城市扩张过程中建设用地景观格局演变特征及其驱动力

剖析城市扩张过程中建设用地景观格局演变特征及其驱动力,不仅有助于城市生态环境问题的解决,而且为城市用地结构优化、城市景观规划等工作提供科学依据。以快速城市化的中型城市—扬州市为例,利用多期(1995、2000、2005、2010年和2015年)Landsat卫星影像、乡镇水平的扬州统计年鉴等数据,运用景观格局分析、增强回归树(Boosted regression trees)等方法,研究建设用地的扩张模式、形态及景观格局,定量探究地理、社会和经济因子对景观格局的影响机制,从而明晰景观格局演变特征及其驱动力。结果表明,1995—2015年,建设用地面积持续增加,填充式(Infilling)、边缘式(Edge-expansion)和跳跃式(Leapfrog)3种扩张模式在各时段均有出现,但其优势度随着城市发展而改变。建设用地的形态在城市扩张的过程中经历着"集聚"和"扩散"的交替变化过程,景观格局则出现了同质化倾向,景观破碎化下降、聚合度增加。地理因子(海拔和到县市中心的距离)对景观格局的综合影响虽然高于社会经济,但它的影响力却随着城市发展呈现出下降趋势,社会经济的作用则逐渐增强。海拔和人口密度的增加会促进景观的破碎化、抑制景观的集聚;人口数量的增加会促进景观的集聚、降低景观的破碎化;其余因子(到县市中心的距离、人均GDP和第二产业占比)的作用则随城市的发展而发生转变,如人均GDP对景观破碎化的作用表现为"促进→抑制"的转变、第二产业则为"抑制→促进"。

Changes in landscape pattern of built-up land and its driving factors during urban sprawl

Understanding changes in the landscape pattern of built-up land its driving factors not only helps to solve the eco-environmental problems in cities but also provides a scientific basis for urban land use optimization and landscape planning. In this study, we used multi-source data (e.g., 30 m resolution Landsat images and statistical yearbooks) to analyze changes in sprawl mode, morphology and landscape pattern of built-up land in a rapidly urbanized medium-sized city-Yangzhou City, Jiangsu Province, China. Furthermore, we employed the boosted regression tree method to quantify the effects of geographic and socioeconomic factors on landscape patterns of the built-up land, which revealed changes in the driving factors with the built-up land sprawl. The results showed that the area of the built-up land gradually increased from 1995 to 2015. All of the three sprawl modes (infilling, edge-expansion and leapfrog) appeared in each period, but their relative dominances varied with the built-up land sprawl. The morphology of the built-up land underwent an alternating process of "diffusion" and "coalescence", while the landscape pattern tended to homogenize (i.e., decreased fragmentation and increased aggregation of the built-up landscape). Geographical factors (altitude and distance from the geographical center of a county) played more important roles in shaping landscape pattern than socioeconomic factors. Moreover, the combined effects of geographical factors decreased with the built-up land sprawl, while socioeconomic effects became more significant. Landscape fragmentation of the built-up land increased with increasing altitude and population density and decreasing total population; meanwhile, the landscape aggregation was inhibited. However, effects of other factors (i.e., distance from the center of a county, GDP per capita, and the proportion of secondary industry) on landscape pattern varied with the built-up land sprawl. For example, the promoting effect of GDP per capita on landscape fragmentation turned to an inhibiting effect, while the inhibiting effect of secondary industry turned to a promoting one.