Research on the impact of impervious surface area on urban heat island in Jiangsu Province

Land surface temperature (LST), vegetation index, and other surface characteristics that obtained from remote sensing data have been widely used to describe urban heat island (UHI) phenomenon, but through impervious surface area (ISA) to describe the phenomenon has only used in a few study areas in our country. In a high urbanization and high population density region like Jiangsu Province, a wide range of extraction of ISA to study its relationship with UHI is less. In this paper, we use multi-temporal remote sensing images as data sources, and extract ISA from it in a large-scale by using decision tree classifier (DTC) and linear spectral mixture analysis (LSMA). Then combine the average surface temperature from the sixth band of Landsat TM by mono-window algorithm for spatial analysis, to assess the change of the urban heat island temperature amplitude and its relationship with the urban development density, size and ecological environment. Finally we use statistical methods to analyze the relationship between ISA, LST and UHI. The results show that ISA has a positive correlation with surface temperature. The ratio of ISA is higher and the difference value of the temperature is larger, thus the UHI will be more obvious.