| The process of urbanization has been speed up after China’s reform and opening up over 30 years. The land problems have been continued to arise when economic development made tremendous achievements. For a long time, local governments formed the land use pattern of "earning money base on lands" which have main features are rapid expansion of urban land, declining abundant quality cultivated land, garden land, woodland,etc, the action are steadily deteriorating the ecological environment and living condition,and wasting and lying land resources. Currently, many cities development are faced with the plight of land resources binding constraint. The new situation demands that we should change land use pattern previously while strengthen the research of predicting, managing and controlling effectively for land use risks, At the same time, the simulation and early warning technology are particularly important and urgent for us to use and manage the precious land resources. The simulation is a process of recurring real things, the central idea is to demonstrat key featrues of the abstract system, the key issues including access effective information, and select key features and performances, similar simplification and assumptions application, and repeatability and validity of the simulation; In geography field we often use mathematical modeling approach to describe the structure and behavior of one geographic system, the models usually have two kinds of static model and dynamic model. The early warning is a pre-release alarm system which can identify, analyze and forecast potential problems and risks, take the necessary and advanced measures to prevent and eliminate them, the goal is minimized probable losses. It can be said that combinating simulation and early waning technology is one of the most economic and easy way to control and prevent urban land use risks or crisis.The paper selected Shenzhen city as the study area which is the representative region of rapid urbanization, the research based on the GIS spatial analysis and the characteristics of urban development of study area, integrated using methods of correlation analysis, principal component analysis, regression analysis, decoupling analysis to research the characteristics, driving forces and economic decoupling process of urban land use change; and on the basis, integrating extended CA model with early warning technology to simulate and warn scenerios and alarm status of future urban land use change. Through this study, one hand, we can enhance the macro-control capability for government to rectify future urban land use; the other hand, it can save intensive valuable land resources to avoid extensive operate by forecasting and pre-controlling risks. The main research contents and conclusions are as follows:(1) Based on the analysis of land use in Shenzhen city from 1990 to 2008, first, we used transfer matrix, expansion intensity index, spatial variation methods to extract basic features of urban land use change in the study area; second, we used correlation analysis, principal component analysis, regression analysis to study of driving forces of urban land use change; last, we used decouoling theory to deep discuss the "positive and negative decouoling "relationship between economic growth and urban land use change at different times. The research result shows:the most important characteristics of urban land use change is increase of urban land and reduction of agricultural land or water in Shenzhen city from 1990 to 2008; the three major driving factors of economic development, population growth and labor force conversion, expansion of secondary and tertiary industry investment are affected urban land use change; at the same time, the research found that a "absolute decoupling" phenomenon has been occurred between economic growth and urban land expansion in Shenzhen city after 2004 year, the degree is reduced year after year that economic development is heavily undependent on the land resources, which benefit from adjusting and urgrading industrial structure.(2) The research selected CA model to simulate and early warn urban land use change in study area. But the standard CA model has some deficiencies which as iteration time or making conversion rule, etc. We need to modify and extend it for practical application. The research started from solving the defects of standard CA model, and proposed to build a constrained extended CA model framwork based on life mechanism. In order to build a extended CA model for simulating and early warning system of urban land use change, we proposed amount land areas to control CA itration time which predicted by Markov chain; use integrated cellular transformation probability which determined by cellular fitness calculation and domain space effection based on life mechanism to replace the complex transformation rules making. After built the extended CA model, we used C++ language to develop software system on VC.NET environment which integrated Markov, Logistic, CA and GIS technology, and applicated it to simulate urban land use change in Shenzhen city from 2010 to 2030. The research result shows:the CA model can sumulate and predict the evolution of urban space in the future, the model accuracy is more than 80%. The model also is the core of early warning system for urban land use.(3) Now, the main research areas of land use early warning are focus on the cultivated land and ecology environment early warning, meanwhile, the study of urban land early warning is still in theory phase, and had few applications of typical areas to provide local government some references for decision-making on land use management. The research try to construct the urban land use early warning system based on existing stydies, the system including early warning content, mechanisem, indicatro system and early warning model,etc. and used the extended CA model to multi-scale(regional scale and street scale) early warn urban land use change in Shenzhen city from 2008 to 2012. The research result shows:CA model is a effective method for urban land use early warning; and the value of early warning of urban land use change is in the level of light and middle range which can be controlled on the whole state in Shenzhen city from 2008 to 2012. |
PDF Full Text Request