Study On The Methods Of Relationship Of Urban Thermal Environment And Vegetation Cover

Urban thermal environment results from many factors such as land use/land cover, the artificial heat sources comes from energy consuming,weather condition in urbanized areas. Many coastal cities in China have launched into the fast development stage in recent years,and the natural surface is continuously substituted by urban artificial surface. The heat effect in the urban region becomes more evident. The continuous deterioration of urban thermal environment has become one of the important factors that affect the sustainable development of urban society and economy. Therefore, the quantitative relationship between urban thermal environment and vegetation cover is focused.There are many kinds of methods to study the quantitative relationship between urban thermal environment and urban vegetation cover. The thesis studies many aspects, for example ,urban thermal environment, urban land use/land cover, the quantitative relationship between urban thermal environment and urban vegetation cover. The use of different methods of empirical research is carried out and compared in the same study area. The thesis provides the reference for selected methods for theoretical research and practical planning.The thesis introduces the process of inversion of land surface temperature. According to the brightness temperature computing model,planetary brightness temperature(PBT) in study area is retrieved by digital numbers derived from the Landsat TM/ETM+ images,which is reflected urban thermal environment. The analysis of distribution character and pattern of urban thermal environment in study area indicates that the range of urban thermal environment has been expanding and its structure is more complicated, while it has evolved from centralized to dispersed distribution. The thesis introduces five kinds of classification of brightness temperature, ISODATA, supervised classification, a robust statistical method etc. The thesis divides and compares brightness temperature of the same study area through five different methods, analysis temporal and spatial characteristics of thermal environment. The result shows that the methods of classification, the number of classes and the indicators of classification are related to the characteristics of urban thermal environment.The thesis summarizes simply three methods, unsupervised classification, supervised classification and multi-step-classification method based on spectral characteristics for land cover. Three methods are applied and compared in the same study area. The results of classification of three methods represent multi-step-classification method can make good use of multi-spectral data. The extraction accuracy and the kappa index are the best in the three methods. The vegetation cover of study area is showed according to NDVI.Using different methods, the thesis analyses the quantitative relationship between urban thermal environment and vegetation cover from different aspects in the same study area. The author uses the classes of brightness temperature divided three methods. The author makes two quantitative models between brightness temperature and NDVI, brightness temperature and vegetation area. The results represents these models are better, the two groups of quantitative relationship are negatively correlated. The authors models brightness temperature and the area of vegetation patches in different temperature classes in the classic district. All the relationships are decreasing according to logarithmic relationship. The author attempts to statistic brightness temperature according to vegetation area. The author makes the model between the area of vegetation patches and brightness temperature, the model between the area of vegetation patches and NDVI. The models represent that increasing the area of vegetation patches can reduce brightness temperature. The rate of decreasing is reducing. Ultimately brightness temperature keeps a steady state. The relationship is complicate between the area of vegetation patches and NDVI.Compared with the previous researches, three innovative ideas are presented, including (1)studying many methods used in the relationship between urban thermal environment and vegetation cover, using and comparing of different methods of empirical research in the same study area. Comparing the results of brightness temperature classification applying different methods.(2)researching the relationship between the area of vegetation patches and brightness temperature in high, normal and low temperature class districts, making models through cutting classical region. (3)trying to statistic brightness temperature according to the area of vegetation patches, making the quantitative relationship model between the area of vegetation patches and brightness temperature, the quantitative relationship model between the area of vegetation patches and NDVI.