Evaluating Anthropogenic CO2 Bottom-Up Emission Inventories Using Satellite Observations from GOSAT and OCO-2

Abstract

Anthropogenic carbon dioxide (CO2) emissions from bottom-up inventories have high uncertainties due to the usage of proxy data in creating these inventories. To evaluate bottom-up inventories, satellite observations of atmospheric CO2 with continuously improved accuracies have shown great potential. In this study, we evaluate the consistency and uncertainty of four gridded CO2 emission inventories, including CHRED, PKU, ODIAC, and EDGAR that have been commonly used to study emissions in China, using GOSAT and OCO-2 satellite observations of atmospheric column-averaged dry-air mole fraction of CO2 (XCO2). The evaluation is carried out using two data-driven approaches: (1) quantifying the correlations of the four inventories with XCO2 anomalies derived from the satellite observations; (2) comparing emission inventories with emissions predicted by a machine learning-based model that considers the nonlinearity between emissions and XCO2. The model is trained using long-term datasets of XCO2 and emission inventories from 2010 to 2019. The result shows that the inconsistencies among these four emission inventories are significant, especially in areas of high emissions associated with large XCO2 values. In particular, EDGAR shows a larger difference to CHRED over super-emitting sources in China. The differences for ODIAC and EDGAR, when compared with the machine learning-based model, are higher in Asia than those in the USA and Europe. The predicted emissions in China are generally lower than the inventories, especially in megacities. The biases depend on the magnitude of inventory emissions with strong positive correlations with emissions (R2 is larger than 0.8). On the contrary, the predicted emissions in the USA are slightly higher than the inventories and the biases tend to be random (R2 is from 0.01 to 0.5). These results indicate that the uncertainties of gridded emission inventories of ODIAC and EDGAR are higher in Asian countries than those in European and the USA. This study demonstrates that the top-down approach using satellite observations could be applied to quantify the uncertainty of emission inventories and therefore improve the accuracy in spatially and temporally attributing national/regional totals inventories.

A Novel Approach for Predicting Anthropogenic CO2 Emissions Using Machine Learning Based on Clustering of the CO2 Concentration

“…We would extend the study areas, such as the American land area, to validate our proposed approach in the future. Another additional study built the grid-based prediction model by generalized regression neural network (GRNN) algorithm using a long time series of data from 2010 to 2019, each grid including XCO 2 and SIF, where the predicted ACE was 4% lower than ODIAC in 2019 for the Chinese region and indicated that larger predicting biases were mainly located around the big cities [28]. This GRNN model needs long-term series data that are difficult to collect when using multiple parameters, which could not further improve the prediction accuracy of the model.…”

“…Consequently, the values of emissions have a regional non-normal distribution, which affects the unbiased modeling and accurate estimation of ACEs. Several studies have attempted to reduce this effect using the sampling methods of training data (e.g., by segmenting the emission intensity into two parts and then constructing an estimation model on a segment basis [27,28]) and demonstrated better validation results for predicting anthropogenic CO 2 emissions at the national scale. However, these previous studies used few predictor variables and insufficiently evaluated the prediction results.…”

A top-down estimation of subnational CO2 budget using a global high-resolution inverse model with data from regional surface networks

“…These estimates at the urban scale have widely been obtained by top-down inversion approaches (more details in supplementary note 1) based on atmospheric CO 2 measurements (e.g. Lauvaux et al 2020 for Indianapolis; Breon et al 2015 for Paris; Verhulst et al 2017 for Los Angeles; Mueller et al 2021 for Baltimore-Washington area; Basu et al 2020 for the United States), which allows independent evaluation and identification of potential quality issues of GHG inventories (Zhang et al 2022). Its applications in the estimation of anthropogenic GHG were reported by Manning et al (2011), Lauvaux et al (2020), andNASEM (2022).…”