Evaluation of shear stress and unit stream power to determine the sediment transport capacity of loess materials on different slopes.

Purpose: This study aims to evaluate the relationship between loess soil-based sediment transport capacity and the most well-known and extensively used shear stress and unit stream power for different steep slopes. This study also determined the suitability of shear stress- and unit stream power-based transport capacity functions for rill flow on non-erodible bed. Materials and methods: Loess soil was collected from Ansai County, which is located in a typical loessial region in China's Loess Plateau. The median diameter of the loess soil was 0.04 mm. The experiment was conducted in a rill flume with a soil-feeding hopper. The slope gradients in this study ranged from 10.51 to 38.39%, and the flow discharges per unit width varied from 1.11 × 10 to 3.78 × 10 m s. The sediment transport capacity was measured for each combination. Results and discussion: Results showed that T can be effectively described by the power function shear stress-based equations for various slope gradients with R > 0.94 and P < 0.01. Shear stress was a good predictor of T for different slope gradients with the Nash-Sutcliffe model efficiency (NSE) from 0.94 to 0.99. Moreover, shear stress was better in predicting T when the slope gradient was above 21.26%. T can be efficiently described by the power function unit stream power-based equations for various slope gradients with R > 0.95 and P < 0.01. Unit stream power was a good predictor of T for different slope gradients with NSE that ranged from 0.95 to 0.99. The unit stream power predicted T better when the slope gradient was above 26.79%. Unit stream power was more satisfied than shear stress for predicting T under different slope gradients. The unit stream power-based LISEM, which was multiplied by 0.62 (i.e., the correction coefficient), predicted well the sediment transport capacity of the rill flow in our experiment, where NSE = 0.93. The shear stress-based Zhang model, which was multiplied by the correction coefficient of 0.77, adequately predicted the sediment transport capacity of rill flow in our experiment, where NSE = 0.81. Conclusions: By performing the controlled rill flume experiments, this study showed that shear stress and unit stream power strongly influenced T for certain slope gradients under non-erodible conditions. [ABSTRACT FROM AUTHOR]