Model Sensitivity in Predicting Extreme Precipitation Event over Beijing

Understanding and forecasting the spatial and temporal distributions of extreme precipitation in urban areas are crucial for effective planning and mitigation efforts. In the present study, we examined the hindcast of an extreme precipitation event during 21-22 July 2012 in the greater Beijing metropolitan area. The primary focus was assessing its sensitivity to two widely used PBL parameterizations (MYJ and YSU), two urban parameterizations (SLUCM and BEP_BEM), two different land-use and land-cover (LULC) datasets for Beijing and different initial conditions. The results show that MYJ PBL scheme performs comparatively better than YSU PBL scheme in capturing near-surface air temperature and the location and timing of the heaviest precipitation. The LULC impacted the spatial distribution of precipitation but its effect on the amount of precipitation was minimal. Overall, using a combination of the MYJ PBL scheme, SLUCM urban parameterization, and locally-enhanced Beijing LULC, and initializing the model simulations at 0000 UTC July 20, 2012, demonstrated superior performance in capturing precipitation levels, despite some spatial discrepancies in the precipitation distribution. The performance of BEP_BEM urban parameterization is similar to SLUCM across various factors such as average rain rate, max rain rate, and rain volume.