RS-GIS based landscape pattern change and its effect on urban thermal environment in main urban area of Kunming, China

The urban eco-environmental effect was generated in the process of urbanization. Urban thermal environment is a comprehensive embodiment of environmental effect of urban landscape ecosystem. Landsat TM/ETM+ data, Modis data and the real-time meteorological data were used to study the effect that landscape changes caused to the urban thermal environment. Fragstats, a famous landscape metrics calculation software, was used to calculate the landscape metrics in the class-level and the landscape-level. Erdas Imagine, a famous RS soft has been used to build model to extract land surface temperature. Modis data and meteorological data at the satellite pass were used to testify the validity of Landsat TM/ETM+ Band6. Annual variation and seasonal variation of thermal environment were revealed. The result showed that the building area increased quickly through occupying land. As a result, land deceased sharply during year 1992 to 2008. The building area changed from 5163ha in year 1992 to 11425ha in 2008. The urban expansion also occupied the forest and grassland. There was 1123ha grass land in 1992, but it couldn't be identified on the Landsat TM images in 2008. Through three periods in 1992, 2000 and 2008, NP(number of patches) was 1283, 1952, 2566 respectively. The largest patch type changed from land to building area and LPI (the largest patch index) increased from 29% to 66%. Six higher LST areas and two lower LST areas were found on 4 Landsat images in year 2000 to 2001. Meanwhile, five same higher temperature areas which weakened at some extent and two new areas with higher LST were founded on Landsat image in 2008. The biggest high temperature area was located in Liangting Village and the higher temperature area of building area moved outside gradually from year 1992 to 2008