Identifying water quality parameter concentrations and their drivers is important for the prevention and control of water environment pollution. In this study, we constructed an inverse model of water quality parameters based on measured water quality parameters and remote sensing spectral data for this study area using artificial neural networks. We investigated the water environment response of the urban water network in the Pearl River Delta under the influence of typhoon rain events and explored their spatial heterogeneity using a multiscale geographically weighted regression model. The results indicate that in regions with a high level of urbanization, the dissolved oxygen (DO) concentration in river water is lower, and the ammonia nitrogen (NH3-N) concentration is higher. Under the influence of typhoon rain events of varying intensities, the response of water quality parameters in the urban water network of Zhongshan City varies. The intensity of rainfall determines the impact of typhoon rain events on water quality parameter concentrations. Our results are expected to improve the understanding of water quality trends under the influence of typhoon rain events and help policymakers and planners better develop water environment control strategies during typhoon transit.