Uncertainty Analysis of Dominating Hydrological Parameters in Diverse Hydrometeorological Micro-Watersheds in Krishna Basin, India

Abstract

Hydrological model plays a vital role in water resources management and serves many applications, including water resources planning development and management, agriculture, and flood production. Rainfall-runoff modeling in any watershed is highly influenced by the different hydrological parameters such as Curve Number (CN_2), Groundwater Delay (GW_DEALY), Base flow (ALPHA_BF), ground water Revap (GW_REVAP), Threshold depth of water (GWQM), and ESCO. Therefore, the rainfall- runoff model was simulated using the Algorithm SUFI-2 of SWAT-CUP in the Soil-Water Assessment Tool (SWAT) in two diverse hydro-meteorological watersheds of Marol (area: ~5158 km2), and Talikot (area: ~2370 km2) in Krishna basin of Southern India, where the unregulated flow exit. The results show that the R2 and NSE-values for the calibration of streamflow in Marol watershed are 0.87 and 0.84, respectively, and in the validation, 0.65 and 0.58, respectively. Similarly, for the Talikot watershed, they are 0.90 and 0.52; and 0.83 and 0.74, respectively. The curve number, groundwater delay and water available capacity of the soil are the most sensitive parameters in the Marol watershed, whereas the threshold water level in the shallow aquifer for the base flow is the additional sensitive parameter. The uncertainty ranges of the different sensitivity parameters are observed for both the watersheds. The calculated coefficient of variation shows that, among the different parameters, the GW_DELAY has relatively high uncertainty, but the curve number (CN_2) and threshold depth of water (GWQMN) parameters have relatively low uncertainties.