Research On The Spatial Variation Of The Urban Residential Land Prices Based On GWR Model Of Shangrao City

Urban land is the main place of production and daily life of urban residents. It isalso the center of economy, politics, culture, education, technology and informationcommunication. Residential land price reflect the health condition of land marketand the supply and demand of land directly. Besides, it is the significant tool forgovernment to carry out macroeconomic regulation and control. With the rapiddevelopment of national economy, it has become an important subject of landresources management that to make rational plans and policy of the whole in order toachieve the sustainable utilization of urban land resources and to prevent the loss ofland assets by studying the spatial differentiation of Residential land price and itsinfluencing factors.There are amount of studies of the spatial differentiation and its influence factorsof Residential land price, most of which are using statistical analysis method toestablish characteristics of price model and quantitating the factors that influence theResidential land price to represent the land price changes of the whole space. Suchanalysis of space structure of land price reflect the whole domain changes of landprice only, theycannot explain the minor changes of the study area precisely.Whentalking about the evaluation results, the characteristic of the price model is unitary andnot intuitive.Since then, we need to choose a better way to analysis the spacedifferentiation of the land price, namely, the geographically weighted regression(GWR) model.In this paper, we defined the connotation of Residential land price dependingon the studies of space differentiation of land price and the theoretical researches bythe scholars in domestic and foreign countries. We considered Shangrao City as thestudy area, investigating and quantitating the data of land transfer market situation in2009-2012. Firstly, using the method of exploratory spatial analysis (ESDA) toexplore the space structure and the space change trend of residential land pricedistribution of Shangrao City.The result showed that the basic residential lands pricesdata conforms to normal distribution and the land prices are present the direction ofinverted U-shapedin the latitude and longitude, especially the land price in longitudedirection showed the distribution of low-high-low from south to north.Secondly,analysis the spatial autocorrelation of residential land prices in Shangrao City by using the Geoda software, calculate the land prices by Kriging interpolation and drawthe residential land price section in different directions by ArcGIS tools. The resultshowed that the residential land prices of Shangrao City present the accumulativeeffect and in a “single” space structure of distribution, with the trend of diminishingfrom center to periphery. In addition, there are two another areas in tThen, by selecting the weighted function and bandwidth, wechose thegeographically weighted regression model to explore the relationship between theresidential land price and the factors.We comprised and analyzed the result of GWRmodel by the analysis of variance, events, intelligent diagnosis andMonteCarlosignificance test. After that we regarded the adjustment space nuclear asthe final model to explain the space structure of residential land prices, combiningwith the analysis of factors on the distribution of residential land price by using GIStools.The result showed that factors such as hospitals, middle schools, elementaryschools, parks,plot ratio and afforestation rate have a significant impact on theresidential land prices. And different locations have different impact on it.Finally, we made a comparison of the two models in the result of goodness-of-fitR2, residual square, finding that the GWR model was better than the Hedonic model.