Potential of European 14 CO 2 observation network to estimate the fossil fuel CO 2 emissions via atmospheric inversions

Yilong Wang, Grégoire Broquet, Philippe Ciais, Frédéric Chevallier, Felix Vogel, Lin Wu, Yi Yin, Rong Wang, Shu Tao

TL;DR

This study explores how a network of 14CO2 (radiocarbon) measurements across Europe can improve estimates of fossil fuel CO2 (FFCO2) emissions using atmospheric inversion models. They found that denser networks and more frequent sampling significantly reduce emission uncertainties, especially in high-emission regions, but a coarse transport model limits accuracy in low-emission areas. The accuracy of these estimates is also highly sensitive to how well prior emission uncertainties are characterized.

What Problem Does This Paper Address?

Accurately tracking fossil fuel CO2 (FFCO2) emissions is crucial for understanding climate change and verifying emission reduction policies. While national emission inventories exist, they often have significant uncertainties, particularly at regional and sub-annual scales. Atmospheric inversion models offer a way to refine these estimates by using observed atmospheric CO2 concentrations. However, the signal from FFCO2 emissions can be masked by natural CO2 fluxes and other atmospheric processes. Radiocarbon (14CO2) is a valuable tracer because its atmospheric distribution is primarily influenced by FFCO2 emissions. This paper addresses the challenge of optimizing the design of a 14CO2 observation network and inversion system to effectively constrain European FFCO2 emissions, considering factors like network density, sampling frequency, and the accuracy of prior emission data.

Methods

The researchers employed a Bayesian statistical framework for atmospheric inversion. They conducted Observing System Simulation Experiments (OSSEs) to evaluate different network configurations and sampling strategies. This involved simulating synthetic "true" emissions and then generating synthetic atmospheric measurements based on these emissions and a global transport model (LMDZv4). The inversion system then attempts to "recover" the original emissions by assimilating these synthetic observations.

Key methodological aspects include:

  • Transport Model: A global transport model (LMDZv4) with a coarse resolution (3.75° x 2.5°) was used. This is a standard for global inversions but is acknowledged as a limitation for regional detail.
  • Observation Operator: This links the simulated atmospheric concentrations to the emissions, accounting for transport and sampling.
  • Prior Uncertainty Characterization: Two distinct approaches were used to model the uncertainty in the prior emission inventories (Bnotion and Bempiric). This is a critical aspect, as the study highlights the strong influence of prior uncertainty on inversion results.
  • Performance Metrics: Two main diagnostics were used:
  • Uncertainty Reduction (UR): The theoretical reduction in emission uncertainty achieved by the inversion.
  • Misfit Reduction (MR): A more practical measure comparing prior and posterior emission estimates against the synthetic "true" emissions, accounting for model-data misfits.

Datasets

  • Prior Emission Inventories:
    • PKU-CO2 inventory: Used for the initial prior emission estimates and for distributing monthly emission budgets into a gridded map.
  • IER-EDG inventory: Used as a "true" emission dataset in the OSSEs, independent of the prior.
  • Atmospheric Transport Model: LMDZv4 (Laboratoire de Météorologie Dynamique) global transport model.
  • Meteorological Data: ECMWF Interim Reanalysis (ERA-Interim) for nudging the transport model.
  • Synthetic Observations: Generated based on the IER-EDG "true" emissions and the LMDZv4 transport model, with added simulated observation errors.

Key Insights and Ideas

  • 14CO2 is a powerful tracer: The study confirms that 14CO2 measurements, when combined with atmospheric inversions, can significantly improve FFCO2 emission estimates.
  • Network density and sampling matter: Increasing the number of 14CO2 measurement sites and sampling more frequently (e.g., daily vs. bi-weekly) leads to substantial improvements in uncertainty reduction, particularly in regions with high emissions.
  • Prior uncertainty is paramount: The accuracy of the inversion results is highly sensitive to how well the prior emission uncertainties are characterized. A mismatch between the assumed prior uncertainty structure and the actual emission errors can lead to degraded estimates, even with good observational data.
  • Coarse transport models limit regional detail: The use of a coarse-resolution transport model, while standard for global inversions, introduces representation and aggregation errors that limit the ability to precisely pinpoint emissions at finer regional scales, especially in low-emitting areas.
  • Trend detection benefits from longer time series: For detecting emission reduction trends, extending the monitoring period (e.g., to 20 years) is more effective than simply adding more stations in the short term.

Key Results

  • NET17 (17 stations, 2-week sampling): Achieved >30% uncertainty reduction for monthly FFCO2 emissions in regions like western Germany.
  • NET43 (43 stations, 2-week sampling): Showed improved uncertainty reduction for regions like eastern France and southern UK, with gains up to 40-70% in high-emitting areas.
  • NET233 (233 stations, 2-week sampling): Demonstrated significant improvements in eastern Europe and the Balkans, but performance remained limited in low-emitting regions due to the coarse transport model.
  • Daily sampling: Further improved uncertainty reduction compared to 2-week sampling, especially in well-sampled, high-emission areas.
  • Misfit Reduction (MR) vs. Uncertainty Reduction (UR): In cases where the prior uncertainty characterization was poor (INV-N), MR values were often negative, indicating a degradation of emission estimates, highlighting the importance of accurate prior uncertainty modeling.

