An updated model-ready emission inventory for Guangdong Province by incorporating big data and mapping onto multiple...

An accurate characterization of spatial-temporal emission patterns and speciation of volatile organic compounds (VOCs) for multiple chemical mechanisms is important to improving the air quality ensemble modeling. In this study, we developed a 2017-based high-resolution (3 km × 3 km) model-ready emission inventory for Guangdong Province (GD) by updating estimation methods, emission factors, activity data, and allocation profiles. In particular, a full-localized speciation profile dataset mapped to five chemical mechanisms was developed to promote the determination of VOC speciation, and two dynamic approaches based on big data were used to improve the estimation of ship emissions and open fire biomass burning (OFBB). Compared with previous emissions, more VOC emissions were classified as oxygenated volatile organic compound (OVOC) species, and their contributions to the total ozone formation potential (OFP) in the Pearl River Delta (PRD) region increased by 17%. Formaldehyde became the largest OFP species in GD, accounting for 11.6% of the total OFP, indicating that the model-ready emission inventory developed in this study is more reactive. The high spatial-temporal variability of ship sources and OFBB, which were previously underestimated, was also captured by using big data. Ship emissions during typhoon days and holidays decreased by 23-55%. 95% of OFBB emissions were concentrated in 9% of the GD area and 31% of the days in 2017, demonstrating their strong spatial-temporal variability. In addition, this study revealed that GD emissions have changed rapidly in recent years due to the leap-forward control measures implemented, and thus, they needed to be updated regularly. All of these updates led to a 5-17% decrease in the emission uncertainty for most pollutants. The results of this study provide a reference for how to reduce uncertainties in developing model-ready emission inventories.