城市地表水体时空演变及其对热环境的影响

以福州市建成区为例,基于Landsat遥感影像对水体类型进行分类,并提取出1989、1996、2006和2014年的地表水体信息,然后与不透水面、植被和地表温度信息结合,运用回归模型定量分析了1989~2014年间福州城市地表水体的时空变化及其对城市热环境的影响。结果表明：① 1989~2014年间,福州建成区各类地表水体均呈不断减少趋势,25 a间水体总面积减少了1 490.67 hm²,其中有70.0%转变成不透水面;② 1989~2014年减少的1 490.67 hm²水体对福州建成区温度上升的贡献达1.03℃,而在水体减少的具体区域,其升温效应可达3.6℃。

Spatiotemporal Variation of Urban Surface Water and Its Influence on Urban Thermal Environment

Shrinkage of urban surface water bodies has caused a series of problems, such as degeneration of urban ecological quality and the intensification of urban heat island phenomenon, which severely affect human living quality. Fuzhou City, as the capital city of Fujian Province in southeastern China, has witnessed a rapid urban expansion process during the last three decades. The rapid urban growth of Fuzhou City has led to the shrinkage of urban surface water bodies and induced a series of urban environmental problems. To reveal the relationship between the change of surface water area and urban heat environment, remote sensing technique was employed using multi-temporal Landsat TM/OLI/TIRS images of 1989, 1996, 2006 and 2014. The use of the modified normalized different water index (MNDWI) has successfully extracted the information of the city’s surface water bodies from the images. The selected thresholds for the extraction of water bodies of the four study years (1989, 1996, 2006, 2014) were 0.1, 0.1, -0.05 and -0.05, respectively. Because the water area of both Min River and Wulong River, which flow through the study area, were basically unchanged in the four-study years, they were masked out in this study. The assessment of the accuracy shows that the overall accuracies of the water extraction of the four years are all greater than 90.0%, which met accuracy requirements. The extracted urban surface water bodies were classified into three types: river, lake and pond by using shape index. Furthermore, the impervious surfaces and vegetation were extracted by two remote sensing indices, i.e., normalized difference impervious surface index (NDISI), and normalized difference vegetation index (NDVI), respectively. While, the land surface temperature (LST) was retrieved using the single channel algorithm (SC). Combined with LST, impervious surfaces, and vegetation information, the extracted water images were used to analyze the spatiotemporal variation of Fuzhou’s surface water in 1989-2014. Regression analysis was carried out to investigate the quantitative relationship between water and LST. The result shows that all three types of surface waters in Fuzhou urban built-up area have decreased substantially in 1989-2014. Water area decreased by 1 490.67 hm² in 25 years and 70% of the decreased waters converted into impervious surfaces. Of the reduced water, river area decreased by 1 490.67 hm², pond area decreased by 951.21 hm², and lake area decreased by 408.15 hm² during the period. The calculation shows that difference in LST between water and impervious surface is 11.12℃, while the difference between water and vegetation was 4.38℃. A significant negative correlation between the proportion of urban surface water and LST has been detected. According to the regression model, the decrease of surface waters in Fuzhou built-up area in 1989-2014 has contributed to the temperature rise by 1.03℃. However, in the local area where water area has reduced substantially, the warming effect is more obvious. Taking Puxia wetland as an example, the decrease of the surface water area of the wetland in 1989-2014 has contributed to the temperature rise by 3.6℃. Obviously, the reduction of urban surface waters has significantly intensified the urban heat island phenomenon of the city.