Land Surface Temperature And Spatiotemporal Patterns Of Surface Urban Heat Island Retrieving From Remotely-sensed Imageries

In this thesis, Landsat TM/ETM+imageries acquired in 1997,2000 and 2002 covering Shanghai metropolitan regions were used to retrieve land surface temperature, to derive impervious surface area (ISA), and normalized difference vegetation index (NDVI) by utilizing spectral mixture analysis (SMA) approach to analyze spatiotemporal patterns of urban surface heat island and the relationships among LST and NDVI and ISA under different urban development density. My main findings are as following:(1) The urban heat island exhibited an increasing trend with urbanization degree in Shanghai metropolitan regions from 1997 to 2002. The urbanization has resulted in continuous decrease of area with lower LST and increase of area with higher LST. Multiple areas in high LST were monitored demonstrating that rapid urbanization has resulted in multi-core urban surface heat islands patterns in Shanghai metropolitan regions.(2) Population and urban landscape dynamics are the main driving forces of LST changes in Shanghai metropolitan regions. The distribution of surface UHI is spatially consists with those of dense population and high percent of impervious surface area. Urban vegetation can effectively mitigate UHI in Shanghai.(3) Urban development density can influence NDVI and UHI. Land surface temperature increases and has large variations with percent ISA, while NDVI decreases with percent ISA.