Seoul carbon emissions estimated with Bayesian inverse modeling of measurements from ground and space

The bottom-up method of estimating carbon emissions from fossil fuel use based on socio-economic databases has significant uncertainty. In particular, this uncertainty increases when the spatial and temporal scales are fine, such as in cities. Therefore, an independent and complementary top-down method has been used to verify carbon emissions estimated by the bottom-up method. This method uses the atmospheric CO₂ measurement, transport model, and the inverse model. In this study, carbon emissions provided by ODIAC were improved using ground and satellite CO₂ observation data measured in 2019 for Seoul. We used atmospheric CO₂ concentration from four observation sites located in Seoul and the column-averaged dry air mole fractions of CO₂ from OCO-2. In order to quantify the footprint, which is the flux sensitivities on observation sites, STILT and X-STILT model was used for the ground observation and satellite observation, respectively. The result showed that prior carbon emissions in specific areas including power plants and airport were underestimated. The carbon emission uncertainty decreased through Bayesian inverse model, and it was found that the calculated observation and emission error covariance were appropriate through the reduced chi-square calculation. We assessed Bayesian inverse modelling of Seoul carbon emission from fossil fuel use using measurement from ground and space, which will enable effective carbon neutrality policy decisions for Seoul.

This work was supported by Korea Environment Industry &Technology Institute (KEITI) through "Climate Change R&D Project for New Climate Regime", funded by Korea Ministry of Environment (MOE) (2022003560006)