| In recent years,the rapid urbanization process and economic development have led to the rapid expansion of urban space,which has brought about problems such as the prominent contradiction between land supply and demand,waste of land resources,degradation of ecosystem services,imbalance of land use structure,which have seriously restricted the comprehensive and sustainable development of cities and society in China.As the middle zone connecting urban and rural elements,the metropolitan suburban areas not only affect urban construction but also affect the development of the social economy.However,the expansion of urban space and intense human activities have led to the incredible dynamic and complexity of land use change in the suburban areas.Land use fragmentation,farmland loss,ecological environment damage and other issues become increasingly prominent.In the context of accelerating the urbanization strategy of "local and nearby" in China,ensuring food security and ecological security in the metropolitan suburban areas while meeting the needs of urban development has become an important issue to be solved urgently in the current territorial spatial planning.Therefore,it is of great practical significance to explore the characteristics and laws of land use change in the metropolitan suburban areas and forecast the future land use pattern,in order to guide the reasonable development of various types of land in these regions,realize the sustainable use of land resources and alleviate the contradiction between human and land.Land use change simulation model is an effective tool to explore land use change dynamics in the future.The research on simulation of urban land use expansion or land use change based on cellular automata(CA)and agent based model(ABM)has been widely carried out.However,most of the previous studies mainly focused on exploring the driving mechanism of land use change from natural and social economy,the influence of stakeholders’ decision-making behaviour on land use change needs to be detected deeply.Moreover,most of the existing ABM based land use models mainly focused on single land type,the driving mechanism and simulation of land use change for multiple land use type are insufficient.Therefore,based on the theories such as sustainable development and complex system,this study takes "driving mechanism-model building-scenario simulation" as logical thinking to simulate the land use change scenarios in the metropolitan suburban areas by using GIS,ABM model,CA model and scenario analysis comprehensively.And an empirical study was carried out in the suburban urban development area of Tongshan District,Xuzhou City,where the land use change is complex and drastic.Specifically,this paper carries out the following research: firstly,land use change characteristics and its driving mechanism analysis.From the aspects of land use structure,transfer matrix and land use change in different directions,the temporal and spatial evolution characteristics of land use in the study area from 2010 to 2018 are explained,and the Logistic regression model is used to analyze the driving mechanism of land use change.Secondly,construction and implementation of SD-SCALUS Land Use Change Simulation Model.Through coupling the CA Model,ABM model and adaptive inertial competition mechanism,a land use change simulation model(Land Use Simulation based on Self-adaptive inertia and competition mechanism,Cellular automata and Agent,SCALUS)based on adaptive inertial competition mechanism,CA and multi-agent is constructed.The model can not only reflect the competition and interaction between various land use types,but also fully reflect the interaction between land use spatial environment and various agents on land use change.Then,the system Dynamics(SD)model is established to simulate the land use structure,and the results are introduced into the SCALUS model as quantitative constraints to construct the SD-SCALUS model,the model is implemented based on the Netlogo platform and Vensim software.Thirdly,model verification and land use change scenario simulation.The land use data of 2010 and 2018 in the study area are used to calibrate and verify the accuracy of the above models.On this basis,the land use changes under the three scenarios of natural development,cultivated land protection and ecological control in 2030 are predicted and compared by setting different quantitative and spatial constraints.Through the above research,the following main research results are obtained:(1)The change of land use in the study area from 2010 to 2018 is complex,the change of land use quantity structure is evident,and the spatial differentiation of land use change is significant.Firstly,the area of urban construction land and other construction land increase,among which the urban construction land increases the fastest,and the new urban construction land mainly comes from cropland and rural residential land,while the area of other land types decrease.Secondly,the spatial distribution pattern of urban construction land,forest land and unused land are more stable,while other land types are prone to land type transformation.Moreover,there are significant differences in land use changes in different land types in different directions.The cropland in the southeast and south reduce the most dramatic,the water area increases in the west direction,no obvious change in the northeast direction,and a decrease in the other directions.The expansion of urban construction land is significant,and the spatial difference is noticeable,in which the southeast direction increases most sharply,followed by the east and south direction.The rural residential area and forest are reduced in different directions,and the other construction land is expanded the most in the north.There is no noticeable change in the unused land.(2)Topography,major roads,railways,major rivers,town centers,and rural roads are the main driving factors affecting the spatial distribution of land use in the study area.The ROC values of cropland,forest,water area,urban construction land,rural residential land,other construction land,and unused land are 0.766,0.835,0.750,0.960,0.643,0.609,0.817,respectively,indicating that the model has an excellent fitting effect.The CA layer conversion rules can be obtained according to the Logistic regression parameters.(3)Coupled with CA model,ABM model and adaptive inertia mechanism,a spatial simulation model of land use(SCALUS)are constructed to explore the changing process of land use pattern in the metropolitan suburban areas with rapid economic development caused by human-land interaction.The results show that in terms of simulation reliability,both visual comparison and precision comparison,the SCALUS model has higher goodness of fit than the Logistic-FLUS model and the FLUS model,and the overall FOM of the model is 0.172,which is 0.047 and 0.089 higher than the Logistic-FLUS model and the FLUS model,respectively.It shows that the land use simulation model based on SCALUS can better reflect the evolution process of land use pattern in the study area,reveal the influence of natural location,decision-making behavior of agents and policy planning factors on land use change,and provide a reliable simulation method for the metropolitan suburban areas to a certain extent.(4)The SD-SCALUS model is used to simulate the land use pattern of the study area under three scenarios: natural development,cultivated land protection and ecological control.The comparative analysis shows that urban construction land and other construction land continue to expand based on the original,and the bare farmland and ecological land can not be protected.Under cultivated land protection,the actual reduction speed of cultivated land is reduced,the amount of cultivated land is well guaranteed,the development speed of urban construction land and other construction land is slowed down,and the area of woodland and water area is significantly reduced.Under the ecological control scenario,the loss of woodland and water area is restrained,the stability of ecological land is increased,and the bare farmland is protected to the maximum extent. |
