ABSTRACT
Robert Horton's theory of sheetwash erosion and rill incision was based on his analysis of sheetwash without rainsplash. We conducted experiments on these processes on savanna hillslopes in Kenya to obtain information on the hydraulics of sheetflow and on sediment transport. These data were used to construct a simple, two-dimensional model of flow and sediment transport over natural microtopography, and thus of the spatial pattern of erosion. The model correctly reproduced Horton's prediction and our field observation that, on a sufficiently steep slope, sheetflow would become unstable and would incise a dense network of rills. When sheetwash was accompanied by rainsplash, the rills were filled with sediment and eradicated. Larger rills, formed by large volumes of sheetflow from hillslope segments 600-900 m long, were also degraded by altering the relative roles of raindrop impact and channel flow. The results support the theory that the initiation and maintenance of rills depends upon the balance between sediment transport by sheetwash, which tends to cause channel incision, and rainsplash, which tends to diffuse sediment from protuberances and to fill channels, thereby smoothing the surface.
