An evaluation of WRF urban canopy models over Bengaluru, India

Numerical modeling of urban climate is essential for understanding mechanisms of the urban heat island (UHI) phenomenon. However, models must first be evaluated to identify their limitations. Over India, the evaluation of the Weather Research and Forecasting (WRF) model is limited. Here, WRF was evaluated over Bengaluru, India diurnally during the dry and wet seasons. Simulations were performed for cases using no urban canopy model (No-UCM), the single-layer UCM (SLUCM), and the multi-layer UCM (MLUCM) with the Mellor-Yamada-Janjić and the Bougeault and Lacarrerè planetary boundary layer (PBL) schemes. The simulations were compared to land surface temperature (LST) observations from the Moderate Resolution Imaging Spectroradiometer for their biases in urban LST, non-urban LST, and UHI intensity. Absorbed shortwave radiation, net longwave radiation, sensible heat, and latent heat were investigated for their possible role in driving the LST biases since they are calculated differently depending on which UCM is used. Results show urban LST was more sensitive to UCM choice than PBL scheme and the use of an UCM reduced urban LST biases, leading to improved simulations of the UHI. Non-urban LST was insensitive to UCM and PBL choice. For the best case, urban LST was underestimated by less than 1 °C during the dry season day and night, and was overestimated by 1.88 °C and 0.08 °C in the wet season day and night. In general, the SLUCM had the least bias for urban LST and UHI intensity due to a near-zero latent heat flux in No-UCM, too much trapping of shortwave and longwave radiation by the MLUCM during daytime, and too much surface cooling at nighttime by the MLUCM. These results can inform future WRF studies that evaluate UHI mitigation strategies over Bengaluru on the best model physics to use.