Comparative Study On Spatial Variation Of Urban Residential Price And Its Influence Factors Based On GWR Model

Residential prices is the result of a variety of factors,and its spatial distribution has certain correlation and heterogeneity.The difference of geographical condition is an important reason for the spatial variation of residential prices.Thus,the study of the spatial variation of residential prices and its influencing factors can provide a scientific basis for the government departments to carry out the residential development planning and formulate targeted price control policies,and to ensure the effective control of house prices growth and promote the healthy development of the real estate market at the same time.This paper is based on the theory of residential spatial variation,location theory and urban spatial structure theory,and taking Kunming and Chengdu as the research object,to study on the spatial distribution of residential prices in different cities by GIS spatial analysis,meanwhile,establishing the general linear regression model and geographically weighted regression model(GWR)to analyze quantitatively the formation mechanism of residential prices spatial variation and the difference of its influencing factors on space.The main conclusions of this paper are as follows:(1)Residential prices of Kunming and Chengdu are showing the inverted "U"shape curve distribution in all directions.The trend is decreasing from city center to the surrounding.The residential price of the two cities showed a significant spatial agglomeration phenomenon in the whole,but there was spatial heterogeneity in the local area.(2)The residential price of Kunming are showing circle pattern.Although there are many residential sub centers,but the characteristics of the single center structure is more obvious.The distribution of residential price in Chengdu is not obvious,and the circle pattern of residential price's space distribution has been broken,the spatial distribution of residential prices in a number of areas is differentiation.(3)Through the establishment of residential price and its influencing factors of the general linear regression model and GWR model,the fitting effect of GWR model is better than the general linear regression model.Secondly,the effects of different factors on the residential price in different scale cities have great difference,for example,the most influential factor to the spatial variation of residential price in Kunming is the distance of CBD,but the most significant factors in Chengdu are the greening rate and the property management fee,and the influence degree of CBD is relatively small.Most of the factors have a significant impact on the residential prices outside the central radiation area,the degree of influence on the residential price in the radiation area is small.(4)The empirical study found that the results obtained by GIS spatial analysis and GWR model are basically consistent with the real estate market's situation in Kunming and Chengdu,indicating the suitability of the two methods in the study of the spatial variation of the housing price and its influencing factors.