Quantification of emission changes is a prerequisite for the assessment of control effectiveness in improving air quality. However, the traditional bottom-up method for characterizing emissions requires detailed investigation of emissions data (e.g., activity and other emission parameters) that usually takes months to perform and limits timely assessments. Here we propose a novel method to address this issue by using a response model that provides real-time estimation of emission changes based on air quality observations in combination with emission-concentration response functions derived from chemical transport modeling. We applied the new method to quantify the emission changes on the North China Plain (NCP) due to the COVID-19 pandemic shutdown, which overlapped the Spring Festival (also known as Chinese New Year) holiday. Results suggest that the anthropogenic emissions of NO₂, SO₂, volatile organic compound (VOC) and primary PM_2.5 on the NCP were reduced by 51 %, 28 %, 67 % and 63 %, respectively, due to the COVID-19 shutdown, indicating longer and stronger shutdown effects in 2020 compared to the previous Spring Festival holiday. The reductions of VOC and primary PM_2.5 emissions are generally effective in reducing O₃ and PM_2.5 concentrations. However, such air quality improvements are largely offset by reductions in NO_x emissions. NO_x emission reductions lead to increases in O₃ and PM_2.5 concentrations on the NCP due to the strongly VOC-limited conditions in winter. A strong NH₃-rich condition is also suggested from the air quality response to the substantial NO_x emission reduction. Well-designed control strategies are recommended based on the air quality response associated with the unexpected emission changes during the COVID-19 period. In addition, our results demonstrate that the new response-based inversion model can well capture emission changes based on variations in ambient concentrations and thereby illustrate the great potential for improving the accuracy and efficiency of bottom-up emission inventory methods.

The observation-based air pollution forecasting method has high computational efficiency over traditional numerical models, but a poor ability in long-term (after 6 h) forecasting due to a lack of detailed representation of atmospheric processes associated with the pollution transport. To address such limitation, here we propose a novel real-time air pollution forecasting model that applies a hybrid graph deep neural network (GNN_LSTM) to dynamically capture the spatiotemporal correlations among neighborhood monitoring sites to better represent the physical mechanism of pollutant transport across the space with the graph structure which is established with features (angle, wind speed, and wind direction) of neighborhood sites to quantify their interactions. Such design substantially improves the model performance in 72-hour PM_2.5 forecasting over the whole Beijing-Tianjin-Hebei region (overall R² increases from 0.6 to 0.79), particularly for polluted episodes (PM_2.5 concentration > 55 µg/m³) with pronounced regional transport to be captured by GNN_LSTM model. The inclusion of the AOD feature further enhances the model performance in predicting PM_2.5 over the sites where the AOD can inform additional aloft PM_2.5 pollution features related to regional transport. The importance of neighborhood site (particularly for those in the upwind flow pathway of the target area) features for long-term PM_2.5 forecast is demonstrated by the increased performance in predicting PM_2.5 in the target city (Beijing) with the inclusion of additional 128 neighborhood sites. Moreover, the newly developed GNN_LSTM model also implies the "source"-receptor relationship, as impacts from distanced sites associated with regional transport grow along with the forecasting time (from 0% to 38% in 72 h) following the wind flow. Such results suggest the great potential of GNN_LSTM in long-term air quality forecasting and air pollution prevention.