中国城市蔓延时空动态识别及驱动因素探测

城市蔓延已成为全球性的热点问题。基于2000-2017年夜间灯光数据和LandScan人口分布数据，以中国343个地级及以上城市为研究单元，利用能够细分城市内部密度差异的蔓延度指标测度城市蔓延状况，并运用趋势分析、探索性时空数据分析技术（ESTDA）和地理探测器等方法，分析中国城市蔓延变化趋势及时空格局，探究影响城市蔓延的驱动因素。结果表明：①2000-2017年，中国城市蔓延整体表现为中高度蔓延，在空间分布上，蔓延度呈现西北高、东南低的空间格局。②18年间，47%的城市其蔓延度基本保持不变，位于省际交界处的城市其蔓延度呈现增大的趋势，而少数省会城市或区域中心城市的蔓延度则呈现减小的趋势。③蔓延度的局部空间稳定性自东南地区向西北、东北地区递减。蔓延度存在较强的空间依赖性，且城市数量由低LISA弯曲度到高弯曲度大致呈现金字塔结构，79%的城市其蔓延度的波动性较弱；蔓延度没有出现明显的时空跃迁，空间凝聚性较强，空间格局较为稳定，存在一定的转移惰性。④第三产业产值、GRP和房地产开发投资额是影响中国城市蔓延的最主要因素。

IDENTIFICATION OF SPATIO-TEMPORAL DYNAMICS AND DETECTION FOR DRIVING FACTORS OF URBAN SPRAWL IN CHINA

Urban sprawl has become a key issue of global concern as a special phenomenon related to abnormal urbanization. DMSP/OLI nighttime light data, NPP/VIIRS nighttime light data, and LandScan population distribution data were used to extract urban areas in China from 2000 to 2017 using the mutation detection method. After this, 343 cities at prefecture-level and above in China were taken as study units and the population sprawl index and the land sprawl index-which together constitute the urban sprawl index (USI)-were modeled to subdivide the differences in density within the cities. The evolutionary trend and spatiotemporal pattern of urban sprawl in China were then analyzed using trend analysis and exploratory spatiotemporal data analysis, while Geodetector was applied to probe the factors driving the changes. The results show the following. 1) There has been moderate or high urban sprawl development in China from 2000 to 2017. In terms of spatial distribution, the USI was high in the northwest and low in the southeast; 2) In 47.2% of the cities studied, the spread index has hardly changed in 18 years. 3) The local spatial stability of the USI gradually decreased from southeast to northwest and northeast. 4) The gross regional product (GRP) of tertiary industry, the general GRP, and investment in real estate development are the most important factors affecting the sprawl in cities at the prefecture level and above in China.