Simulation Studies On Land Use And Land Cover Change In Ertok Banner Based On GIS And RS

At present, land use change has become a hot spot in the international academic community. Land use change model is an important way to explore the future land use change and the effects of land use change on environment. Meanwhile, to understand the interrelationship between land use change and factors of impact on it is the basis of the model building. Currently, it is still difficult to construct land use pattern formation and evolution of mechanism model combining with natural environment factors and socio-economic factors. As a result, experiences statistical model has been playing an important role in this field.This paper takes the Ertok Banner, which is located in western Ordos Plateau in Inner Mongolia as an example, to analysis its land use pattern formation and driving forces mechanism of changing in a small-scale spatial extent, using CLUE-S model to simulate the land use change of the study area, which is exploited on the basis of the Logistic regression model and is used to forecast the space module of the land use pattern evolution and the non- space module of the land use demand. The conclusions of this research are as follows:(1) First, the paper explains the concept, research contents, research progress of the land use and the land cover, and application of RS and GIS in this field. Meanwhile, introduces the actuality of the driving factors of land use. On the basis of these, the author introduces several land use change models that are more influential currently, and explains why to choose CLUE-S model.(2) Second, taking the Logistic regression model as a framework, on the basis of the land cover data of the study area in 1990 which is obtained using RS and GIS technology, in combination with the nature influence factors including soil erosion, grade, slope direction, soil texture, temperature, precipitation and so on ,as well as the socio-economic factors which include the population, towns, roads, livestock etc, the author sets up the Ertok Banner's simulation model of the land use space pattern, and confirms the main driving factors of restricting all types of land change in this study area.To ensure the accuracy of model, the paper designs a variable based on the result that whether a certain land type will appear in 1500m×1500m grid cells. By dealing with the nature control factor sequentially, and inputting data and operating to the social economic factors and the interannual changes in meteorological factors to provide the ample candidate variable for the logistic regression analysis, and though the strict collinearity diagnosis and filter variable to pick up more meaningful factors. In order to improve the model's explain ability, this paper adopts determinant based on the classification form. The result of each land type regression analysis shows: the overall correct classification rate of each land type entirely reaches more than 73%, so the author thinks that the driving factors which have been selected from regression equation have better explanation ability for each land type. This result has laid the foundation for further study of the land use changes imitation of the Ertok Banner.Based on the two periods (1990,2005) data of the land use pattern of Ertok Banner, the paper applies the land use/ coverage change simulation model—CLUE- S model which is currently extensive to be applied in international, and establishes the Ertok Banner LUCC model which has good space expression ability and strong forecast ability. Meanwhile, based on the land use data in 1990, this paper simulates the land use change pattern in 2006. The result shows that this model is more capable of simulating the Ertok Banner land use / cover change, kappa index up to 67%. So the author finally proves that as a better model, CLUE- S mode has the ability of simulating the regional land use time-space changes, so it is worth recommending it toward domestic counterparts.