Research On Dynamic Simulation Of Multi-scenario Land Use Change Based On CLUE-S Model

The process of land use change is extremely complicated,mainly because it is not only the result of a single factor in the process of change,but a complex process formed by the doping of multiple factors.With the rapid development of social economy,once the regional land use is unreasonable,it will easily affect the sustainable use of regional resources and the state of harmony and stability.In addition to the land use change simulation model,in addition to finding unreasonable conditions in the process of regional land use change,it can also simulate future land use changes,and further provide a reference basis for future regional land use changes.Taking Xinfu District of Xinzhou City as an example,this paper analyzes the land use changes in Xinfu District(2005,2016),and analyzes the land use changes in the region from 2005 to 2016 from three different perspectives.Ten driving factors were selected for the actual situation in the study area,and a logistic regression model was established to quantitatively analyze the relationship between land use types and driving factors.The CLUE-S model was used to simulate and predict the land use change in Xinfu District in 2016,and the accuracy was verified.After the verification,three different scenarios were constructed to simulate the land use pattern change in Xinfu District in 2025.Different landscape indexes were selected,and the simulation results were analyzed.Finally,corresponding suggestions were made for the future land use in Xinfu District based on the research results.conclusion as below:(1)The land use pattern of Xinfu District changed from 2005 to 2016.The main types of land use that have changed are urban and rural construction land,cultivated land,grassland,and forest land.The area of construction land has increased significantly.The total number of areas increased was 1620 hectares,mainly concentrated near the central urban area.(2)Select ten driving factors such as elevation,slope,and distance from the railway to establish a logistic regression model to reveal the quantitative relationship between the various categories and driving factors.The fit degree verification of this part of the model mainly uses the ROC curve method.The results show that the fitting degree of all classes in Xinfu District is greater than 0.7,which indicates that the selected driving factors have a strong ability to explain the land use types in Xinfu District.It also shows that the main driving factors for different land use types are different.(3)Based on the 2005 land use data of Xinfu District as a basis,the CLUE-S model was used to simulate and predict the land distribution pattern of Xinfu District in 2016,and the Kappa value accuracy test was performed on the simulation map and the actual land use status map.The kappa value obtained is0.9689,which indicates that the CLUE-S model can be used to simulate and predict the future spatial distribution pattern of land use in Xinfu District.(4)Construct three different scenarios of natural growth,food security,and ecological protection.Through adjustment of model parameters,the spatial distribution pattern of land use in Xinfu District based on different scenarios is obtained.The results show that the trend of land types in Xinfu District under the natural growth scenario is the same as that in 2005-2016.The area of urban and rural construction land is increasing,and the area where the construction land area is increasing is mainly concentrated near the central urban area.Under the food security scenario,the cultivated area of the region is strictly controlled.The cultivated area is concentrated and continuously distributed,and high-quality cultivated land resources in the study area are protected.Avoiding invasion of forest land and grassland while developing various land use types under ecological protection scenarios is extremely important for solving regional ecological problems.After comparing and analyzing different simulation results,four landscape pattern indexes,namely,Aggregation Index(AI),Contagion index(CONTAG),Shannon's Diversity Index(SHDI),and Shannon's evenness Index(SHEI),were selected to reflect the landscape under different scenarios.Simulation results.Under the scenario of food security,Aggregation Index and Contagion index is the highest,which indicates that the protection of cultivated land can guarantee the amount of cultivated land area,which results in the concentrated land use type and the best connectivity in this scenario.Shannon's Diversity Index and Shannon's evenness Index are high under the ecological protection scenario,which indicates that various land use types have better diversity and uniform distribution in this scenario,and the landscape types are developing in a balanced direction.