Abstract Description: In recent decades, rapid urbanization and unprecedented economic growth in China has occurred concurrently with a greater reliance on fossil fuels. This led to increased air pollution levels in China. In response, the government implemented various policies to reduce pollutants. Since the nationwide Air Pollution Prevention and Control Action Plan (APPCAP) began in 2013, the annual average PM2.5 concentration has decreased by 22.3 µg/m³ from 2014 to 2023, with a maximum reduction of about 40 µg/m³ in the BTH area. While O3 concentration has decreased in Anhui, Shanxi, Yunnan, and Hainan provinces, it has risen in other parts of the country, with the highest increase in summertime average maximum daily 8-hour average of 59.6 µg/m³ in Beijing, followed by 35.5 µg/m³ in Zhejiang and 32.9 µg/m³ in Guangxi.
It is unclear, however, to what extent short-term meteorological variability has influenced these trends. Considering the regional nature of pollutant variability, we incorporate average regional reanalysis meteorology and observed PM2.5 and O3 concentrations for 2,028 stations across China from 2014-2023 to understand and quantify meteorology induced pollutant concentration variability at daily scales. To do so, we train a hybrid Kolmogorov-Zurbenko random forest model to predict daily contributions of meteorological variability.
We find changing meteorological contributions to PM2.5 and ozone concurrent with changing concentrations across China throughout the study period. For Beijing, the capital of China, results show an overall decrease of 180 μgm-3 in 2011 to 100 μgm 3 in 2020 in the variance of meteorology-induced PM2.5 concentration at daily scales, whereas we see the opposite trend for O3, with increasing meteorology-induced variability of an average of 4.4% per year after 2014. Our study helps in understanding changing trends in PM2.5¬ and O3 concentration variability across the country by quantifying links between long-term policy implementation and short-term meteorological contributions to pollutant concentrations.
