Analysis of thermal environment and urban heat island using remotely sensed imagery over the north and south slope of the Qinling Mountain, China

Aim to acquire the spatial characteristics of the thermal environment and urban heat island, near-surface air temperature is retrieved from Landsat ETM+ image using the BP neural network algorithm in support of observed meteorological data in this paper. Surface temperature and Evapotranspiration were retrieved from Landsat ETM+ data using the mono-windows algorithm and the air temperaturebased algorithm respectively. Then the relationships of near surface air temperature, surface temperature, Evapotranspiration and NDVI are utilized to analyze thermal environment of different land cover classes in the both sides of the Qinling Mountain and the urban heat island of the Xi’an city. Meanwhile semivariogram and histogram could be used to express the spatial heterogeneity and differences of thermal environmental factors in the representative windows selected from the northern and southern slope of the mountain respectively. From systemic analysis, it can be seen that there is significant meaning to understand the inner mechanism of thermal environmental factors’ differences in the both sides of the mountain by analyzing the spatial information of near-surface air temperature, surface temperature, Evapotranspiration and NDVI from remote sensing image, as well as the urban heat island effect of the Xi’an city.