广州市核心区城市绿地降温效应研究

城市化在快速推进社会和经济发展的同时, 也造成了严重的城市热环境问题, 绿地被认为是降低城市温度的有效途径。以广州市核心城区为研究区域, 基于2011年6月份和9月份的Landsat-5遥感影像提取城市地表温度信息, 以高分辨率影像获得城市绿地信息, 运用GIS空间分析和建模方法综合分析城市绿地景观格局的降温效应。结果表明, 研究区城市地表温度空间异质性十分强烈, 莫兰指数(Moran’s I)结果显示9月份温度集聚程度更高, 随时间变化低-低空间关联模式面积减少程度最大; 与普通的线性回归模型和空间滞后模型相比, 空间误差模型更能表达城市绿地格局对地表温度的影响, 空间误差模型的决定系数(R2)值比其它两种模型在两个月上都高出20%; 空间误差模型结果显示绿地面积百分比、平均斑块大小(MPS)和最大斑块指数(LPI)与地表温度呈负相关, 而边界密度(ED)和最大面积指数(LSI)呈正相关关系。

Cooling effect of urban green space of Guangzhou core area

While urbanization is rapidly advancing social and economic development,it also causes serious urban thermal environmental problems.Green space is regarded an effective way to reduce urban temperature.In this study,the core urban area of Guangzhou was selected as study site.Urban surface temperature in June and September in 2011 was extracted based on Landsat-5 remote sensing images,and urban green space information was examined with high-resolution image.The cooling impacts of urban green landscape pattern on urban surface temperature was studied with GIS spatial analysis.The results show that the spatial heterogeneity of urban surface temperature in the study area is very strong in June and September.The Moran's I results show that the temperature concentration is higher in September,and the areas of low-low spatial correlation mode decreases with time.Compared with the linear regression model and spatial lag model,the spatial error model explores the influence of urban green space pattern on surface temperature better,and the decision coefficient(R²)value of the spatial error model is 20%higher than that of the other two models.The spatial error model results show that the percentage of green space area,Mean Patch Size(MPS)and Largest Plaque Index(LPI)are negatively correlated with surface temperature,while Edge Density(ED)and Largest Shape Index(LSI)are positively correlated.