Research On Dynamic Analysis And Prediction Model Of Urban Expansion Based On RS And GIS—a Case Study Of Central District Of Ganzhou City

Urbanization promotes the expansion of urban space,which leads to the land reduction,vegetation destruction and soil erosion.So,it is necessary to monitor the dynamic changes in urban development in order to analyze the changes in urban land use,which can grasp the regular of urban land use changes from a macro perspective and provide certain reference value for optimizing urban spatial layout.Simulating urban evolution trends,scientifically predicting the direction of urban development,is useful to facilitate the provision of appropriate supporting information for rational development of urban development policies.Therefore,the analysis of the dynamic changes of urban expansion and the study of the dynamic evolution of cities have been the research focus of urban issues.This paper takes the urban main area of Ganzhou City as the research object,analyzes its urban expansion characteristics,predicts the urban development trend in the region,and analyzes the driving force of the urban development of Ganzhou City,which has important practical significance.It can be used for providing reference value on urban planning,industrial upgrading and infrastructure improvement,policy formulation and implementation.In this paper,Landsat series remote sensing images from 1987 to2017 in downtown Ganzhou City and socio-economic statistics from 1987 to 2016 in Ganzhou City were used as the main data sources for carrying up corresponding research.The main research contents of this article are below:(1)Study on land use information extraction method.Through the comparison and analysis of remote sensing imagery land use classification methods using a single normalized building index,soil vegetated index,and improved normalized water body index and support vector machines for remote sensing image classification,selects the higher accuracy land use information extraction classification method.The results show that the classification accuracy is higher based on the support vector machine,and the method has the learning function,and the learning classifier can be output.The next time the classification can directly provide a good classifier.(2)Study on urban expansion characteristics.According to the SVM classification image of land use in the main urban areas of Ganzhou City in 1987,1995,2003,2013 and 2017,the characteristics of expansion were analyzed from the aspects of urban expansion speed and expansion intensity.The results show that the area of urban construction land in downtown Ganzhou City has increased from 19.32km2 in1987 to 130.15km2 in 2017.In the course of nearly 30 years of development,the area of urban construction land has increased by 120.83km2.The urban expansion speed of the downtown area of Ganzhou City is moderately fast and the expansion trend is relatively stable.(3)Study on urban expansion driving force.Principal component analysis method is used to analyse the selected socio-economic elements of the eight sub-categories of economy,transportation and population from the economic statistical data from the year of the study,which determines the driving factors.From the experimental results,it can be concluded that economic growth,population increase,and improvement of traffic conditions are the main factors affecting the urban expansion of Ganzhou City.The analysis shows that the proportion of the primary industry in the GDP gradually decreases,while the share of the secondary and tertiary industries represented by industry and service industries in the GDP continues to increase,and the improvement of the traffic conditions lays the necessary hardware foundation for urban development.The support of government policies is conducive to guiding the direction and scale of urban expansion.(4)Based on case-based reasoning and cellular automata theory,a combination prediction model with a case-based reasoning model for cellular automata was explored and set up.Under the premise of satisfying the forecasting accuracy,the city's main urban area construction land in 2025 was forecasted based on the historical data.The results show that the simulation accuracy obtained by this model is relatively high,and it has good practicality for the prediction of the evolution of the downtown area of Ganzhou City.