Evaluation of the effects of grassland distribution on erosion processes of the loess hillslopes under simulated rainfall

Vegetation distribution are of great significance to control soil erosion and water and soil loss on slope.  In order to further explore the influence of slope vegetation distribution on the process of sediment yield and discharge, and quantitatively analyze the relationship between vegetation distribution and sediment yield and discharge. Through field rainfall simulation, under the condition of 15° slope, the sediment yield and runoff were observed under the conditions of different vegetation coverage (40%, 60%), different rainfall intensity (30, 60, 90 mm/h) and different vegetation distribution positions (relative distance 0, 0.2, 0.4, 0.6, 0.8, 1). The results showed that: (1) under a certain slope and rainfall intensity, the runoff yield and sediment yield increased rapidly at first and then tended to be stable under different vegetation distribution conditions. (2) In this study, the average runoff yield and sediment yield firstly decreased and then increased with the increase of the relative distance. The average runoff yield of the slope with the relative distance of 0.2 was the minimum. (3) The random forest algorithm shows that rainfall intensity and vegetation coverage have important effects on runoff yield, and rainfall intensity and vegetation relative distance have important effects on sediment yield. (4) When the vegetation coverage was 40%, the optimal vegetation relative positions were 0~0.36 and 0~0.31, respectively, with the main objective of reducing runoff and sediment. When the vegetation coverage was 60%, the optimal relative vegetation positions were 0~0.43 and 0~0.22, respectively, to reduce runoff and sediment. This study shows that slope vegetation distribution has an important effect on sediment yield and runoff. Under the same vegetation coverage, the smaller the relative distance of the grass belt, the better the effect of reducing runoff and sediment. The research results can provide theoretical basis and data support for optimal allocation of vegetation in the process of ecological restoration.