Land Use Change Analysis And Study On Land Use Prediction Method In Beijing

Economic development and population growth affect the utilization of land resources.In the study of global environmental change,land use change is an important field.In previous studies,scholars mainly studied changes in the structure and quantity of land use,and in recent years,the scholars constantly shifted to the study of spatial distribution of land use change.The simulation of the spatial distribution of future land use by coupling of several mathematical models is also the focus of the research..In this paper,the study area is Beijing,the three year land use data is extracted and the dynamic change analysis is carried out to find the main factors that lead to the change.Thus the model is constructed to predict the spatial pattern of land use in 2020.The conclusions are as follows:(1)according to the national land use classification standard,it can be divided into six categories: grassland,Cultivated land,forest land,water and wetland,built-up land and unutilized land.Using object-oriented method,Landsat is used as data source to classify the three period data of 2005/2010/2015.According to the vector interpreted by the project team,the accuracy analysis and manual revision are carried out,The final data accuracy can meet the requirements of this paper.(2)to analyze the dynamic change of land use in Beijing,used the range of land use changes,the speed of land use change and the spatial change of land use.Through analysis,it is found that the construction land expanded to the periphery of the city center,the unutilized land is greatly reduced,the cultivated land is reduced,and the water and wetland land are not greatly reduced.the range of land use decreased in turn from unutilized land,cultivated land,construction land,grassland,water area and wetland and forest land.The largest single type of dynamic change is the unutilized land.The annual comprehensive land use dynamics of Beijing city in 2005-2010 and 2010-2015 years are 0.35 and 0.38 respectively,which shows that the city of Beijing has been developing at a smooth speed in the last ten years.(3)select the elevation,the slope,the population density,the per capita GDP,the distance from the administration center,the distance from the river,the distance from the highway,seven factors,and add the land use neighborhood information to construct the Markov-ANN-CA coupling model,and use the model to simulate the spatial distribution of land use in 2015 and comparison with data of classification in 2015.the overall accuracy is 96.81% and the KAPPA coefficient is 0.9189,which verifies the reliability of the model.And this model is used to predict the land use in 2020,so as to provide a basis for revising relevant plans.