Land Use Changes, Driving Forces Mechanism And Simulate Model In The Development Coastal Region In China

Since the coastal region development tactic was going to further from 80th in 20 century, society and economy was development in a high speed in this region. At the same time, land use and land cover changed very rapidly. Zhejiang province, lie in the east coastal region of China, became one of the regions that were in the rapidest development speed. In these twenty years, urban sprawled extensively and crop land decreased quickly, so crop land became more and more scarcely. Besides that, environment deterioration and crop land downgrading were limiting the harmonious and rapid development of economy. So study on the land use changes is urgently. It is useful to using land rationally, optimizing land resources. It can provide scientific foundations for sustainable land using.Study area is Zhejiang coastal region. The remote sensing (RS), geography information system (GIS) and global position system (GPS) technologies are use to detect the land use changes in 1985, 1993 and 2001, the precise land use maps were obtained by TM/ETM+ images interpretation . The spatial analysis method of GIS and the statistical method were used to analysis the drivers of the land use changes. The driving factors include the social, economic, demographical, technological and policy factors. The model, based on the spatial characteristic and cellular automata was introduced into study area. The applications were improved to forecast the land use changes.The main conclusions are the followings.Land use changes were detected. Crop land was decreased at large extent from 1985 to 2001. The total area decreased was 277600 ha among sixteen years. Urban area was the other one changed distinctively, and increased by 211009 ha among sixteen years. Garden plot, forest and water areas changed a little, and all increased by small areas. Mudflat area was increased in someextent. All kinds of land use were changed more from 1993 to 2001 than from 1985 to 2001. According to zonal analysis, Hangzhou is the zone that faced with the most quickly changing of land use. And Yongjia was the one that faced with the least changes of land use.The relationship between land use changes and driving forces were investigated.(1) Driving forces on nature. The distance from urban center had the most influence to land use changes, and transportation had the more influence, then the elevation and slop are the lest influence to land use changes. Four levels' transportation had different affection on land use changes. Provincial highway had strongest affection, national highway had stronger, railway and speedway had weaker.(2) Driving forces on human being. Total population had closest relationship with areas of all kinds of land use. But with the social and economical development, the profits from per area of ground became the most important driver to land use changes. Land use policies were divided into three phases to explain the influence on land use changes. Land use changes were limited by policy before 1985. With the land use policy and land management changing, the land use changes were speedup from 1986 to 1993. Laws and regulations on land uses' interconversions were established soundly after 1994, then the land use changed quickly.(3 ) The relationships between driving forces on nature and on human being. The natural driving force is the basement of occurring on land use changes, and geography location is the fundamental condition of the land use changes. Driving forces on human being affect the speed of land use changes. Land use changes in different way by strong influence on these driving forces when social and economic is developing rapidly. And land policies and management measurements push land use changes in special way.SLfiUTH model, based on cellular automata principle, was introduced to simulate the land use changes. After technical developed and application improved, Zhejiang land use changes were simulated. According to error matrix, accuracy of the simulation results in nine years was above 80%, and in sixteen years was above 74%. After control coefficients of the model were adjusted further, accuracy of cropland area and urban area simulating in most of counties were above 90%.