Analyzing And Simulating The Spatial Pattern Of Land Use Change At Township Level

The study of land use change is developing from the macroscopic scale to the typical regional microcosmic scale.The study on the land use change characteristics and driving mechanism of small scale in typical areas is an important way to deeply understand the mechanism of macro land use change.In this study,Mapo and Niulanshan town was selected as the research object,which is the typical urban-rural fringe of Beijing.Based on the 2001 aerial remote sensing image and 2015 satellite remote sensing image,Geographic Information System(GIS)and geometric model were used to analyze the dynamic process and spatial structure of land use change in the study area from 2001 to 2015.Based on the Logistic regression model,the driving mechanism of land use change was quantitatively analyzed.The CLUE-S model was used to simulate the spatial pattern of land use in 2025,and predicted the trend of land use change in the future.Results indicated that:(1)The matrix landscape of Mapo and Niulanshan town were changed from cultivated land to construction land from 2001 to 2015,the transformation of cultivated land to construction land was the dominant characteristic of land use change in the study area.Overall,cultivated land,water area,unused land were decreasing,construction land and forest land were increasing,and the garden land was more stable.(2)There were significant spatial differences of land use change in Mapo and Niulanshan town.Cultivated land presented obvious gradient trend from the southeastern to the northwest in the study area,the trends were "rapid disappearance-slow disappearance-stable-rapid increase".There was a clear gathering area for the construction land expansion,and up to 62%of the village construction land was the increase state.The spatial change of forest land showed obvious polarization state,46.5%of the village forest land was the increase state,32.6%of the village forest land was the disappearance state.The spatial change of garden land was disorderly,and there was no obvious gathering area.The state of water area development was relatively stable.Unused land expansion was mostly appear in the edge of the study area.(3)As studied in the small scale,it was found that the explanatory power of the driving factors varied with the grid resolution,and the model fitting effect was optimal under 30m x 30m resolution;the land use change in Mapo and Niulanshan town was influenced by socioecononic factors and natural factors,but there were significant differences in the driving factors of different land use type.The comprehensive explanatory power of the driving factors on land use change from high to low as follows:water area,unused land,cultivated land,forest land,construction land,garden land.From the dominant driving factor,the location factors had a leading role in the change of construction land,garden land and unused land,socioeconomic factors had a leading role in the change of cultivated land,forest land and water area.The human factor survey proved the rationality of the current forest construction from the perspective of human demands.(4)Using CLUE-S model to simulate and predict dlstribution of land use in 2015,the kappa coefficient was 0.814,which showed satisfactory results.The model was applicable to multi-scenario prediction in the future land use spatial pattern.Finally,the research simulated and predicted three scenarios about natural development,food security,ecological security.The results showed that the urbanization trend of the study area was still intense,under the three scenarios,the land use change trend was mainly based on the expansion of construction land,main sources of expansion were the conversion of cultivated land.Over the past 15 years the construction land development scale of eastern region had gradually matured,the future direction of expansion was mainly in the central and western regions.