Research On The Simulation And Prediction Of Construction Land Change Of Xiangxiang City Based On LCM Model

In recent years,land use/cover change(LUCC)has become a research hotspot around the world.Construction land is the most intense and intuitive one in land use change,characterized by its openness,nonlinearity,and dynamic characteristics,which determine that it is difficult for the traditional research methods to truly reproduce the quantitative and spatial changes.Therefore,how to solve the problem of non-linear control of conversion rules,spatial structure change of land use and driving factors in the land use spatial variation simulation,scientifically and rationally studying changes in construction land,simulating future changes,and sustainable use is essential.Xiangxiang City is a county-level city under the jurisdiction of Xiangtan City in Hunan Province,and is located in the core pilot area of the "Two-Oriented Society" building in Changsha,Zhuzhou and Xiangtan.In recent years,the economic and social development of Xiangxiang City has been rapid,and land use has changed dramatically.Taking Xiangxiang City as an example,this study attempted to combine Canonical Correspondence Analysis(CCA)with the land use change model(LCM)based on the IDRISI platform to simulate and predict changes in land use for land use.It provides a theoretical basis for the efficient use of land for urban construction and scientific management of land.Based on the data of land use change in Xiangxiang City,this study used the ArcGIS platform to perform data processing before the introduction of basic data.Using the construction land expansion rate model and space gravity center migration model,the spatial-temporal changes of construction land from 2009 to 2014 in Xiangxiang City were analyzed in multiple levels.In order to explore the influence degree of driving factors on the change of construction land,a model correspondence analysis method was used to analyze and rank the relationship between construction land change and driving factors in Xiangxiang City.On this basis,the converted basic data and impact factors were introduced into the LCM model,and the potential model was transformed using a multilayer perceptron(MLP)neural network.After training the model,the change of construction land in Xiangxiang City in 2014 was simulated.Through the accuracy verification,which can be used for the simulation and prediction of the construction land change in Xiangxiang City in 2020.The main research conclusions are as follows:(1)This study used land expansion rate model and space center of gravity migration model to analyze and study the land use status of Xiangxiang City from time and space level,and obtained the conversion status of various land use lands to construction land during the period from 2009 to 2014.The results show that all types of land use are actively transformed into construction land,especially the proportion of cultivated land and forest land.Moreover,the effective supplement of cultivated land is insufficient,resulting in a significant reduction in the amount of cultivated land in Xiangxiang in recent years.The center of gravity of the regional construction land is gradually shifted from the southeast of the city to the northwest,with an offset distance of 376 meters.(2)The correlation between land use change and driving factors was studied through canonical correspondence analysis(CCA)method.With limited data and shorter intervals,the CCA distribution map overcoming the limitations of the traditional research methods is that it is difficult to clearly reveal the nonlinear relationship between construction land changes and driving factors.(3)This study is based on the Validation function in the LCM.The land use status map 2009 and 2014,and the land use type simulation map for 2014 are tested for accuracy and adjusted for errors.Using the VALIDATE test in the Change/Time series module in GIS Analysis,the test result has a higher kappa index.There is a high degree of consistency between the 2014 land-use simulation map and the 2014 land-use status map.It shows that the land use simulation results in 2014 are relatively good and can be used to predict the changes in the scale and spatial structure of construction land in Xiangxiang City in 2020.(4)In order to improve the scientificity of the simulation results,this study used ANN-CA mode to simulate the change of construction land in the same year in Xiangxiang City,and compared the simulation results of the two models.The results show that compared with the ANN-CA model,the four Kappa coefficients of LCM have a very high degree of matching,and the LCM model runs faster,validates LCM model has obvious advantages for the simulation and prediction research of regional land use changes.This model has a good effect on complex and nonlinear issues such as land use change.