Automatic Detection and Dynamic Analysis of Urban Heat Islands Based on Landsat Images

Given rapid global urban development, increases to impervious surfaces, population growth, building construction, and energy consumption result in the heat island (UHI) phenomenon. However, spatial extent of UHIs is not clearly mapped many UHI studies based on a remote sensing approach. Therefore, we developed method extract during period from 2000 2021 Nanjing, China, explored impact two- three-dimensional expansion intensity. After cropland effects (i.e., bare soil) were eliminated, our proposed combines Getis-Ord-Gi* standard deviation normalized difference vegetation index (NDVI STD) area Landsat 5 8 images using land surface temperature (LST) autocorrelation characteristics seasonal variation vegetation. Our results show following: (1) Bare farmland has large influence extraction UHI—combined with NDVI STD, soil was highly reduced, strongly improving accuracy extraction. (2) The dynamics are consistent changes built-up Nanjing at both temporal scales, but increase green ratio, mature areas trends decrease due cooling effect space. (3) accumulation GDP promote vertical buildings. When two-dimensional city reaches saturation, intensity primarily affected by expansion.