基于网络爬虫技术获取房产时空数据,结合时间序列和空间自相关分析方法,在"小区—街道—区级"多尺度上探索广州市住宅容积率时空模式与演化规律。结果表明:1近三十年广州市居住空间呈现中心城区集聚开发—向南、北方向近郊区拓展—往南、北、东方向远郊区外扩的演变过程,容积率呈现波动式增长态势。2随着房地产的快速发展,住宅容积率逐渐呈现显著的空间自相关特征,体现了住宅开发与规划有序性的提高。3区级尺度上容积率呈现较强的自相关性,越秀区与天河区为HH集聚区,从化区为LL集聚区。街道尺度上容积率呈现出显著的空间不均衡性,区位交通条件较好的街道成为容积率HH集聚区,而LL集聚区主要分布于生态旅游重镇。
Analyses of the spatial and temporal characteristics of residential Floor Area Ratio(FAR) are of great significance for urban planning and real estate development. The majority of the current studies focus on FAR estimation, calculation, planning and the relationship between FAR and economic development. However, the evolution process and the spatial pattern of FAR lack significant attention. Using residential big data obtained from the Internet, this study investigates the spatial-temporal pattern of FAR at multiple levels by combining time series analysis and spatial autocorrelation methods. The results of the study are as follows:(1) In the past 30 years, the residential space of Guangzhou has expanded from urban central area to the suburban areas. The fluctuation and increase trends in residential FAR have occurred.(2) In the initial stage of residential development before 1994, the FAR at the residential quarter level has no obvious spatial autocorrelation. In the rapid development stage(1995-2004) and the stable equilibrium stage(2005-2014), the FAR of newly-built residential quarters shows a significant spatial autocorrelation pattern. The spatial agglomeration has grown over these three intervals, thus reflecting the gradual improvement of residential development and planning guidance.(3) At district(country) level, the residential FAR has demonstrated a strong spatial autocorrelation. High- high clusters are concentrated in Yuexiu and Tianhe districts, whereas lowlow clusters are concentrated in Conghua district. The spatial differentiation of FAR has emerged at the street(town) level. With the influence of location, transportation, infrastructure construction and land price, several streets in the urban core, such as Dongfeng, Binjiang, and Liede, have become the high- high concentrated areas of residential FAR. Moreover, the lowlow clusters of FAR are mainly distributed in some tourist towns, such as Wenquan, Shawang,and Nansha towns.