A Generalized Storage Function Model for the Water Level Estimation Using Rating Curve Relationship

Abstract

This study proposes a novel generalized storage function (GSF) model for water level estimation from the rating curve relationship by considering (i) the spatial distribution of rainfall over the basin and (ii) incorporating all the possible inflow and outflow components to reduce the uncertainties involved. The proposed GSF model, along with three other models, was then applied in two watersheds of Japan to examine its applicability in different types of watersheds with optimized parameters: (i) the Iga watershed, a semi-urban watershed and (ii) the Oto watershed, a rural watershed. Further, the proposed model’s effectiveness was identified based on hydrograph reproducibility, Akaike information criterion, and Akaike weight. The results showed that the GSF model performed well in both watersheds compared to the other models. Moreover, the Morris global sensitivity method has used to analyze the sensitivity of the GSF model parameters for the objective function of root mean square error.