Study On The Spatial-temporal Characteristics Of Heat Island In Hefei And Air Temperature Inversion Based On Remote Sensing Data

In the context of rapid progress of urbanization,the urban heat environment was one of the main factors that dominate the urban environment.The rapid advance of urbanization made the urban thermal environment highlighted.The warming effect of impervious surface is one of the important factors leading to environmental problems of urban thermal.Up to now,research about the optimal coverage of impervious of urban is insufficient,few studies on the warming effect of impervious surfaces.Air temperature was monitored according to meteorological observation techniques until now,obtaining the high-resolution air temperature has always been one of the problems of research and application.Existing literatures all research on the conversion relationship between temperature and air temperature,a variety of inversion algorithms for land surface temperature inversion only.Study on air temperature inversion has not been reported yet.Regarding the issue above,this paper taking Hefei as the study area,analyzed the time and spatial pattern in heat island effect of Hefei city by using remote sensing data of Landsat-8.Estimation of the impervious surface by using remote sensing data,studied on the warming effects of impervious surface combine with meteorological observation data.Get the optimal threshold of coverage of urban impervious surface.Take lead to air temperature retrieval at 1.5m height by using remote sensing data,and output air temperature spatial distribution,then try to explore the feasibility of the temperature retrieval technique.Obtained the spatial distribution of urban air temperature in the city base on conversion model between land surface temperature and air temperature,carry on air temperature revised by using meteorological data.Breaking on the constraints of traditional air temperature interpolation in geo-statistical methods,air temperature distribution was achieved on transition space from large-scale to small-scale,and break on the limitations of practical application.Results showed that:(1)The heat island effect in Hefei in all year around,its intensity is strongest in summer and weakest in spring and winter.The heat island areas concentrate in villages,towns and bare areas in winter and concentrate in main urban areas in other seasons with weaker intensity.The urban heat island was closely related to the underlying surface,and the heat island effect mainly appears in impervious surface and bare areas,its heat-island intensity larger.The urban heat island distribution has negative correlation with normalized difference vegetation index,and has positive correlation with normalized difference building index.(2)The impervious surface coverage distribution has an exponential correlation with land surface temperature,with the increase of impervious surface coverage,land surface temperature exponential growth.The impervious surface coverage distribution has a linear relationship with air temperature.This paper proposed a new index,modified normalized difference impervious surface index(MNDISI),for estimating impervious surface and the overall accuracy can reach 86.1%.With the increase of impervious surface coverage,land surface temperature rises faster than the rate of rise in air temperature.When impervious surface coverage reached at 43%,impervious surface warming effect and other surface feature's cooling effect arrived in a dynamic equilibrium.When the area of impervious surface coverage less than 43%,and the cooling effect of other land surface integrated plays a dominant role,so the surface temperature rises slowly.The optimal impervious surface coverage of the city should be less than 43%,the urban heat island effect should be alleviated with the improvement of urban surface pattern,such as increase the proportion of vegetation and water.(3)It was feasible to inversion of air temperature by using remote sensing data.This paper presents a method in divided underlying surface of city based on NDVI index,this method can be applied to air temperature inversion,the division method was universal and the result of the division is closer to the land surface in actual.Then obtain the conversion models between land surface temperature and air temperature in different underlying surface based on the method.Studied on the inversion of air temperature at 1.5m height according to the models.The inversion results show that the average absolute error of air temperature was 0.88 ?,and there were 62.93% data error were less than 1 ?.