to be the most important link between sediment movement and landform evolution (Slaymaker and Spencer 1998). In mountain environments sediment fluxes are heavily influenced by topography and glaciation. The accumulation, storage and release of sediment in mountain areas affected by glaciation operate on different time- and space-scales (Church and Ryder 1972; Ballantyne 2002a). Process rates and operation times changed in the past, thus generating a sequence of landforms that compose today’s land surface. Sediment storage landforms are often assembled in a nested manner, creating neighbouring, overlapping, or underlying landform patterns. The role of sediment storage within a sediment budget approach is often based on estimations only. However, geophysical methods, high resolution digital terrain data and GIS techniques open up new possibilities for the quantification of sediment storage volumes. This study analyses the spatial distribution of sediment storage landforms and quantifies sediment volumes in the high Alpine Turtmann Valley in the Swiss Alps. The sediment flux system generally includes the transport and storage of fine and coarse solid materials and dissolved matter. As this study is based on the actual distribution of sediments on the land surface, it concentrates on solid sediments only. The following main questions will be addressed: • How are sediment storage landforms distributed in the Turtmann Valley? • What kind of functional relationships exists between these landforms? • How can the sediment storage volume be quantified for an entire Alpine meso-scale catchment? • How much sediment is stored in the Turtmann Valley? • Which storage landform types store the largest quantities of sediment? 1