Performance evaluation of modified versions of soil conservation service curve number method for estimating runoff from watersheds of Jharkhand state (India)

Abstract

Rainfall and runoff are important phenomena for watershed management. One of the commonly used methods to estimate the volume of direct runoff for a given rainfall event is the soil conservation service curve number (SCS−CN) method. However, the researchers throughout the world felt a need for improvement or overhauling of the method; resulting in many modifications in the recent past. Recently, an improved version of SCS−CN model called as Sahu−Mishra−Eldho (SME) model was reported incorporating hydrologically sounder procedure for accounting antecedent moisture in the Mishra−Singh (MS) model, an important modified version of the SCS−CN method. The present study evaluates performance of SME model, MS model and original SCS− CN method by applying them to the rainfall−runoff datasets of selected four watersheds (Adda−1, Chitankhari, Indra and Karimati) of Jharkhand state (India). Three statistical criteria, namely the standard error (SE), root mean square error (RMSE), and normalised root mean square error (NRMSE) were used as performance evaluation criteria to assess model performance. The results obtained and their analysis indicated that the MS model and the SME model perform equally well and much better than the original SCS−CN model (with λ = 0.2 or 0.3). Moreover, the MS model and the SME model show consistent performance compared to the original SCS−CN model. As the SME model has theoretically many advantages over the MS model and the duo perform equally well, the SME model is suggested for the practical applications, especially for the watersheds of Jharkhand (India).