Improved simulation of very heavy rainfall events by incorporating WUDAPT urban land use/land cover in WRF

A World Urban Data Analysis and Portal Tool (WUDAPT) based Local Climate Zone (LCZ) classification and mapping is done for Mumbai, India. Prior WUDAPT-based urban studies have primarily focused on urban heat island (UHI) assessment, and this is the first study to assess the impact on rainfall simulations. Accordingly, in this study, the impact of incorporating the LCZ map in the Weather Research and Forecasting (WRF) model for simulating very heavy rainfall events over Mumbai is assessed. The classified LCZ map is incorporated in the WRF model, and the model performance is compared against the Control, for four recent very heavy rain events. Comparing the results from paired simulations reveals that the rainfall amount and spatial variability is significantly different when the LCZ framework is used. Implementation of WUDAPT data, introduces heterogeneity in the morphological characteristics of the city, which contributes to microscale feedbacks that appear to organize and impact the mesoscale convergence/divergence and convection fields in the WRF model. The study results show variability within the cases highlighting the need for additional investigations in different cities. Overall, the incorporation of WUDAPT LCZs in WRF results in notably improving the model performance for heavy rainfall simulation in urban areas. © 2020 Elsevier B.V.