Glacial landscape evolution by subglacial quarrying: A multiscale computational approach

Abstract

Quarrying of bedrock is a primary agent of subglacial erosion. Although the mechanical theory behind the process has been studied for decades, it has proven difficult to formulate the governing principles so that large-scale landscape evolution models can be used to integrate erosion over time. The existing mechanical theory thus stands largely untested in its ability to explain postglacial topography. In this study we relate the physics of quarrying to long-term landscape evolution with a multiscale approach that connects meter-scale cavities to kilometer-scale glacial landscapes. By averaging the quarrying rate across many small-scale bedrock steps, we quantify how regional trends in basal sliding speed, effective pressure, and bed slope affect the rate of erosion. A sensitivity test indicates that a power law formulated in terms of these three variables provides an acceptable basis for quantifying regional-scale rates of quarrying. Our results highlight the strong influence of effective pressure, which intensifies quarrying by increasing the volume of the bed that is stressed by the ice and thereby the probability of rock failure. The resulting pressure dependency points to subglacial hydrology as a primary factor for influencing rates of quarrying and hence for shaping the bedrock topography under warm-based glaciers. When applied in a landscape evolution model, the erosion law for quarrying produces recognizable large-scale glacial landforms: U-shaped valleys, hanging valleys, and overdeepenings. The landforms produced are very similar to those predicted by more standard sliding-based erosion laws, but overall quarrying is more focused in valleys, and less effective at higher elevations.