Spatial Differentiation Of Residential Land Price In Nanchang City

This paper analyses the existing theories and research methods of land price spatial differentiation,and makes an empirical study on the residential land market in Nanchang city.The data of residential land price comes from the evaluation report of residential land demarcation in Nanchang urban area in July 2018 issued by Jiangxi Provincial Natural Resources Department.The land price is revised to evaluate the land price under the condition of standard plot land through the residual method and the revision method of base land price coefficient.The information of residential land price in Nanchang urban area in the report is collated and the spatial database is established.Then,using the exploratory spatial analysis tool(ESDA)of ArcGIS,this paper analyzed the distribution pattern,trend analysis,spatial autocorrelation analysis and land price profile of land price data samples,summarized the spatial differentiation law of land price data in Nanchang,and then compared different Kriging methods to confirm the use of ordinary Kriging method to generate land price by spatial interpolation.According to the two-dimensional map,the spatial differentiation law of land price in Nanchang is displayed intuitively.Finally selected 11 sub-categories of five categories as research object which are volume ratio,traffic conditions,educational resources,landscape ecology and life services,and innovatively used network analysis to participate in the construction of OLS and GWR models to analyze the impact.The relationship between factors and land price and the spatial differentiation of land price are analyzed.This paper mainly draws the following conclusions:First,the original data of residential land price in Nanchang does not conform to normal distribution,and the data after logarithmic transformation obeys normal distribution;the land price in are all inverted U-shaped;the sample points of land price have positive spatial correlation and local agglomeration;and the distribution trend of land price shows the approximate concentric circle decreasing to the edge of the urban center with Chaoyang New Town to Plate-Binjiang to Avenue-Hongguzhong Avenue as the main part,and has strong spatial autocorrelation.Its anisotropy belongs to geometric anisotropy.The results of Kriging spatial interpolation show that the high value area of land price is distributed in the central urban area,and radiates outward as the core.Second,construct the road traffic network in the residential area of Nanchang city.Through the analysis function of the nearest facility point in ArcGIS network analysis,find the nearest facility point for the accident point(land price sample point)and find out the shortest path.Improve the shortcomings of using Euclidean distance in the previous research model,which can not reflect the real distance of the city,the model using this method is more accurate.Third,establish the ordinary least squares model and the geographically weighted regression model.The comparison to GWR model analysis,the price of residential land in Nanchang is not only affected by its own factors,but also by the surrounding location factors.Among them,the volume ratio is the most influential factor to the spatial differentiation of land price,followed by water system,subway station,primary school,bus station,third-class hospital,people's government,middle school,Park square,shopping mall and colleges and universities.