Inversion of CO emissions in Greater Bay Area over southern China using a WRF-STILT-Bayesian framework

Carbon monoxide (CO) is a major atmospheric pollutant with adverse health effects on humans. Moreover, CO can indirectly prolong the lifetime of methane and contribute to global warming. Therefore, understanding the spatial distribution of CO emissions is crucial for designing much-needed strategies to control this pollutant. In this work, a hybrid Weather Research & Forecasting–stochastic time-inverted Lagrangian transport (WRF-STILT)–Bayesian inversion framework was constructed to correct CO emissions over the Greater Bay Area (GBA) for February 2019 and February 2020. After adjusting CO emissions, the average root mean squared error (RMSE), normalized mean error (NME), and correlation coefficient (R) for the simulated CO concentrations in February 2019 and 2020 changed from 0.31 ppm to 0.12 ppm (a 61% reduction), 0.35 to 0.13 (a 63% reduction), and 0.47 to 0.87 (an 85% increase), respectively. The updated CO emissions were then used as input for the Comprehensive Air Quality Model with Extensions (CAMx), a Eulerian model, to further validate the method. The results again indicated that the simulation performance was improved substantially, with a 58% increase in the average R value, a 62% reduction in the RMSE, and a 68% reduction in the NME. This validates the effectiveness of the proposed method in correcting CO emissions. According to the updated emission data, CO emissions over the GBA during the Spring Festival and the COVID-19 lockdown period were 8.3% and 19.6% lower than during normal periods, respectively. These results highlight the importance of accounting for such atypical events in emission estimation and air quality modeling. Analysis of the source areas contributing to CO concentrations in population centers of major GBA cities showed that the average contributions from local emissions and emissions from other GBA cities were 45.5% and 38.8%, respectively. The method developed in this work can be further used for CO adjustment in other regions and contribute to a deeper understanding of the characteristics of this important pollutant.