Effects of urban warming on surface temperature: Integrating the boosted regression tree approach and regional warming sensitivity index

The urban heat island effect has become a prominent urban environmental issue in the process of urbanization. However, attempts to differentiate regional sensitivities in the process of surface warming are rare. In this study, we designed the regional warming sensitivity index by adopting the boosted regression tree (BRT) approach. With this index, we quantified regional sensitivity differences in the process of surface warming. The results show that: 1) the warming sensitivities of Shanghai's jurisdictions in descending order are Huangpu District, Yangpu District, Hongkou District, Jing' an District, Putuo District, Xuhui District, Changning District, Baoshan District, Minhang District, Pudong New District, Jiading District, Fengxian District, Songjiang District, Qingpu District, Jinshan District, and Chongming District. 2) The difference of regional warming sensitivity is obvious under different temperature gradients, and the regional variability of temperature changes is large under the highest temperature gradient. 3) The warming sensitivity of land covers has obvious regional difference. 4) The regional warming sensitivity index based on the BRT method has a good indicating significance and high sensitivity. The exploration of regional warming sensitivity could provide scientific bases for regulating the urban thermal environment in deteriorating urban thermal environments.