The relative importance of landslides and soil creep for eroding steep soil-mantled hillslopes

Abstract:

Unlike gentle hillslopes that erode primarily by soil creep, soil-mantled hillslopes that exceed a threshold angle (typically around 30°) can erode by a combination of processes that are catastrophic (e.g., shallow landslides) and gradual (e.g., soil creep and ravel). Because landslides can quickly move large volumes of soil, it is tempting to assume that landslides dominate erosion on all hillslopes steeper than the threshold angle. However, we have little data to assess the relative role of landslide erosion as a function of hillslope angle. Here, we address this knowledge gap using landslide databases from California and Switzerland that include measurements of landslide size and topographic position. We find that landslide size decreases as a function of hillslope angle while landslide density is approximately constant for all hillslopes steeper than the threshold angle. Consequently, shallow-landslide erosion peaks near the threshold angle (~30° for most landscapes) and decreases by a factor of two on hillslopes that are just 5° steeper. Moreover, in these landscapes and in other published datasets of soil thickness, soils are thinner on steeper hillslopes. These observations have two possible explanations: that steeper hillslopes with thinner soils have slower overall erosion rates at our study sites; or, that erosion by gradual processes (soil creep and ravel) increases on steeper slopes, thereby thinning soils and limiting landslide occurrence. We argue that this second explanation is more consistent with existing theory for soil production, soil creep, and landslide mechanics. Thus, our results suggest that shallow (soil) landslides are most erosive near the threshold hillslope angle, and that soil creep and ravel may be increasingly dominant on steeper hillslopes.