Advancing ecological quality assessment in China: Introducing the ARSEI and identifying key regional drivers

Abstract

Accurate analysis of regional ecological quality and its drivers is crucial for the sustainable development of human society. The remote sensing eco-index (RSEI) has been widely used to monitor changes in ecological quality in many countries or regions, but it ignores the problem of declining air quality caused by economic development and population growth. Consequently, an improved remotely sensed ecological index (ARSEI) was developed to evaluate China's ecological environment quality by incorporating aerosol optical depth (AOD) into the index system. Additionally, a random forest regression model was used to rank the importance of ecological indexes in the ARSEI. Furthermore, a geographical detector was utilized to assess the impact of natural and socioeconomic factors on the spatial heterogeneity of the ARSEI in six geographic regions of China, identifying their primary drivers. The research findings revealed the following: (1) There are similarities and differences in the order of importance of ecological indicators in the ARSEI across the six geographic regions. (2) ARSEI values significantly increased in 24.70% of China's areas, primarily in the Northeast Plain, Loess Plateau, and Tarim Basin, while they significantly decreased in 5.35% of the areas, mainly in the Qinghai-Tibetan Plateau, the northern part of the Tianshan Mountains, eastern coastal cities, and central urban agglomerations. (3) Rainfall and vegetation conditions are the main factors affecting environmental quality in the Three-North region (XB, HB and DB). In the southern (XN, ZN and HD) regions, vegetation cover and land use change, population density and PM2.5 concentrations were greater than the influence of climate factors. The interaction of socioeconomic factors, including PM2.5, land use change, and population density had a greater impact on the spatial heterogeneity of the ARSEI in the southern regions. The results of this study can provide data support for the coordinated development of regional ecosystems and socioeconomics.

Keywords

• Random forest
• Regional applicability
• Influencing factors
• Spatiotemporal variation