The Three Gorges Reservoir region suffers from severe soil erosion that leads to serious soil degradation and eutrophication. Interrill erosion models are commonly used in developing soil erosion control measures. Laboratory simulation experiments were conducted to investigate the relationship between interrill erosion rate and three commonly hydraulic parameters (flow velocity V, shear stress τ and stream power W). The slope gradients ranged from 17.6% to 36.4%, and the rainfall intensities varied from 0.6 to 2.54 mm·min–1. The results showed that surface runoff volume and soil loss rates varied greatly with the change of slope and rainfall intensity. Surface runoff accounted for 67.2–85.4% of the precipitation on average. Soil loss rates increased with increases of rainfall intensity and slope gradient, Regression analysis showed that interrill erosion rate could be calculated by a linear function of V and W. Predictions based on V (R2 = 0.843, ME = 0.843) and W (R2 = 0.862, ME = 0.862) were powerful. τ (R2 = 0.721, ME = 0.721) did not seem to be a good predictor for interrill erosion rates. Five ordinarily interrill erosion models were analyzed, the accuracy of the models in predicting soil loss rate was: Model 3 (ME = 0.977) > Model 4 (ME = 0.966) > Model 5 (ME = 0.963) > Model 2 (ME = 0.923) > Model 1 (ME = 0.852). The interrill erodibility used in the model 3 (WEPP) was calculated as 0.332×106 kg·s·m–4. The results can improve the precision of interrill erosion estimation on purple soil slopes in the Three Gorges Reservoir area.
Hydrological processes play important roles in soil erosion processes of the hillslopes. This study was conducted to investigate the hydrological processes and the associated erosional responses on the purple soil slope. Based on a comprehensive survey of the Wangjiaqiao watershed in the Three Gorges Reservoir, four typical slope gradients (5°, 10°, 15°and 20°) were applied to five rainfall intensities (0.6, 1.1, 1.61, 2.12 and 2.54 mm·min-1). The results showed that both surface and subsurface runoff varied greatly depending on the rainfall intensity and slope gradient. Surface runoff volume was 48.1 to 280.1 times of that for subsurface runoff. The critical slope gradient was about 10°. The sediment yield rate increased with increases in both rainfall intensity and slope gradient, while the effect of rainfall intensity on the sediment yield rate was greater than slope gradient. There was a good linear relationship between sediment yield rate and Reynolds numbers, flow velocity and stream power, while Froude numbers, Darcy-Weisbach and Manning friction coefficients were not good hydraulic indicators of the sediment yield rate of purple soil erosion. Among the three good indicators (Re, v and w), stream power was the best predictor of sediment yield rate (R2 = 0.884). Finally, based on the power regression relationship between sediment yield rate, runoff rate, slope gradient and rainfall intensity, an erosion model was proposed to predict the purple soil erosion (R2 = 0.897). The results can help us to understand the relationship between flow hydraulics and sediment generation of slope erosion and offer useful data for the building of erosion model in purple soil.