Study on remote sensing retrieval of urban thermal environment and the temporal-spatial distribution in Beijing-capital zone

Urban surface temperature is higher than surrounding regions, which is associated with the differences between the urban surface and it surroundings and arising serious of environmental problem. In the paper, we retrieved land surface temperature of Beijing-capital zone by spit windows algorithm using Moderate Resolution Imaging Spectroradiometer (MODIS) image at I km resolution, which kept good similarity with LST product of MODIS, then we analyzed the temporal-spatial change dynamic using synthesis eight days LST product. It was found that the daytime UHI demonstrates distinctive seasonal variation, with the minimum 21 degrees C in all four seasons. The UHI effect of night is more serious than that of day time in Jan. and Jul.; while vice verse in Apr. and Oct. The UHI intensity rank is Apr., Jan., Oct., Jul. by decreasing in day time respectively, and Jan., Apr., Jul., Oct. in night time. The spatial autocorrelation analysis indicates that the distribution of thermal environment is spatial clustering of similar LST value. Distribution of UHI effect and the correlationship between landuse/landcover and UHI effect are analyzed. The conclusion shows that the urban heat island mainly results from the difference of the surface thermal characteristics between urban and rural area.