Eco-Environmental Quality Assessment in China’s 35 Major Cities Based on Remote Sensing Ecological Index
Hui Yue,Ying Liu,Yao Li,Yang Lu
DOI: https://doi.org/10.1109/access.2019.2911627
IF: 3.9
2019-01-01
IEEE Access
Abstract:The ecological conditions in urban area are greatly changed during the process of industrialization and urbanization of China. The pressure-state-response (PSR) framework is the most popular method to evaluate the ecological quality by integrating a set of remote sensing and statistical indicators into one index through a weighting method. However, a completely remote-sensed ecological index (RSEI), integrating normalized difference vegetation index (NDVI), Wet, land surface temperature (LST), and the normalized differential build-up and bare soil index (NDBSI) through principal components analysis (PCA) method, has been proposed to assess the regional ecological quality. The publications about urban ecological evaluation by RSEI often focus on only one city or a certain area and there are few types of research on the ecological quality assessment by RSEI of 35 major cities in China. In this paper, we employed RSEI to monitor the changes in the ecological quality in China' 35 major cities. The results of RSEI were compared to that of PSR and stepwise regression method was applied to establish the quantitative relationship among RSEI, NDVI, Wet, NDBSI, and LST. The results show that there are 18 cities with ecological quality deteriorated, mainly located in the east and southwest of China (Shanghai, Guangzhou, Hongkong, Macao, Nanjing, Haikou, Shijiazhuang, and Xi'an), and 17 cities with better ecological quality, mainly located in the north and central area of China (Beijing, Tianjin, Shenzhen, Taipei, Fuzhou, Chongqing, and Jinan), from 1990 to 2015. The 3D-scatter plots of RSEI, NDVI, Wet, NDBSI, and LST demonstrate that the levels of very bad and bad mainly situate in where with a high density of built-up and low vegetation cover and soil water content. The PSR map, acquired from integrating 17 indicators, is quite similar to that of RSEI generated by merging only four remote-sensed indicators. This indicates that RSEI can be adopted to characterize regional ecological quality. Take the quantitative equation of Shanghai in 2015 as an example, every 1.46 decrement in NDBSI or each 3.72 increments in NDVI value can result in one increment in RSEI value and the ecological quality can be improved. Specifically, the expansion of the built-up area can lead to ecological degradation, and vegetation construction can promote eco-environmental quality.