Cellular Automata Simulation Of Urban Land Use Change Based On Random Forest

Building a land use model to simulate the process,pattern and mechanism of land use change is the focus of academic and industrial circles.Based on the research object of Jinan Central District,taking land use,DEM,road,population and economy as the data source,this paper obtains the land use change information of the research area from the angle of time,land use transfer,landscape pattern.Under the natural development scenario model,RF-CA-Markov and LR-CA-Markov models were used to simulate the land use evolution process and predict the spatial distribution pattern of land use in 2025.The main research work and conclusions of this thesis are as follows.(1)In this paper,the Landsat image data of the study area in 2005,2010,2015 and2020 were collected,and the image preprocessing such as radiometric correction,projection and cropping were carried out.The random forest method was used for image classification,and the multi-temporal land use status data of five land types,such as construction land,cultivated land,forest land and water area,are obtained.(2)The temporal and spatial changes of land use in the study area from 2005 to 2020 were analyzed from the perspectives of time,land use transfer and landscape pattern,etc.(1)From the perspective of time,the trend of land use change in the study area was the continuous increase of construction land,while other land use types fluctuated in a larger range.(2)The transfer matrix of land use was used to calculate the transfer direction of each land use type.The change of land use structure was obvious,mainly from construction land,cultivated land and forest land.(3)The landscape single index and landscape level index directly reflect the land use situation of the study area.The results of single index show that the construction land is dominant,the spatial agglomeration is increasing,the fragmentation degree of patch is decreasing,and the relationship with other land types is increasing.The proportion of cultivated land is declining,and the degree of fragmentation was deepening constantly.The fragmentation of forest land and water area is increasing,and the spatial connectivity is decreasing.The proportion of other land use types in the landscape pattern declined.The landscape continuity of urban land decreased,the degree of fragmentation and the complexity of spatial structure increased,the landscape structure maintained a good diversity,and the distribution of patch types was relatively uniform.(3)The RF-CA-Markov model was constructed to simulate the land use evolution process of the study area during 2010-2015 and 2015-2020,and the land use simulation maps of 2015 and 2020 were obtained,which had a high consistency with the current situation maps of 2015 and 2020.In order to improve the reliability of the simulation results,the LR-CA-Markov model was selected for comparison and verification.The experiment show that compared with LR-CA-Markov model,RF-CA-Markov model was more suitable to predict the future land use pattern,and the land use pattern in the study area from 2020 to 2025 will keep the development trend from 2015 to 2020.