DEVELOPMENT OF SCS-CN MODEL FOR RUNOFF ESTIMATION USING SOIL MOISTURE PROXIES

Abstract

The Soil Conservation Service Curve Number (SCS-CN) technique created by the USDA-Soil Conservation Service (SCS, 1972) is generally utilized for the estimation of direct runoff for a given rainfall events from little rural watersheds. The initial soil moisture assumes a critical part in re-organizing of the SCS-CN technique and empowers to anticipate irrational sudden jumps in surface runoff estimation and this has incited the idea of incorporating the physical soil moisture accounting (SMA) technique to create enhanced SCS-CN based models for accurate runoff prediction capabilities from rainfall realistically. Applying the idea of SMA method with altered parameters in the initial moisture conditions over the watershed, Michel et al. (2005), Water Resour Res 41 (2):1–6, built up an enhanced SCS-CN model (MSCS-CN), which could be an alternate to existing SCS-CN technique. Though, their model still acquires a few reasonable constraints and irregularities for runoff prediction capabilities. The Soil Moisture Proxies (SMP) accounting procedure for runoff estimation is one of the recent developments and has been a key component of transmutation of the existing SCS-CN method to various improved variants. The SMP/SMA methodology depends on the idea that higher the initial soil moisture higher will be runoff and vice-versa. Therefore, the present study is an new attempt and remodelling the existing SCS-CN method by improved Soil Moisture Proxies/Accounting (SMP/SMA) procedure, thereby showing the way for a revised version of the Michel et al.(MSCS-CN) model. The study reveals that new SMP procedure has been incorporated as a continuous function for antecedent soil moisture in the MSCS-CN model and obviates the sudden jumps in runoff estimations and resolves the impediments of earlier SCS-CN and MSCS-CN models. iii This improved model i.e. IMSCS-CN model was applied for judging the accuracy and acceptability and its appropriateness over the MSCS-CN and existing SCS-CN models. These models were applied to the available rainfall–runoff events based data sets from 33 watersheds which are located across the USA. Their relative performance was assessed utilizing Nash and Sutcliffe efficiency (%) (NSE) and root mean square error (mm) (RMSE) tests and Ranking and Grading System (RGS). It was found that the IMSCS-CN scores most astounding mark (90 marks with rank I) trailed by MSCS-CN with 70 (marks with rank II), and SCS-CN model with aggregated total of 38 mark (marks with rank III) with respective to the maximum of 99. In view of the general outcomes acquired from this study, it can be concluded that the IMSCS-CN model is one of the most excellent model and has several advantages and performs better than the MSCS-CN and SCS-CN models.