Abstract: Urban impacts on spatial and temporal rainfall variabilities present significant challenges for effective urban flood mitigation and adaptation strategies. However, the physical mechanisms underlying these impacts remain unclear. In this study, we conducted modeling analyses using the Weather Research and Forecast (WRF) model, combined with thunderstorm identification and tracking algorithms, to investigate the influence of urban areas on warm-season thunderstorms in Nanjing, China. Our findings reveal divergent urban impacts on the structure and evolution properties of thunderstorms based on different pre-storm synoptic conditions. When the synoptic conditions are weak, the urban heat island effect dominates, enhancing convective activities over urban areas. This leads to a reduction in the number of storm cells but an expansion in spatial coverage, ultimately resulting in increased rainfall over downtown areas. Conversely, when the synoptic conditions are strong, the urban canopy effect becomes prominent, slowing down storm movement and increasing the frequency of small storm cells over urban regions. These storm cells exhibit distinct "sharp" structures in terms of rainfall distribution and tend to intensify over downwind areas due to moist convergence. As a result, both downtown and downwind regions experience enhanced rainfall. This study improves our understanding of urban rainfall modification and offers valuable insights for storm nowcasting algorithms and the design of urban-specific rainfall events.