Factors Underlying Spatiotemporal Variations in Atmospheric PM2.5 Concentrations in Zhejiang Province, China

Abstract

Fine particulate matter in the lower atmosphere (PM2.5) continues to be a major public health problem globally. Identifying the key contributors to PM2.5 pollution is important in monitoring and managing atmospheric quality, for example, in controlling haze. Previous research has been aimed at quantifying the relationship between PM2.5 values and their underlying factors, but the spatial and temporal dynamics of these factors are not well understood. Based on random forest and Shapley additive explanation (SHAP) algorithms, this study analyses the spatiotemporal variations in selected key factors influencing PM2.5 in Zhejiang Province, China, for the period 2000–2019. The results indicate that, while factors influencing PM2.5 varied significantly during the period studied, SHAP values suggest that there is consistency in their relative importance as follows: meteorological factors (e.g., atmospheric pressure) > socioeconomic factors (e.g., gross domestic product, GDP) > topography and land cover factors (e.g., elevation). The contribution of GDP and transportation factors initially increased but has declined in the recent past, indicating that economic and infrastructural development does not necessarily result in increased PM2.5 concentrations. Vegetation productivity, as indicated by changes in NDVI, is demonstrated to have become more important in improving air quality, and the area of the province over which it constrains PM2.5 concentrations has increased between 2000 and 2019. Mapping of SHAP values suggests that, although the relative importance of industrial emissions has declined during the period studied, the actual area positively impacted by such emissions has actually increased. Despite developments in government policy, greater efforts to conserve energy and reduce emissions are still needed. The study further demonstrates that the combination of random forest and SHAP methods provides a valuable means to identify regional differences in key factors affecting atmospheric PM2.5 values and offers a reliable reference for pollution control strategies.

Integration of shapley additive explanations with random forest model for quantitative precipitation estimation of mesoscale convective systems

“…Their results show that SHAP can eliminate the difference in the importance of features caused by the selection of different indicators, which is meaningful for revealing the objective phenomenon that has not been found, and it helps to fill the gap in the interpretability of machine learning models. Compared with other studies that use machine learning to predict precipitation (Min et al, 2019;Li et al, 2021a;Lao et al, 2021;Lao et al, 2021), nevertheless, our present work focuses on interpreting the output of the model, exploring the features and models, and the influence between different features. SHAP is an after-the-fact interpretation approach that is suitable for any machine learning model.…”

“…Traditional ground-station observations of precipitation exhibited extremely high measurement accuracy on the point scale, but they cannot accurately reflect the precipitation on the regional scale owing to the sparse distribution and network density of stations. Ground-based radar observations can give the spatial and temporal distribution of precipitation within a 300-km radius range, but their spatial coverage cannot be scaled up to the global scale (Li et al, 2021a). With the development of meteorological satellites, satellite-based quantitative precipitation estimation (QPE) technology has been greatly conducted (Tang et al, 2015;Yang et al, 2018;Zheng et al, 2021).…”

“…Covering the fields of clinical medicine (Wang et al, 2021), finance (Pérez-Castrillo and Sun, 2022), structural engineering (Mangalathu et al, 2020) and the environment (Tang et al, 2022). For instance, based on the RF algorithm and SHAP model, (Li et al, 2021a) analyzed the temporal and spatial variations of selected key factors affecting PM 2.5 (fine particulate matter) in Zhejiang Province, China. The results showed that the factors influencing PM 2.5 varied greatly during the study period, but the Shapley values indicated that their relative importance was consistent.…”

Maximum Likelihood Algorithm Detects Coastal Wetland Changes in Two Contrasting Coastal Wetlands in Louisiana

“…Wetlands also act as a carbon sink, helping to mitigate the effects of climate change. The loss of wetlands is a significant environmental concern, and efforts are being made to restore and protect wetlands in these region (Lane et al, 2021;Li et al,2020;Liu et al, 2020;Keim et al, 2019) Overall, it is evident that Barataria Bay has increased in the open water and built-up areas land cover category but decreased in the forest and marsh categories.…”

Negative Air Ion (NAI) Dynamics over Zhejiang Province, China, Based on Multivariate Remote Sensing Products

“…Secondly, the widely adopted simple empirical model fails to include multi-factor interactions [19]. The drawbacks of the statistical models mentioned above encourage the use of advanced data-driven machine-learning models, such as Random Forest and ANN models, which have special characteristics to simulate the unknown relationships among different parameters [7,13,20,21].…”