To address the challenges of determining water source composition using isotope tracing methods in large-scale alpine basins, as well as the low accuracy of rainfall-snowmelt runoff simulations and the difficulty in quantifying flood risk encounters, a distributed hydrological model, iRainSnowHydro, was established for water source partitioning and runoff simulation in the Shigu Basin of the upper reaches of the Jinsha River. Based on the division results of rainfall and snowmelt water, combined with observed flow data, the Copula function was used to construct two-dimensional and three-dimensional joint distributions to calculate the encounter risks of wet and dry for rain, snow, and flood combination events. The results indicate that the iRainSnowHydro model performs well in simulating daily runoff at the upper reaches of the Jinsha River, with Nash efficiency coefficients exceeding 0.8. There are significant seasonal differences in the water source proportion, with snowmelt water in spring accounting for a high and stable proportion of(27±6)%. In the twodimensional joint distribution, the risk of synchronous wetdry encounters between rainfall and streamflow is the highest at 56.34%. In the threedimensional joint distribution, the risk of synchronous wetdry encounters for rainsnowflood is 15.39%, and the probability of extreme weather events involving synchronous occurrences of rain, snow, and flood is 2.55%.