Study On Land Use Change Base On RS And GIS

Researches on Land Use Change now are the frontier and hot problems for discussion in the researches on the global environmental change. However, studing on the dynamic change cases of Land Use Change in typical areas are one of the most important ways to understand global environmental change. Urbanization is becoming faster and faster in China now.It is very important to study how to use the land under the rapid Urbanization.In the paper ,Xi'an city which is one of the biggest city in west of China, is selected as a typical area for this case study. The method used are remote image classification and model researches.Remote classification is the basis for land use change.The Remote imange used is Landsat TM/ETM in 1988,1997 and 2002. Three years land use situation maps are got by using supervised classification tools, process-based classification tools and object-oriented classification tools.Found the best classification by using geography, statistics, fractal theory and quantitative research methods,and finally found that process-based classification tools is the best method ,becase its overall classification accuracy and Kappa is better than other two.So the image classification by process-based classification tools are taken for study.By using 1988 and 1997 land use situation maps to predict land use forecast maps in 2002 with model method to compare the situation maps and forecast maps in 2002 to find the degrees of similarity and accuracy. The prediction model using includes theMarkov model and LCM models. Comprehensive the three moldes used to find that the Kappa between land use forecast maps and the situation maps in are accuracy which reaches 0.8328.