Challenges of Simulating Small-Scale Erosion Processes in Catchment Models

Abstract:

Soil erosion is linked to overland flow processes that often occur in patchy patterns within the landscape. This patchiness arises from local differences in soil surface conditions, anthropogenic effects, vegetation cover, microtopography, and rainfall distribution and is not typically represented in catchment models. The continuity of overland flow is extremely important in predicting the severity of event-based erosion and the delivery of sediment to channels. Without some connectivity across the hillslope, surface water will reinfiltrate and any erosion that develops will be minor and local in nature. However, if runoff patches are connected by the development of other disturbances or overwhelmed by runoff during large events, the connectivity of overland flow will increase and erosion will be enhanced. A number of studies have shown that levels of both overland flow and erosion are generally greater in small plots compared to hillslopes because of opportunities for reinfiltration as runoff progresses. However, if runoff becomes concentrated and begins to incise rills and gullies, these ‘hot spots’ may increase erosion at larger scales and actually connect runoff from otherwise minor sediment contributing portions of the landscape into erosion channels. When external sources of water augment these dynamic systems (e.g., road and trail runoff), severe erosion consequences may occur. Capturing this important behaviour in catchment sediment models requires the identification of erosion hot spots and other anthropogenic effects that affect runoff, evaluation of thresholds for rill and gully incision, and assessment of topographic attributes that facilitate sediment linkages from source areas to channels. In many managed sites, a small percentage of the catchment contributes most of the sediment, thus modelling needs to identify these source areas and their linkages to streams in a parsimonious manner.