Abstract
We present a near-real-time global gridded daily CO$_2$ emissions dataset (GRACED) throughout 2021. GRACED provides gridded CO$_2$ emissions at a 0.1° × 0.1° spatial resolution and 1-day temporal resolution from cement production and fossil fuel combustion over seven sectors, including industry, power, residential consumption, ground transportation, international aviation, domestic aviation, and international shipping. GRACED is prepared from near-real-time daily national CO$_2$ emissions estimates (Carbon Monitor), multi-source spatial activity data emissions, and satellite NO$_2$ data for time variations of those spatial activity data. GRACED provides the most timely overview of emissions distribution changes, which enables more accurate and timely identification of when and where fossil CO$_2$ emissions have rebounded and decreased. Uncertainty analysis of GRACED gives a grid-level two-sigma uncertainty value of 19.9% in 2021, indicating that reliability was not sacrificed for higher spatiotemporal resolution. Continuing to update GRACED in a timely manner could help policymakers monitor energy and climate policies' effectiveness and make adjustments quickly.
Introduction
Traditional global carbon inventories, such as EDGAR or CEDS, serve as the gold standard for climate science but are often hindered by a significant "data lag," sometimes exceeding a year. Furthermore, these inventories typically provide only monthly or annual averages. This lack of near-real-time data makes it difficult to assess how sudden events—such as pandemic lockdowns, economic shocks, or extreme weather—impact emissions as they happen. To effectively manage carbon neutrality goals, the scientific community must move from retrospective accounting to active monitoring. The Global gRidded dAily CO$_2$ Emissions Dataset (GRACED) addresses this gap by downscaling national daily totals into a fine-grained global grid.
Methodology and Data Integration
GRACED employs a "top-down" spatial gridding approach to produce high-resolution daily estimates. The workflow integrates three primary pillars of data:
- National Totals: Daily national-level sectoral emissions are sourced from Carbon Monitor, tracking major sectors including power, industry, and transport.
- Spatial Proxies: Baseline spatial patterns are derived from the EDGAR database to distribute national totals onto a 0.1° grid.
- Satellite Refinement: To capture dynamic human activity (e.g., localized lockdowns), the team utilizes TROPOMI NO$_2$ satellite data. Since NO$_2$ is co-emitted with CO$_2$ but has a shorter atmospheric lifetime, it serves as a high-resolution proxy for active fossil fuel combustion.
A critical step in this methodology is the alignment of sectoral categories between the near-real-time national data and the spatial proxy databases. This ensures that the temporal variations from Carbon Monitor are applied to the correct spatial footprints in EDGAR.
Key Insights and Results
The 2021 Rebound and Sectoral Trends
Following the significant global dip in 2020, 2021 marked a clear rebound in emissions. The daily average emission per grid cell rose from 3,830 kgC/d in 2020 to 4,092 kgC/d in 2021. However, recovery speeds varied by sector. While the power and industry sectors in regions like China rebounded early, international aviation and shipping only showed significant recovery in the latter half of 2021.
The Weekend Effect
GRACED reveals a pronounced "weekend effect," where global emissions are significantly lower on Saturdays and Sundays (averaging 241 kgC/d per grid cell less than weekdays). This trend is primarily driven by reduced industrial activity and ground transportation in high-income regions.
Spatiotemporal Precision
The dataset achieves a 0.1° × 0.1° resolution (approximately 11km at the equator) updated daily. Despite this high resolution, the grid-level two-sigma uncertainty is maintained at ±19.9%. Seasonal analysis shows that the first quarter (January–March) remains the highest emission period globally, largely due to heating demands in the Northern Hemisphere.
Limitations and Future Directions
While GRACED represents a significant advancement, it relies on the assumption that sub-national CO$_2$ changes strictly follow NO$_2$ variations. Atmospheric conditions can introduce noise into NO$_2$ readings, potentially affecting proxy accuracy. Additionally, some sectors still utilize 2019 spatial patterns, which may not fully reflect structural economic changes post-COVID-19. Future work will focus on extending the time series and refining these proxies to provide a truly "live" dashboard for global climate policy effectiveness.
Summary updated