Limitations and Future Work

  • Coarse Transport Model Resolution: The study acknowledges that the coarse resolution of the LMDZv4 transport model is a key limitation, leading to representation and aggregation errors that hinder precise regional emission estimates.
  • Future Work: Investigating the use of higher-resolution regional transport models is suggested to improve accuracy.
  • Prior Uncertainty Characterization: The sensitivity of results to the prior uncertainty configuration is a major finding. The study highlights the difficulty in perfectly matching assumed prior uncertainties to actual emission errors.
  • Future Work: Emphasizes the need for more detailed and systematic evaluations of emission inventory uncertainties and their spatial/temporal correlations.
  • Other 14CO2 Fluxes: The study assumed that FFCO2 is the dominant influence on 14CO2.
  • Future Work: For real-world applications, accounting for uncertainties from other 14CO2 fluxes (e.g., from nuclear facilities, oceans, biosphere) is necessary.
  • Temporal Correlation of Errors: While some temporal correlations were modeled, the study notes that longer-term correlations in observation errors might be important.
  • Future Work: More complex models accounting for varying levels of annual posterior uncertainties and temporal correlations in transport and other error sources are needed.

Summary updated

Abstract

Combining measurements of atmospheric CO2 and its radiocarbon (14CO2) fraction and transport modeling in atmospheric inversions offers a way to derive improved estimates of CO2 emitted from fossil fuel (FFCO2). In this study, we solve for the monthly FFCO2 emission budgets at regional scale (i.e., the size of a medium-sized country in Europe) and investigate the performance of different observation networks and sampling strategies across Europe. The inversion system is built on the LMDZv4 global transport model at 3.75∘ × 2.5∘ resolution. We conduct Observing System Simulation Experiments (OSSEs) and use two types of diagnostics to assess the potential of the observation and inverse modeling frameworks. The first one relies on the theoretical computation of the uncertainty in the estimate of emissions from the inversion, known as “posterior uncertainty”, and on the uncertainty reduction compared to the uncertainty in the inventories of these emissions, which are used as a prior knowledge by the inversion (called “prior uncertainty”). The second one is based on comparisons of prior and posterior estimates of the emission to synthetic “true” emissions when these true emissions are used beforehand to generate the synthetic fossil fuel CO2 mixing ratio measurements that are assimilated in the inversion. With 17 stations currently measuring 14CO2 across Europe using 2-week integrated sampling, the uncertainty reduction for monthly FFCO2 emissions in a country where the network is rather dense like Germany, is larger than 30 %. With the 43 14CO2 measurement stations planned in Europe, the uncertainty reduction for monthly FFCO2 emissions is increased for the UK, France, Italy, eastern Europe and the Balkans, depending on the configuration of prior uncertainty. Further increasing the number of stations or the sampling frequency improves the uncertainty reduction (up to 40 to 70 %) in high emitting regions, but the performance of the inversion remains limited over low-emitting regions, even assuming a dense observation network covering the whole of Europe. This study also shows that both the theoretical uncertainty reduction (and resulting posterior uncertainty) from the inversion and the posterior estimate of emissions itself, for a given prior and “true” estimate of the emissions, are highly sensitive to the choice between two configurations of the prior uncertainty derived from the general estimate by inventory compilers or computations on existing inventories. In particular, when the configuration of the prior uncertainty statistics in the inversion system does not match the difference between these prior and true estimates, the posterior estimate of emissions deviates significantly from the truth. This highlights the difficulty of filtering the targeted signal in the model–data misfit for this specific inversion framework, the need to strongly rely on the prior uncertainty characterization for this and, consequently, the need for improved estimates of the uncertainties in current emission inventories for real applications with actual data. We apply the posterior uncertainty in annual emissions to the problem of detecting a trend of FFCO2, showing that increasing the monitoring period (e.g., more than 20 years) is more efficient than reducing uncertainty in annual emissions by adding stations. The coarse spatial resolution of the atmospheric transport model used in this OSSE (typical of models used for global inversions of natural CO2 fluxes) leads to large representation errors (related to the inability of the transport model to capture the spatial variability of the actual fluxes and mixing ratios at subgrid scales), which is a key limitation of our OSSE setup to improve the accuracy of the monitoring of FFCO2 emissions in European regions. Using a high-resolution transport model should improve the potential to retrieve FFCO2 emissions, and this needs to be investigated.
BibTeX
@inproceedings{wang2018potential,
  author = {Yilong Wang and Grégoire Broquet and Philippe Ciais and Frédéric Chevallier and Felix Vogel and Lin Wu and Yi Yin and Rong Wang and Shu Tao},
  title = {{Potential of European <sup>14</sup> CO <sub>2</sub> observation network to estimate the fossil fuel CO <sub>2</sub> emissions via atmospheric inversions}},
  booktitle = {Atmospheric},
  year = {2018},
  doi = {10.5194/acp-18-4229-2018},
}
Figure