Soil effective clay content and sediment load reduce soil detachment rate by rill flow

Abstract

Soil detachment and sediment transport are the two primary rill erosion processes. The sediment load in rill flow undergoes continuous variation due to the continuity and feedback of soil detachment and sediment transport processes, potentially impacting the soil detachment rate in the next stage. However, numerous studies focus on the soil detachment process by clear rill flow, and the few existing studies on soil detachment process by sediment-laden rill flow fail to consider the effect of soil properties. Therefore, this study was conducted to analyze the effect of soil properties and sediment load on the soil detachment rate by sediment-laden rill flow (D_r), decipher the variation in D_r, and establish a model equation to predict D_r, where soil properties and sediment load were introduced. An indoor rill flume simulation experiment was conducted under combinations of five soil types, five slopes, five flow discharges, and five sediment loads. The results revealed that the D_r of Shenmu sandy loess is the largest with a mean of 2.06 kg m⁻² s⁻¹, followed by D_r of Ansai loess (1.57 kg m⁻² s⁻¹), Yangling clay loess (1.37 kg m⁻² s⁻¹), Dingbian sandy loess (1.19 kg m⁻² s⁻¹), and Changwu loess (1.14 kg m⁻² s⁻¹). The effective clay content was the optimal soil property index correlation with D_r. D_r decreased with increasing sediment load and effective clay content. Variation partitioning revealed that the explanatory fraction of flow discharge was the highest (0.32), followed by the sediment load (0.21), slope (0.19), and effective clay content (0.14). There are interactions among the influencing factors in process of soil detachment. The sediment load level and effective clay content could inhibit the influence of flow discharge on the D_r. The higher the slope, flow discharge, and sediment load level, the greater the influence of the effective clay content on the D_r. D_r by sediment-laden rill flow can be modeled using a quaternary power function of the slope gradient, flow discharge, sediment load, and effective clay content (R² = 0.863). Introducing the sediment load and effective clay content as factors in the model equation of D_r can improve the simulation precision. These findings aid in advancing the development of a physical process-based rill erosion model.