Optimization of Runoff Curve Number Considering the Effect of Slope on Small Watershed in Loess Plateau.

Slope has an important effect on the runoff production process. Slope correction of runoff curve number in the SCS-CN runoff production model is important to improve the model prediction accuracy, In order to conduct an investigation of the effect of average watershed slope in the runoff production process, in this paper, based on Huang's slope modification formula, the first 40 out of 70 rainfall runoff measurements in the study area from 1955 to 1981 were used as the validation data, and a1 and a2 parameters in Huang's slope modification formula were re-determined, so as to propose the slope modification formula of the CN value for the study area under three different antecedent moisture conditions, and then the SCS-WJG method, which is of good applicability to the region, was established and compared to the SCS-CN look-up table method, and the SCS-Huang method, which is a direct transfer of the Huang's slope modification formula. The conclusion is as follows: the λ-value of the initial rainfall loss rate determined using the ergodic method is 0.09, so the standard value of 0.2 is not applicable to the study area. After comparing the size of the CN values obtained by the three methods, it is found that the CN value obtained by the SCS-Huang method is the largest, followed by the SCS-CN look-up table method, and the smallest is obtained by the SCS-WJG method, so the CN value obtained by the SCS-WJG method is more suitable for the study area as shown in the simulation results. Comparison of the SCS-CN lookup table method and the SCS-Huang method reveals that the NSE of the SCS-Huang method is lower and the RMSE is higher; therefore, it is necessary to propose a slope correction formula for the CN value applicable to the study area and establish the SCS-WJG method. When λ = 0.09, the NSE of the SCS-CN lookup table method, and the SCS-Huang method are -2.85 and -3.25, respectively, while the NSE of the SCS-WJG method is increased to 0.86, and the RMSE of the SCS-WJG method is reduced by 77%, and 78%, respectively, compared with the first two methods. The simulation accuracy of the SCS-WJG method is significantly better than that of the remaining two methods, indicating that the SCS-WJG method has good regional applicability and can provide a reference for the simulation of runoff production in small watersheds. [ABSTRACT FROM AUTHOR]