Surface reconstruction and derivation of erosion rates over several glaciations (1Ma) in an alpine setting (Sinks Canyon, Wyoming, USA)

At middle to high latitudes, many alpine valleys have been shaped by glaciers associated with periods of Pleistocene glaciation. Present glaciated valleys are characterised by broadened valley floors and U-shaped cross sections, continuously formed by glacial activity from initially V-shaped, fluvial cross sections. Sinks Canyon (Wind River Range, USA) is a glaciated valley characterised by a typical U-shaped cross section, containing till from several glacial advances over a range of at least 1Ma. The morphostratigraphic records indicate a fourfold difference in ice surface elevation between the youngest and oldest glacial periods, which is not easily explained by the present-day valley topography. To assess possible evolution scenarios of Sinks Canyon, we modelled the palaeovalley topography using a geographic information system (GIS) filtering technique in combination with temporal reference points from relative and numerically dated glacial deposits. Ice thicknesses were calculated using the shallow ice approximation.In our model, the valley became shallower and the topography smoother with increasing years back in time. The results suggest that valley topography with ages between 640 and 1000ka can clearly be distinguished from the present-day topography. Surfaces with ages of 130–200ka (attributable to MIS 6; Bull Lake glaciation) still could be discerned from present-day topography, but with relatively high uncertainties. The method did not work for topography less than ~100ka or older than ~1Ma. Erosion depths were calculated using the differences between present-day elevation and the modelled surfaces. Calculated erosion rates were within the range of reference values for glacial erosion (0.001 to 1mm a⁻¹). Glacial erosion appears to have removed 0.52 to 0.72mm a⁻¹ of rock within a time frame of 1Ma, assuming 200ka of aggregated glacial flow. If the glacial occupation was longer or the impact of fluvial erosion was not negligible (as assumed), then the calculated rate is lower. If the model assumes a less extended time of glacial flow (<100ka), the rate exceeds 1mm a⁻¹. The method used for surface reconstruction and erosion rates calculation is highly dependent on (i) the reliability of the temporal anchor points and (ii) the assumption of physical glacier conditions (basal shear stress).