Abstract: With the rapid advancement of urbanization, urban landscape patterns have undergone significant changes, further intensifying the complexity of urban climate.Although previous studies have examined the impact of urbanization on extreme rainfall, there remains a significant research gap in understanding how different types of urban landscapes and their future evolution influence extreme rainfall events with varying durations.In this study, we construct a time-varying non-stationary generalized additive model for location, scale, and shape (GAMLSS) based on six urban landscape indicators, combine it with the Patch-generating Land Use Simulation (PLUS) model to explore the impact of future urban landscape configurations on extreme rainfall.The results indicate that: ① The non-stationary GAMLSS model based on urban landscape indicators outperforms traditional stationary models in simulating extreme urban rainfall.② For short-duration extreme precipitation events (≤3 h), the impervious surface percentage (ISP), patch diversity (PD), and Shannon′s diversity index (SHDI) appear more frequently in the optimal non-stationary models.In contrast, for long-duration extreme rainfall events (> 3 h), PD and SHDI exhibit an increase in relative frequency, while ISP shows a decrease.③ Under future scenarios of different land use development, although the design rainfall values of different land use scenarios show slight differences, the spatial distribution trends remain consistent.For short-duration heavy rainfall, areas with high design value are mainly concentrated in the central-northern region.In contrast, areas with high design value for long-duration heavy rainfall have a wider impact range, covering not only the central-northern region but also the southeastern coastal areas.This study can provide critical insights into extreme rainfall forecasting and flood risk assessment under the background of urbanization, contributing to the optimization of urban planning and disaster mitigation strategies.