Characterization of Soil Surface Roughness Effects on Runoff and Soil Erosion Rates under Simulated Rainfall

Abstract

This study aims at identifying the influence of soil surface roughness from small to large aggregates (random roughness) on runoff and soil loss and to investigate the interaction with soil surface seal formation. Bulk samples of a silty clay loam soil were sieved to four aggregate-size classes of 3 to 12, 12 to 20, 20 to 45, 45 to 100 mm, and packed in soil trays set at a 5% slope. Rainfall simulations using an oscillating nozzle simulator were conducted for 90 min at an average rainfall intensity of 50.2 mm h(-1). Soil surface roughness was measured using an instantaneous profile laser scanner and surface sealing was studied by macroscopic analysis of epoxy impregnated soil samples. The rainfall simulations revealed longer times to initiate runoff with increasing soil surface roughness. For random roughness levels up to 6 mm, a decrease in final runoff rate with increasing roughness was observed. This can be attributed to a decreased breakdown of the larger roughness elements on rougher surfaces, thus keeping infiltration rate high. For a random roughness larger than 6 mm, a greater final runoff rate was observed. This was caused by the creation of a thick depositional seal in the concentrated flow areas, thus lowering the infiltration rates. Analysis of impregnated soil sample blocks confirmed the formation of a structural surface seal on smooth surfaces, whereas thick depositional seals were visible in the depressional areas of rougher surfaces. Therefore, from our observations it can be learned that soil surface roughness as formed by the presence of different aggregate sizes reduces runoff but that its effect diminishes due to aggregate breakdown and the formation of thick depositional seals in the case of rough soil surfaces. Sediment concentration increased with increasing soil surface roughness, due to runoff concentration in flow paths. Nevertheless, final soil loss rates were comparable for all soil roughness categories, indicating that random roughness is only important in influencing runoff rates and the time to initiate runoff, but not in influencing sediment export through soil loss rates.