基于MGWR模型的福清市二手住宅价格空间分异及影响因素研究

Spatial Differentiation and Influencing Factors of Second-hand Housing Prices in Fuqing City Based on MGWR Model

为探索中小城市二手住宅价格空间分异特征及影响因素,以福清市2019年二手房交易数据为例,基于MGWR模型进行量化研究,并与经典GWR模型对比分析,结果表明:(1)福清市二手住宅价格具有显著全局空间自相关性,由局部空间自相关性分析得出,住宅价格“高-高”聚集区主要分布在融侨新城一带,“低-低”聚集区主要布局在中心老城区。(2)MGWR模型回归结果优于经典GWR模型及传统OLS模型,适用于中小城市住房价格研究。(3)不同指标因子回归系数空间异质性显著,其中,建筑容积率、到医院、中小学校以及大型商业综合体距离回归系数存在正负差异;小区绿化率和汽车站回归系数为正值,对房价具有正向影响;建成时间回归系数为负值,对房价具有抑制作用。

Taking the second-hand housing transaction data of Fuqing city in 2019 as an example and based on MGWR model, this paper explores the spatial differentiation characteristics and influencing factors of second-hand housing prices in small and medium-sized cities by conducting quantitative research and comparing it with the classical GWR model.The results show as follows:(1)the second-hand housing prices in Fuqing city have significant global spatial auto-correlation;from the local spatial auto-correlation analysis, it is concluded that the “high-high” accumulation area of housing prices is mainly distributed in Rongqiao New City, and the “low-low” accumulation area is mainly located in the central old city;(2)the regression results of MGWR model are better than that of classical GWR model and traditional OLS model, which is suitable for studying housing prices in small and medium-sized cities;(3)the spatial heterogeneity of regression coefficients of different indicators is significant;among them, there are positive and negative symbolic differences in building volume rate, distance regression coefficients to hospitals, primary schools, secondary schools and large commercial complexes;the regression coefficients of residential greening rate and bus station are positive, exerting positive impact on housing prices;the regression coefficient of construction time is negative, imposing inhibitory effect on housing prices.