Abstract:

The accuracy of hydrological forecasting is closely related to the multi-source uncertainty in the modeling process, and its interactive effects will further increase the prediction uncertainty. To reduce its impact, the Jinghe River basin was taken as the research object and the global sensitivity analysis method (LH-OAT) was used to effectively obtain the “available” parameter set of SIMHYD Hydrological model. Secondly, the impact of multiple uncertainties on runoff and soil moisture simulation was explored based on the hydrological model driven by meteorological elements obtained from the Statistical Downscaling Model (SDSM). Finally, the analysis of variance (ANOVA) method was used to dynamically quantify the relative contribution of various sources of uncertainty and their interactions to hydrological prediction uncertainty. The results show that the annual average runoff of the Jinghe River basin has a significant decreasing trend year by year. The “available” parameter group can better reproduce the hydrological process of the watershed, but its uncertainty has a significant impact on the uncertainty of the simulation results. The relative contributions of parameter, climate model, and climate change scenario uncertainties to monthly scale runoff are 30%, 40%, and 10%, respectively, and 75%, 15%, and 5%, respectively to soil water content. The interaction between multiple uncertainties before and after the flood has significantly increased. The research results are particularly important for reducing the uncertainty of hydrological prediction and improving the accuracy of hydrological simulation.

Key words: multiple uncertainties, climate change, hydrologic model, interaction effect, Yellow River Basin