Assessing Uncertainty in Hydrological Processes Using a Fuzzy Vertex Simulation Method

Abstract

In this study, a fuzzy vertex simulation method (FVSM) is developed for assessing parameter uncertainty on modeling outputs of Xiangxi River Basin in China. The FVSM couples a fuzzy vertex analysis approach with a semidistributed hydrological model to reflect effects of uncertain parameters on hydrological possesses. In FVSM, the topography-based hydrological model (TOPMODEL) is used for predicting streamflow of saturated area in the catchment by tackling the distribution of the topographic index; the fuzzy vertex analysis method specializes in dealing with ambiguous coefficients based on fuzzy sets theory and α-cut analysis technique. The modeling outputs (i.e., Nash-Sutcliffe efficiency coefficient and the sum of the squared errors) indicate a good performance of TOPMODEL in describing the daily streamflow at four hydrological stations in the study basin. The uncertainties of two parameters (M and T0) are examined under different α-cut levels. The results indicate that flow interval is wide under a low degree of plausibility (i.e., a low α-cut level); conversely, a high degree of plausibility would lead to a narrow interval. Sensitivity analysis and factorial analysis are conducted, the results of which imply that the effect of M is more significant than the effect of T0 on hydrological modeling simulation. Results also reveal that the interactive effect of M and T0 is small, which can be neglected in the subsequent simulation. The findings can help quantify the multiple uncertain parameters and their interactions on modeling simulation as well as enhance the hydrological model’s capabilities for water resources management.