Tracking industrial relocation and air pollution redistribution in China: a multi-dimensional satellite-based...

Conventional Air Quality Index (AQI) systems in China are typically based on the single pollutant with the highest concentration, which limits their ability to capture cumulative exposure and multi-pollutant health risks. This study develops an integrated assessment framework and discusses how large-scale changes in air quality can be used to explore spatial patterns potentially associated with industrial relocation and environmental transformation at the national scale. We propose a multidimensional air quality evaluation system comprising the Total Atmospheric Quality Index (TAQI), Potential Atmospheric Quality Index (PAQI), and Effective Atmospheric Quality Index (EAQI), which respectively represent total pollutant load, toxicity-weighted hazard intensity, and governance effectiveness. To facilitate the identification of spatial patterns potentially associated with industrial relocation, air quality dynamics were further coupled with vegetation greenness (NDVI) and economic activity (GDP). Based on this coupling, the Vegetation-Air Quality Index (VAQI), Financial-Air Quality Index (FAQI), and an integrated Industrial Transfer Intensity (ITI) indicator were constructed to screen regions where concurrent environmental and economic changes may indicate intensified industrial transfer processes. An ERA5-based meteorological background comparability analysis was further conducted to assess whether large-scale differences in weather conditions between 2015 and 2019 could confound the interpretation of air-quality redistribution patterns. Applying this framework to China during 2015-2019, we find that national total pollutant load (TAQI) declined by 16.3% and toxicity-weighted hazard intensity (PAQI) by 32.2%, while signals indicative of industrial relocation intensified in northwestern and southwestern regions. The derived ITI shows strong spatial correspondence with independent satellite-based indicators of anthropogenic activity, including VIIRS nighttime light change (r = 0.68, p < 0.001) and Proba-V built-up expansion (r = 0.73, p < 0.001). Overall, this study provides a transferable, satellite-supported framework and discusses how to use it to identify emerging patterns of pollution redistribution and potential industrial relocation hotspots. The results offer policy-relevant evidence to support differentiated air quality governance and sustainable industrial transition in rapidly developing economies.