Research On Simulation And Forecast Of Construction Land-use Change Based On CA_Markov Model

Sustainable land use is the basis of sustainable social development strategy, which attaches great importance to the change of construction land. Because of the openness, nonlinearity, uncertainty and dynamic natures, the change of construction land in quantity and space is hard to be described accurately by the traditional methods. The Markov model has been widely applied in predictive modeling on land use change, and also the cellular automata (CA) model with powerful spatial operation ability is a good choice. So in this paper, the combination of these two models can demonstrate their effective ability on simulating the spatial variation of the construction land change system.The Li Shan town which is the center of the Sui County has been taken as the research object in this paper. Combining the ArcGIS9.3, IDRISI17.0and the transfer matrix method to build CA_Markov model, the evolution pattern of Li Shan town’s construction land can be founded. Then, the short-term expanse pattern of the construction land can be forecasted. Based on these analyses of the evolution pattern and the prediction result, the scientific and rational land use planning can be developed which is suitable for the development of Li Shan.This paper mainly includes the following researches:At first, the suitable image set of construction land conversion has been defined. The land use suitable image set is composed of the suitable images of construction land change and the probable images of other kinds of land use. It was used to improve the cellular automata theory transformation rules, and make the simulation results close to reality. This paper selected the basic farmland protection areas, the waters and the existing construction land as restrictive conditions, and formulated the six constraint factors including transportation distance factor, the waters distance factor and so on as transformation conditions in operation of CA_Markov.Secondly, this paper validated the accuracy of the CAMarkov model. It verified the accuracy of the CAMarkov model in three aspects including the quantity error, cell size and cell distribution. The prediction research of construction land distribution in2015was created through choosing the suitable scale of data layer measured30m*30m. In terms of number error, this paper predicts the data of201l,and compares the obtained results with the present situations. It shows that the errors of all kinds of land use are not more than5%. Even the error of the traffic land, construction land and the waters does not exceed1%. Moreover, in the aspect of spatial distribution, this paper adopts the method of spatial superimposed analysis. It predicted the land use distribution map of2011, and superimposed the result on the actual land use distribution map of2010. The superposition results show that accuracy rate can achieve95.8%.At last, this paper analyzed the construction land’s quantity change and spatial distribution of Li Shan town. The prediction results demonstrated that the number of construction area of the Li Shan town has been increased progressively in the study period. At the end of2015, the construction land of Li Shan will reach2000hectares. On the point of conversion type, the new construction land is composed of cultivated land, grassland, garden land and bare land; on the point of spatial distribution, the new construction land are mainly distributed in the center towns and both sides of major traffic roads in the middle south area. These changes are basically same as the changes from2009to2011.