Chararcterizing Land Surface Temperature In Relation To Percent Impervious Surface,Landscape Metrics And Land Cover Types From Landsat TM/ETM+

With economic development, urbanization has been accelerated in recent years in the city of Fuzhou, FuJian province of China. Rapid change on land surface property and its patterns may lead to change of thermal properties in urban areas.One of the main impacts of rapid urbanization is the effect of urban thermal environment. The raising temperature not only consume more resources but also may hazardous to human health.We hope to provide reasonable city planning, theoretical basis and scientific guide for the sustainable development of urban ecosystem health by the analysis of spatial pattern, temporal variation and influences factor of urban thermal patterns.In this study, integrating quantitative remote sensing retrieval theories and methods and Spatial analysis of GIS, we studied Spatial distribution pattern of thermal environment in Fuzhou, relationship between land use/land cover and land surface temperature; we also introduced to study landscape metrics on thermal forcing and impervious surface area(ISA) viewed from the landscape ecology view. The method was shown below. Two Landsat TM/ETM+images acquired on June151989and March4,2001were used as data sources. The normalized difference vegetation index(NDVI), and impervious surface area(ISA),which was extracted by applying linear spectral mixture Analysis,were used to estimate land surface temperature (LST).Using threshold continuum approach and range approach to discretize percent ISA into different categories.landscape metrics such as cohesion,AI,LPI etc are calculated based on different FISA categories. The main find are followed:(1) Urban heat island mainly distributed in the downtown areas of Fuzhou, such as GuLou and TaiJiang.The trend of the Urban heat island scope is increasing.(2) The surface temperature was different with land-use types. The successive order of temperature was Bare land,building, grass, forest land and water. The main heat source of land-use types was building. The cooling effect of forest land and water were obvious, the effect of grass not obvious and had the tendency to decrease. (3) The areas increased in LS A were broadly in line with the trend of expansion of the city.There was a positive linear relationship between LST and impervious surface over the region,The correlation coefficient is0.66(1989) and0.71(2001).The temperature increased by0.76in1989s and0.71in2001s as every0.1increase of FISA.The surface temperature showed a negative correlation with NDVI.(4) Thermal field landscape patch density level was high from1989to2001. Especially in the areas where the high intensity of heat island. The greater the heat island, the landscape fragmentation was more obvious. The patch density of high temperature area and extremely high temperature area in1989are3.7to6.8times than1989.There was complex geometry and low similarity between thermal field patches.(5) With the increase of spatial extent,patch density reduce,heterogeneity was reduced, which made little feature combined. The shape of patch become more and more simple and the similarity was strong. The landscape metrics trend to stable as the increase of spatial extent. The more complex the shape of patch of FISA, the more intense the heat island effect. The FISA’s temperature was well correlated with the continuity of the patch, the temperature increase as the continuity of the patch strengthen.