Simulation Of Land Use Change Based On CLUE-S Model And GIS

Land-Use and Land-Cover Change (LUCC) is a current hot topic in researching global environment change. With the development of the social economy and the progress of technological, research direction and content of LUCC is continuously deepening and expanding. In consideration of China's huge population base and relatively limited land resources, the research of LUCC is of more practical significance. One of the core content of LUCC research is analysis and simulation of regional land use change.Wuzhong County, Suzhou City as the study area in this paper, firstly, based on the data of land use change for years, gray model is applied to predict the demand for land use types of future years. Then, logistic model is used to calculate the correlation between type of land use of study area and nature, economic society drive factor, and the best scale of land use change simulation in the study area is gotten through the ROC test. Afterwards based on the land use data of2005, by inputting above-mentioned research findings into the CLUE-S model, the spatial distribution pattern of land use in the study area in2009can be simulated and map of actual land use in2009is used for validation. Finally, the map of land use spatial pattern distribution of Wuzhong County in2020is simulated.The main conclusions include:(1) For the non-spatial analysis module in the CLUE-S model, applying the gray model for quantitative prediction of land use structure in the study area of the target years, predictions of area data of each land utilization type are more reliable. The data of land use type is from the land alteration data in Wuzhong County. The use of existing built land use database can reduce the data errors from the interpretation of remote sensing image.(2) The regression analysis of correlation between various kinds of spatial, non-spatial driving factors and land use type is the core of simulation by CLUE-S model. It makes the effects of a variety of driving factors on the spatial distribution to be realized. In this paper, the ROC test results of Logistic regression analysis shows that simulation results are ideal: farmland0.939, Corner0.832, woodland0.978, other agricultural land0.896, land for construction of0.888, water area0.982, all degrees of fitting are greater than0.8. It shows that there is a good relevance between the selected driving factors and land use types.(3) According to the logistic regression model, comparing pros and cons of different analog scales, it can be showed that there are some differences in β coefficient of different analog scale. After comparison,200*200m can be the best scale of constructing simulation of spatial distribution pattern of land use in the study area.(4) In view of the results, the transformation and spatial distribution of main land types including cultivated land, garden plots, forest land, other agricultural land, land for construction and so on accords with the developing trends of economic society and arrangement in urban and rural areas.