| The rapid advance of urbanization brings the progress and prosperity of modern China and also causes serious environmental problems at the same time.The city's rapid development caused great changes of the structure and type in the underlaying surface,and the resulting effect of urban heat island has become a problem in the field of interdisciplinary across.Remote sensing as an effective way for getting surface information can obtain the surface temperature through the relevant algorithm,which provides surface information of the accuracy in time and space for the research on urban heat island and meets the dynamic observation analysis and requirements of urban heat island in time and space.The retrieval of component temperature abandoned the ideal assumptions of homogeneous pixel and tried to use a different model to describe the complex structure of the surface scale in sub-pixels,which is closer to the real condition of natural surface and more accurate for the study of urban heat island.The study is in the area of Nanjing in Jiangsu province.Nanjing as the central city in the political,economic and cultural aspects of the Yangtze river delta region developed particularly rapidly in recent 20 years.With the advancement of economic development,the city scale continuously outward expanded and heat island effects changed obviously,as a result,choosing Nanjing has great significance and research value in the study area of the heat island.In this study the author chose four phases of the Landsat images as the main data source,which used the algorithm of decomposition of mixed pixels to carry on Surface temperature inversion based on Landsat images,compare the inversion results for the accuracy evaluation and further analyze related links of impervious layer and the surface temperature with the aid of statistical methods.To analyze Nanjing urban heat island is based on component temperature.In the changes of the heat island spatial pattern,the landscape ecology method has been used.Through the calculation of landscape index and aspects of plaque type level and landscape level,the evolution characteristics of the landscape pattern of Nanjing heat island landscape index in recent 20 years has been analyzed,thus achieved the following results and understanding:(1)Based on the decomposition of mixed pixels algorithm,the surface temperature of the study area was inversed to obtain three sets of temperature,and the results of surface composition temperature were compared and analyzed,also,the accuracy was verified.The results show that the component temperature can show the characteristics of surface temperature more accurately.(2)We used impervious layer temperature to carry out statistics and research on the central urban area of Nanjing city,and studied the distribution pattern and changing characteristics of urban heat island through temperature difference classification method in software.The results show that the closer the urban center is,the larger the area of the hot island area.The farther away from the center of the city,the larger the low temperature zone and the sub-central temperature zone,the increasing area of the hot island in Nanjing in the last 20 years,which shows an increasing trend(3)The study uses ecology knowledge based on landscape index to analysis the urban thermal environment.The results indicate that an upward trend in total heat island landscape patches and density.The area,proportion,number of plaques,density and the average area of individual plaques in the high-grade heat island type patch tended to rise,and the shape of the whole hot island landscape tended to be complex,but the distribution of each type in the area of the heat island landscape gradually tended to be even. |
PDF Full Text Request