Abstract: This study leverages the WRF mesoscale meteorological model to simulate summer urban meteorological conditions in Chengdu, using four different combinations of land surface types and urban canopy schemes. Specifically, it incorporates 2021 Chengdu Local Climate Zone (LCZ) land cover data and traditional MODIS land cover data alongside a single-layer Urban Canopy Model (UCM) and multilayer Building Environment Parameterization (BEP). The simulations are evaluated using meteorological station data across various LCZ surface types. Key findings reveal that temperature is more sensitive to changes in land cover data, while wind speed is particularly responsive to alterations in urban canopy schemes. The combination of LCZ surface data and the BEP model closely aligns with observed diurnal variations in urban temperature and wind speed, accurately capturing the spatial distribution of decreasing temperature and increasing wind speed from Chengdu’s urban core to its outskirts. The BEP model’s detailed representation of the urban canopy significantly improves upon the UCM model, particularly in correcting the overestimation of nighttime temperatures and wind speeds, leading to more accurate predictions. Moreover, LCZ surface data substantially enhances temperature simulations in compact low-rise, open mid-rise, and open low-rise areas and reduces overestimated wind speeds found with traditional land surface data, especially in densely built-up areas. These results suggest that the LCZ classification effectively complements traditional urban land surface categorizations by offering detailed building morphology and enhanced surface roughness within the urban canopy. Incorporating LCZ-based surface data with the BEP model in WRF markedly enhances the accuracy of Chengdu’s summer meteorological simulations, offering valuable technical support for urban climate forecasting and air quality prediction.