The COVID-19 pandemic poses a serious global threat to human health. In China, the government immediately implemented lockdown measures to curb the spread of this virus. These measures severely affected transportation and industrial production across the country, resulting in a significant change in the concentration of air pollutants. In this study, the Euclidean distance method was used to select the most similar meteorological field during the COVID-19 lockdown period. Changes in the concentration of air pollutants in China were analyzed under similar meteorological background conditions. Results indicate that, compared with data from 2015–2019, air quality in China significantly improved; with the exception of ozone (O₃), the concentration of major air pollutants declined. Compared with baseline conditions, the reduction of air pollutants in China from 25 January to 22 February 2020 (Period 2) was the most significant. In particular, NO₂ decreased by 41.7% in the Yangtze River Delta. In Period 2, the reduction of air pollutants in areas other than Hubei gradually decreased, but the reduction of NO₂ in Wuhan reached 61.92%, and the reduction of air pollutants in various regions after February 23 was significantly reduced. By excluding the influence of meteorological factors and calculating the contribution of human activities to atmospheric pollutants by linear fitting, in Period 2 the effect of artificial controls on NO₂ in Wuhan attained 30.66%, and reached 48.17% from 23 February to 23 March (Period 3). Results from this investigation provides effective theoretical support for pollution prevention and control in China.

Near-surface ozone is a secondary pollutant, and its high concentrations pose significant risks to human and plant health. Based on an Extra Tree (ET) model, this study estimated near-surface ozone concentrations with the high spatiotemporal resolution based on Himawari-8 aerosol optical depth (AOD) data and meteorological variables from 1 January 2016 to 31 December 2020. The SHapley Additive exPlanation (SHAP) method was employed to evaluate the contribution of AOD and meteorological factors on ozone concentration. The results indicate that (1) the ET model achieves a sample-based cross-validation R² of 0.75–0.87 and an RMSE (μg/m³) of 17.96–20.30. The coefficient of determination (R²) values of the model in spring, summer, autumn, and winter are 0.81, 0.80, 0.87, and 0.75, respectively. (2) Higher temperature and boundary layer heights were found to positively contribute to ozone concentration, whereas higher relative humidity exerted a negative influence. (3) From 11:00 to 15:00 (Beijing time, UTC+08:00), ozone concentration increases gradually, with the highest occurring in the summer, followed by spring. This study has obtained high spatial and temporal resolution ozone concentration data, offering valuable insights for the development of fine-scale ozone pollution prevention and control strategies.