TL;DR
This paper introduces GCP-GridFED, a new high-resolution (0.1°) global dataset of monthly fossil CO2 emissions from 1959-2018. This dataset is crucial for atmospheric inversion models, which use it as a prior to better understand the sources and sinks of atmospheric CO2, and it also includes estimates of related oxygen uptake.
What Problem Does This Paper Address?
Accurately tracking the sources and sinks of atmospheric CO2 is fundamental to understanding climate change and evaluating mitigation efforts. While global carbon budgets are regularly assessed, a key challenge lies in disaggregating these global totals into spatially and temporally resolved emissions. Atmospheric inversion models, which use observed atmospheric CO2 concentrations to infer surface fluxes, rely heavily on accurate prior estimates of fossil CO2 emissions. Existing datasets often lack the necessary spatial resolution or temporal detail, or are not directly consistent with the national inventory data used in global carbon budget assessments. This paper addresses this gap by providing a gridded dataset that bridges national inventory data with high spatial and temporal resolution, making it directly usable for inversion models.
Methods
The GCP-GridFED dataset was created by scaling monthly gridded emissions from the Emissions Database for Global Atmospheric Research (EDGAR, version 4.3.2) for the year 2010 to match national annual emissions estimates from the Global Carbon Project (GCP) for the period 1959-2018. This involved a multi-step process:
- Grouping EDGAR sectors: EDGAR's detailed activity sectors were grouped to align with GCP-GridFED's source classes (e.g., coal, oil, natural gas, cement, bunker fuel).
- Extracting national data: Country-specific emission grids were extracted from EDGAR.
- Summing emissions: Monthly emissions were summed across the year and across grid cells for each country and source class.
- Calculating scaling factors: Annual scaling factors were derived by comparing the summed EDGAR emissions with the corresponding national annual emissions from GCP-NAE.
- Applying scaling factors: These scaling factors were applied to the monthly gridded EDGAR emissions to produce the final GCP-GridFED dataset.
- Collation: Gridded emissions from all nations were merged to create a global output.
Additionally, the dataset includes gridded estimates of O2 uptake, calculated by applying established oxidative ratios to the CO2 emission estimates for fossil fuels. Uncertainty estimates were also incorporated, based on national inventory uncertainties.
Datasets
- Emissions Database for Global Atmospheric Research (EDGAR) version 4.3.2: Used for its high spatial (0.1°) and temporal (monthly) resolution of emissions for the year 2010.
- Global Carbon Project (GCP) national annual emissions (GCB-NAE): Provided the consistent, long-term national emission totals for the period 1959-2018, which served as the target for scaling.
- EU GISCO Countries 2016 dataset: Used for country masks to extract national data from EDGAR.
- Oxidative Ratios (OR): Literature-based values were used to estimate O2 uptake from CO2 emissions.
Key Insights and Ideas
- Bridging the gap: The core innovation is harmonizing bottom-up national inventories with the high-resolution, spatially explicit data needed for top-down atmospheric models. This consistency is vital for closing the global carbon budget.
- O2 uptake as a proxy: Including O2 uptake estimates, derived from CO2 emissions, provides an additional data stream that can be used in coupled CO2-O2 inversion models, potentially improving flux estimates.
- Temporal consistency: By scaling a single year's monthly distribution (2010 EDGAR) across the entire time series, the dataset captures long-term trends while maintaining a consistent seasonal pattern derived from EDGAR.
Key Results
- Gridded CO2 emissions: The GCP-GridFED dataset provides monthly fossil CO2 emissions at a 0.1° spatial resolution for the period 1959-2018, disaggregated by fuel type (oil, coal, natural gas) and including cement production and international bunker fuels.
- Gridded O2 uptake: Estimates of O2 uptake due to fossil fuel oxidation are also provided.
- Consistency with national data: Global annual emissions from GCP-GridFED were found to be within 0.0077% of the GCB-NAE input data, demonstrating strong consistency.
- Seasonality preserved: The monthly seasonality observed in the EDGAR 2010 data was successfully transferred to the GCP-GridFED dataset.
Limitations and Future Work
- Seasonality assumption: The assumption that the 2010 EDGAR monthly distribution is representative of all years might not fully capture inter-annual variability in emissions seasonality, although a heating and cooling degree day (HCDD) correction was applied to address this.
- Spatial disaggregation uncertainties: The dataset's uncertainties do not fully account for the complexities of spatial and temporal disaggregation of national emissions, which can be significant at finer resolutions.
- Omission of diffuse sources: The dataset does not explicitly account for diffuse CO2 sources from non-CO2 carbon emissions (like CO and CH4) or fugitive methane emissions, which are considered for future development.
- Future updates: The authors plan to update the dataset annually to align with ongoing GCP global carbon budget assessments, incorporating new data and methodologies (e.g., EDGARv5.0, India and USA-specific datasets).
Summary updated