Soil detachment capacity of shallow overland flow in Earth-Rocky Mountain Area of Southwest China

•Effective stream power was used to predict Dc with good performance.•Stream power was the optimal parameter for predicting Dc for purple soil.•Linear equation stream power was the optimal model for predicting Dc for purple soil.•Power equation flow discharge and slope model was useful for predicting Dc of purple soil. In process-based erosion models, soil detachment capacity (Dc) is a pivotal parameter that is typically estimated by hydraulic parameters. Purple soil in Earth-Rocky Mountain Area of Southwest China is a unique soil resource with characteristics including fast physical weathering, soft texture, and rich nutrient reserves. However, the prediction accuracy of Dc and its relationship with certain hydraulic parameters remains poorly understood for purple soil, which is defined as an Entisol in the US Soil Taxonomy system. The objective of this study was to investigate the relationship between Dc of shallow overland flow and hydraulic parameters to establish an accurate model for purple soil. A series of 825 flume experiments were conducted in a 12 m × 0.3 m hydraulic flume. Flow discharge was 8.33 × 10−3 to 41.67 × 10−3 m2 s−1, slope gradient was 8.7% to 39.1%, and rill length was 0.5 m to 10 m. Results showed that flow discharge and slope, stream power, and effective stream power were effective predictors of Dc. However, flow depth, flow velocity, shear stress and unit stream power were relatively poor predictors. Stream power was the optimal parameter for predicting the Dc of shallow overland flow. Flow discharge, slope and effective stream power were also useful parameters for predicting Dc. Linear equation between Dc and stream power was the optimal model for predicting Dc, which was recommended when flow depth and flow velocity data were available; however, power equation among Dc, flow discharge and slope also showed good performance in predicting Dc, which was recommended when only flow discharge and slope data were available. The results of this study provide theoretical guidance for predicting sediment yield in Earth-Rocky Mountain Area of Southwest China as well as a decision-making basis for controlling soil and water loss in Earth-Rocky Mountain Area of Southwest China.