Effects of soil surface roughness on interrill erosion processes and sediment particle size distribution

Soil surface roughness significantly impacts runoff and erosion under rainfall. Few previous studies on runoff generation focused on the effects of soil surface roughness on the sediment particle size distribution (PSD), which greatly affects interrill erosion and sedimentation processes. To address this issue, a rainfall-simulation experiment was conducted with treatments that included two different initial soil surface roughnesses and two rainfall intensities. Soil surface roughness was determined by using photogrammetric method. For each simulated event, runoff and sediment samples were collected at different experimental times. The effective (undispersed) PSD of each sediment sample and the ultimate (after dispersion) PSD were used to investigate the detachment and transport mechanisms involved in sediment movement. The results show that soil surface roughness significantly delayed runoff initiation, but had no significant effect on the steady runoff rate. However, a significant difference in the soil loss rate was observed between the smooth and rough soil surfaces. Sediments from smooth soil surfaces were more depleted in clay-size particles, but more enriched in sand-size particles than those from rough soil surfaces, suggesting that erosion was less selective on smooth than on rough soil surfaces. The ratio of different sizes of transported sediment to the soil matrix indicates that most of the clay was eroded in the form of aggregates, silt-size particles were transported mainly as primary particles, and sand-size particles were predominantly aggregates of finer particles. Soil surface roughness has a crucial effect on the sediment size distribution and erosion processes. Significant differences of the enrichment ratios for the effective PSD and the ultimate PSD were observed under the two soil surface roughness treatments. These findings demonstrate that we should consider each particle size separately rather than use only the total sediment discharge in assessing and modeling interrill erosion processes. (C) 2017 Elsevier B.V. All rights reserved.