Calculation of CO2 emissions from coal-fired power plants based on OCO-2/3 satellite observations and divergence model

• Published in: Energy (Oxford) 2025-01, Vol.315, p.134303, Article 134303
• Main Authors: Yuan, Tao, Xue, Yong, Wang, Chunbo, Gao, Pengyuan, Zhao, Bingjie, Zhao, Liang, Jin, Chunlin, Li, Shengwei, He, Botao
• Format: Article
• Language: English
• Subjects: CO2 emissions; divergence; OCO-2/3; Power plants
• ISSN: 0360-5442
• DOI: 10.1016/j.energy.2024.134303

The annual Global Carbon Budget projects fossil carbon dioxide (CO2) emissions of 36.8 billion tons in 2023, up 1.1% from 2022 Global fossil fuel emissions primarily result from the combustion of coal, oil and natural gas. Coal-fired power stations are responsible for more emissions than any other fossil fuel, representing approximately 41% of global fossil CO2 emissions in 2023. The bottom-up inventory approach lacks independent third-party data validation. At the IPCC plenary, it was agreed that emission inventories based on atmospheric concentrations should be validated alongside the top-down approach. In this paper, Orbiting Carbon Observatory 2/3 (OCO2/3) observations, paired with a divergence algorithm, are employed to quantify emissions from coal-fired power plants. CO2 emissions from 28 plants worldwide using 30 cases. A comparison of these CO2 emission estimates with emission inventories yielded an R2 value of 0.73 across the 30 cases. Compared with Carbon Brief's emission inventory results, the average percentage difference of our algorithm in China is 22.15%, while the average percentage difference of the OCO-3 Snapshot Area Mapping Observation Model (SAM) is only 6.11%. In the U.S., the average percentage difference compared to the hourly emission inventory data EPA (the US Environmental Protection Agency) reaches 16.67%. This result coincides with the average deviation of 15.1% in the Gaussian plume model for the U.S. power plants. In the European thermal power plant emission estimates, the average percentage difference is about 25% lower compared to the proxy emission values. Sensitivity tests on planetary boundary layer heights (PBLH), the derived regional background, and wind speeds suggest our divergence method is quite robust. This has been demonstrated to be a quick and effective method for estimating carbon emissions from coal-fired power plants. [Display omitted] •Power plant emissions based on a divergence algorithm and locally robust regression background estimation methods.•Analyzing the impact of vertical diffusion on emission estimates using CO2 profiles at different heights.•Tested sensitivity of the algorithm to different background concentration range choices.•The estimated R2 of CO2 emissions amounted to 0.708., with an average percentage difference of 20.87%.•The percentage difference between EPA and estimates from 4.87% to 46.46%, with an average of 13.24 %.