Abstract:

Urban agglomeration makes urban thermal environment problems more complicated. Based on four Landsat remote sensing images taken from 2010 to 2020, through the integration of landscape ecology and statistical analysis methods, bivariate spatial autocorrelation analysis was carried out on iso-green and iso-therm lines in the urban agglomerations in central Shanxi Province to explore the influence of urban agglomeration landscape pattern on the surface thermal environment. The results showed that (1) from 2010 to 2020, the area of the construction land in the central urban agglomeration of Shanxi Province increased every year, and the spatial distribution of all cities showed a trend of shifting to Taiyuan. (2) The land surface temperature in the central urban agglomeration of Shanxi Province increased, and an urban heat island effect gradually appeared. From 2013 to 2016, the center of the high temperature region moved 30.03 km to the northeast and shifted to the north of the central axis of Taiyuan, and the connectivity of the thermal environment among cities was strengthened. (3) Changes in landscape pattern in urban agglomerations had significant influences on the thermal environment: ① Among all landscape types, the contribution indexes of construction land, cultivated land and grassland were the highest, while the contribution indexes of forest land and water body were negative. The contribution value of forest land was -0.975, and the absolute value of the contribution index was the largest; ② based on the landscape pattern index, the aggregate index, the mean patch area, the largest path index, and the percent of landscape index were negatively correlated with land surface temperature at the class level, whereas the patch density and edge density were positively correlated with LST of woodland; at the landscape level, the landscape index such as the landscape shape index LSI, the patch density PD, and Shannon's evenness index SHEI, were positively correlated with LST. The largest correlation coefficient was observed in 2013. The more complex the structural components and spatial configurations of different landscape types, the stronger the influence of landscape pattern on LST; ③ the bivariate spatial autocorrelation of iso-green lines and iso-therm lines in the study area was negative. The spatial correlation types were dominated by High-High agglomeration and Low-Low agglomeration. The High-High area was mainly distributed in the construction land area, and the Low-Low area was located in the green area. This study provides theoretical reference for urban agglomeration development planning and urban agglomeration thermal environment improvement.

Key words: central shanxi urban agglomeration, green landscape, thermal environment, landscape pattern, dynamic evolution, sptaial autocorrelation

摘要：

城市群化使得城市热环境问题变得更为复杂。以2010-2020年4期Landsat影像数据为基础，运用空间分析、景观格局指数与数理统计等方法，分析山西中部城市群景观格局动态演变和热环境分布及变化特征，并定量探究城市群景观格局演变对其热环境的影响作用。研究表明，（1）近10年来山西省中部城市群的建设用地逐年增加，在空间分布上均有向太原市偏移的趋势。（2）研究区地表温度总体呈上升趋势，城市群热岛效应逐渐显现，2013-2016年，高温区重心向东北方移动30.03 km，转移至太原市中轴线偏北方位置，各城市间热环境的连通性加强。（3）城市群的景观格局演变对其热环境有着明显的影响：①各景观类型中建设用地、耕地及草地热环境贡献指数较高，林地与水体贡献指数为负，其中林地贡献值为-0.975，贡献指数绝对值最大；②从景观格局指数来看，在类型水平上，林地的聚集度指数、平均分布斑块面积、最大斑块所占景观面积比例、景观类型比例与地表温度（LST）呈负相关，斑块密度、边缘密度与LST呈正相关，而在景观水平上，景观形状指数、斑块密度、香浓均匀度指数等与LST呈正相关，2013年各相关系数最大，不同景观类型的结构组分与空间构型越复杂，景观格局对LST的影响越强；③反映城市群绿地景观格局的等绿线与对应区域等温线双变量空间自相关呈显著负相关性，空间集聚类型以“低-高”聚集和“高-低”集聚为主，“高-高”聚集区主要分布于建设用地区域，“低-低”聚集区位于绿地景观聚集区域。该研究为城市群发展规划和城市群热环境改善提供了理论参考。

关键词: 山西中部城市群, 绿地景观, 热环境, 景观格局, 动态演变, 空间自相关