Spatiotemporal characteristics analysis of multi-factorial air pollution in the Jing-Jin-Ji region based on improved sequential ICI method and novel grey spatiotemporal incidence models.

Air pollution shares the attributes of multi-factorial influence and spatiotemporal complexity, leading to air pollution control assistance models easily falling into a state of failure. To address this issue, we design a framework containing improved data fusion method, novel grey incidence models and air pollution spatiotemporal analysis to analyze the complex characteristics of air pollution under the fusion of multiple factors. Firstly, we improve the existing data fusion method for multi-factor fusion. Subsequently, we construct two grey spatiotemporal incidence models to examine the spatiotemporal characteristics of multi-factorial air pollution in network relationships and changing trends. Furthermore, we propose two new properties that can manifest the performance of grey incidence analysis, and we provide detailed proof of the properties of the new models. Finally, in the Jing-Jin-Ji region, the novel models are used to study the network relationships and trend changes of air pollution. The findings are as follows: (1) Two highly polluted belts in the region require attention. (2) Although the air pollution network under multi-factorial fusion obeys the first law of geography, the network density and node density exhibit significant variations. (3) From 2013 to 2021, all pollutants except O ₃ show improvement. (4) Recommendations for responses are presented based on the above-mentioned results. (5) The parameter analyses, model comparisons, Monte Carlo experiments and model feature summaries illustrate that the proposed models are practical, interpretable and considerably outperform various prevailing competitors with remarkable universality.
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2024 Elsevier Inc. All rights reserved.)