Prediction of daily PM2.5 and ozone based on high-density weather stations in China: Nonlinear effects of meteorology, human and ecosystem health risks

In recent decade, despite the implementation of strict emission reductions, PM2.5 and surface ozone (O3) in China are still at high pollution level. In this study, based on observations at air quality stations (∼1500 mainly in urban region), daily PM2.5 and MDA8 (the daily maximum 8 h average) O3 concentrations in high-density and uniform meteorological sites (∼2000) over China during 2017–2019 were retrieved using generalized additive models (GAMs) for different regions on a seasonal scale, in order to establish 0.1° gridded datasets and investigate their spatiotemporal characteristics, nonlinear effects of meteorology, and ambient air pollution-attributable health risks on a regional scale. This model can provide high accuracy in the estimation of daily PM2.5 (O3) in 2018 with overall cross-validation R2 and RMSE of 0.81 (0.80) and 12.73 (17.63) μg/m3 and predictive capability in 2017 and 2019 with R2 of 0.67 (0.70) for PM2.5 (O3). Temperature, relative humidity and visibility are three main meteorological factors driving PM2.5 (O3) response. PM2.5 generally exhibits an exponential decrease with increasing visibility and an increase with increasing relative humidity, while is suppressed at the thresholds of 60% and 70%. For O3, in northern China, positive response of temperature is more important and O3 suppression exists at 36 °C ∼ 41.5 °C, whereas negative response of relative humidity is dominating in southern China. Overall, annual averaged PM2.5 (O3) values across China reduced (increased) by 18.7% (11.2%) between 2017 and 2019, with the highest concentrations and variation in North China. These variations contributed to a 14.9% reduction in PM2.5 attributed premature deaths, but a 10.4% increase in O3 exposure related deaths. Generally, central-eastern China still exists greater health risk with per capita premature mortality ranging from 0.80‰ to 1.19‰ for PM2.5 and 1.91‰ to 2.60‰ for O3. Additionally, O3 vegetation damages demonstrate a decline (increase) by 1.3% (21.7%) in North China with higher O3 pollution (southern China with slight pollution) between 2017 and 2019. These findings highlight the essentiality for the synergistic control of PM2.5 and O3 based on regional differences in China.