Improving WRF-BEP+BEM performance in simulating wind-temperature-humidity in high density urban areas: A case study of a megacity

This study aimed to improve the performance of the Weather Research and Forecasting model coupled with the Building Effect Parameterization and Building Energy Model (WRF-BEP+BEM) in simulating air temperature, specific humidity, and wind speed within the urban canopy. To achieve this goal, WRF experiments were conducted over Xi'an City, a megacity in northwestern China, and the results were compared with the observations from 25 meteorological stations located in urban areas. Three key findings emerged from this study. First, the WRF-BEP+BEM was highly sensitive in locations where the local climate zone (LCZ) type in the model grid differed from the real one, leading to additional temperature errors. Increasing the grid resolution effectively reduced such errors and improved the simulation of the inter-urban heterogeneity of temperature. Second, the current WRF-BEP+BEM was found to underestimate the dynamic-thermal effects of vegetation due to its use of the “Tile approach.” In this study, the dynamic-thermal effects of grass and trees on the urban canopy were parameterized following an integrated approach, and the root mean square errors (RMSEs) of air temperature, specific humidity, and wind speed were reduced by 0.44 ℃, 0.45 g/kg, and 0.09 m/s, respectively. Finally, the dynamic parameters of vegetation in the land surface model were found to be inappropriate, leading to large wind speed errors in natural LCZs in urban areas. In this study, the zero plane displacement height and momentum roughness length in the land surface model were modified based on the results of scenario large eddy simulations, and the RMSE of wind speed in natural LCZs was reduced by 1.31 m/s.