Spatiotemporal evolution and risk thresholds of PM2.5 components in China from the human health perspective

Abstract

PM_2.5 is a significant global public health hazard, with its components closely linked to various fatal diseases, thereby significantly increasing mortality rates. This study analysed the spatiotemporal evolution of PM_2.5-related mortality and death rates in China using risk attribution methods based on PM_2.5-component and population data. The study used a LightGBM model based on Bayesian and SHAP algorithms to identify the concentration thresholds at which the components of PM_2.5 affect mortality. The results showed that from 2001 to 2022, the mortality rates influenced by the five PM_2.5 components showed a consistent downward trend, decreasing by 101,000–529,000, with a mean annual decrease of 2.2–4.6%. The relative importance of organic matter (OM), nitrate (NO₃^-) and ammonium (NH₄⁺) in influencing mortality increased by 6.3, 17.4 and 4% respectively, while the relative importance of black carbon (BC) and sulphate (SO₄^2-) in influencing mortality decreased rapidly to approximately 2%. The contribution of OM and SO₄^2- to mortality exceeded 30%, with thresholds of 8.6–10.3 and 9.6–10.5 μg/m³, respectively. The thresholds for the effects of NO₃^-, NH₄⁺, and BC on mortality were 6.8–9.4, 3.4–6, and 1.8–3.5 μg/m³, respectively. This study provides effective insights for policymakers to help formulate targeted air pollution control measures and optimise public health interventions to support human health and sustainable societal development.