Evaluation of runoff prediction models using Curve Number based analogy and empirical approach in micro-watersheds in North-Western tract of India

Abstract

The runoff prediction serves as ultimate criteria in designing hydrological structures, water resources, reservoirs, and ponds etc., in the North-Western tract of India. Keeping these points in view, the study was conducted with hypothesis in evaluating the runoff prediction models employing curve number- based analogy and empirical technique in the micro-watersheds. In this regard, the characterised soil moisture retention parameter (S) parameter and Initial abstraction (Ia), were employed in predicting runoff using different Curve Number based models such as Traditional Curve Number (CNM), Mishra’s (MRM), Michel’s (MCM) and Ajmal’s (AJM). The performance evaluation of the models was tested using goodness of fit procedures viz., Root Mean Square Error (RMSE), Nash Sutcliffe efficiency (NSE) and Per cent Relative Error (PRE). However, to quantify monthly, and annual runoff, the CNM model performed better over the other models at Patiala-Ki-Rao MW-I, MW-III and at Saleran MW-I, MRM model at Patiala-Ki-Rao MW-II, MW-IV and Saleran MW-II, MW-III and MW-IV. The quantified runoff in relation to contribution through extreme rainstorms was close to the observed runoff through AJM model in MW-I, MCM model in MW-II and MW-IV, and MRM model in MW-III respectively at Patiala- Ki-Rao. Similarly, the quantified runoff in relation to extreme rainstorms was close to the observed runoff at Saleran through MRM model in MW-I, CNM model in MW-II and MW-III, and AJM model in MW-IV, respectively. However, upon evaluating the runoff prediction models in the tract , the empirical model (regression analysis) quantified runoff quite accurately with 91.9 % ( R2 ) variability in runoff by considering runoff as dependent variable and other 5- factors (Slope steepness, potential retention parameter, initial abstraction, time of concentration and lemniscate ratio) as independent variables.