Evaluation of SCS-CN Method for Incorporation of Antecedent Precipitation

Soil Conservation Service Curve Number (SCS-CN) method currently known as Natural Resources Conservation Services curve number (NRCS-CN) method, is one of the extensively used reliable, simple, and attractive model in practical hydrology for direct surface runoff (rainfall-excess) prediction of a given storm, initially developed for direct surface runoff estimation in small and medium agricultural watersheds of the USA, later extended to other geographical regions of different land use land cover, and climatic conditions of different parts of the earth viz, to rural, urban, forest, experimental. Besides of various improvements, the method has also been extended to a number of hydrological application beyond its initial purpose. This study evaluates incorporation of antecedent precipitation (P5) in place of the antecedent moisture (Mishra and Singh 2004 models) in the Soil Conservation Services-Curve Number (SCS-CN) method using a large set of rainfall-runoff data from 234 small to large experimental watersheds from USDA-ARS. Three variants of the proposed models (M3, M6, and M9), Mishra and Singh models (M2, M5, and M8), and existing SCS-CN (M1, M4, and M7), out of which the first two models of proposed, Mishra and Singh, and existing models are one parameter or CN based and the third model of each are two parameters (CN and λ) models are considered. Employing the widely used performance evaluation goodness of fit (GoF) criterion of root mean square error (RMSE) and ranking based grading system indicate that all the variants of proposed models in turn performing far better than variants of Mishra and Singh model followed by variants existing model, respectively.