LAUSR

Abstract Urbanization can induce land cover changes that impact land surface temperature (LST). Many factors can influence the magnitude of urban heat, such as vegetation and aerosols. This work uses linear correlation, composite analysis, multiple linear regression, and random forest to determine the leading controls on urban LST of Bengaluru, India in the dry and wet seasons during daytime and nighttime from 2003–2018 using data from the MODerate Resolution Imaging Spectroradiometer and the European Centre for Medium-Range Forecasts ERA5 reanalysis. Results show that for the dry and wet season daytime, vegetation was the leading factor (linear correlation R=–0.74 and R=–0.34 with urban LST) since reduced vegetation limits evaporative cooling. For the dry season nighttime, vegetation was the leading factor (R=–0.52). Limited evaporative cooling during daytime can increase surface heat retention at night. For the wet season nighttime, specific humidity was the leading factor (R=0.21) since increased water vapor enhances downward longwave radiation and warms the surface. Therefore, urban heat is primarily controlled by vegetation in Bengaluru. However, since vegetation and specific humidity are related, mitigation strategies that increase vegetation must not increase water vapor substantially, otherwise urban heat may amplify during the wet season nighttime.