2018 年 74 巻 4 号 p. I_973-I_978
Various Storage Function (SF) models have been widely used among different parts of the world in which the Urban SF (USF) model is a newly developed SF model mainly for urban watersheds. In this study, we aim to conduct the performance evaluation of USF model and to compare the results with the conventional SF models of Hoshi, Prasad, Kimura, and the linear model. The reproducibility of hydrograph was evaluated using the performance evaluation criteria's of Root Mean Squared Error, Nash Sutcliffe Efficiency (NSE), and other error functions of peak, volume, and time to peak. The results revealed that higher values of NSE for USF model indicate that the hydrograph reproducibility by USF is more reliable than that reproduced by other SF models. Further, AIC and Akaike Weight (AW) were used to compare and identify the best model among all based on the information criteria perspective. The USF model received the lowest AIC score and highest AW in most of the events which indicate that the USF is the most parsimonious model compared with other SF models.