Carbon Monitor Cities near-real-time daily estimates of CO2 emissions from 1500 cities worldwide

Da Huo, Xiaoting Huang, Xinyu Dou, Philippe Ciais, Yun Li, Zhu Deng, Yilong Wang, Duo Cui, Fouzi Benkhelifa, Taochun Sun, Biqing Zhu, Geoffrey Roest, K. R. Gurney, Piyu Ke, Rui Guo, Chenxi Lu, Xiaojuan Lin, Arminel Lovell, Kyra Appleby, Philip L. DeCola, Steven J. Davis, Zhu LiuScientific 2022 · DOI

TL;DR

This paper introduces "Carbon Monitor Cities," a novel dataset providing near-real-time daily CO2 emissions for 1500 cities worldwide from 2019-2021, broken down by five key sectors. This dataset addresses the critical need for timely, high-resolution urban emissions data to support climate action, especially in low-income regions.

What Problem Does This Paper Address?

Accurate and timely emissions data is crucial for cities to effectively monitor and manage their climate mitigation efforts, particularly as many have set ambitious net-zero targets. However, existing city-level emissions inventories are often outdated, lack temporal resolution, or are difficult to compare due to inconsistent methodologies and spatial definitions. Most inventories are compiled at the national level, making it challenging to disaggregate emissions to the city scale, especially for smaller cities or those in low-income countries where data is scarce. This gap hinders effective urban climate policy development and progress tracking.

Methods

The Carbon Monitor Cities (CM-Cities) dataset is built upon a four-stage workflow:

  1. Gridded Daily CO2 Emissions (GRACED): Daily national emissions from the Carbon Monitor dataset are spatialized using global datasets like the Global Carbon Grid (GID) and EDGAR, along with satellite data (TROPOMI NO2) to capture daily variations. This creates a global, high-resolution (0.1°x0.1°) daily emission map.
  2. Spatial Disaggregation: The gridded emissions are then disaggregated to city level using two spatial definitions: Functional Urban Areas (FUA) and Global Administrative Areas (GADM). This allows for comparisons and differentiation between administrative and broader urban area emissions.
  3. City-Level Corrections: To address biases in residential and ground transport sectors, city-specific data like TomTom congestion indices and heating degree days (HDD) are used to refine estimates.
  4. Error Correction and Validation: Statistical methods are employed to identify and remove outliers, and the resulting dataset is validated against existing city-level inventories.

Datasets

  • Carbon Monitor: Provides near-real-time national daily fossil fuel CO2 emissions.
  • Global Carbon Grid (GID): Offers global CO2 emissions data from major point sources.
  • Emissions Database for Global Atmospheric Research (EDGAR): Provides sectoral emissions data.
  • TROPOspheric Monitoring Instrument (TROPOMI) NO2: Used as a proxy for daily CO2 emission variability.
  • TomTom Congestion Data: Used for ground transport sector corrections.
  • ERA5 Reanalysis Data: Provides air temperature data for calculating heating degree days (HDD).
  • Global Human Settlement Layer (GHSL) FUA: Defines functional urban areas.
  • OECD FUA: Used for functional urban areas in OECD countries.
  • Global Administrative Areas (GADM): Provides administrative boundaries.
  • Flightradar24: Used for aviation sector emissions.

The paper also compares CM-Cities with existing datasets such as CEADs, MEIC, Vulcan, and CDP.

Key Insights and Ideas

  • Bridging the Timeliness Gap: The most significant contribution is providing daily, near-real-time emissions data for cities, which is a major improvement over annual or monthly inventories. This allows for much more agile climate action and monitoring.
  • Dual Spatial Definitions (FUA vs. GADM): Offering estimates for both Functional Urban Areas and administrative boundaries is crucial. This acknowledges the ambiguity in defining "a city" and allows for more nuanced analysis, differentiating between emissions produced within a city's administrative limits versus those from its broader commuting zone.
  • Focus on Low-Income Regions: The dataset includes estimates for cities in low-income countries, where high-quality emissions data is often severely lacking, thereby enabling climate action in previously underserved regions.
  • Sectoral Breakdown: Disaggregating emissions into five key sectors (power, residential, industry, ground transport, aviation) provides granular insights into the sources of urban emissions, aiding targeted mitigation strategies.

Key Results

  • The CM-Cities dataset provides daily CO2 emission estimates for 1500 cities across 46 countries from January 2019 to December 2021.
  • The overall uncertainty range for the dataset is estimated at ±21.7%.
  • Comparisons with existing datasets (CEADs, MEIC, Vulcan, CDP) show good agreement, with R² values generally above 0.7 for total emissions and varying for specific sectors, indicating the dataset's reliability. For example, R² values between CM-Cities and CEADs for China were 0.96 for total emissions.
  • The dataset successfully captures the impact of events like the COVID-19 pandemic on urban emissions, showing significant drops in ground transport and aviation during lockdowns.

Limitations and Future Work

  • Limited City-Specific Bottom-Up Data: Except for the ground transport sector, the dataset relies on downscaled national inventories and spatial proxies, rather than city-specific bottom-up activity data. This is a limitation, particularly for cities in less developed nations.
  • Data Quality in Low-Income Regions: While the dataset aims to provide estimates for cities lacking inventories, inherent errors and missing values in input data for less developed nations are acknowledged. The authors consider these results as meaningful first-order estimates.
  • Future Work: The authors suggest ongoing efforts to calibrate and update the workflow to improve data quality. Future updates to the dataset are planned. The methodology for spatial disaggregation and the use of both FUA and GADM definitions are highlighted as a way to address inter-dataset comparison challenges.

Summary updated

Abstract

Building on near-real-time and spatially explicit estimates of daily carbon dioxide (CO<sub>2</sub>) emissions, here we present and analyze a new city-level dataset of fossil fuel and cement emissions, Carbon Monitor Cities, which provides daily estimates of emissions from January 2019 through December 2021 for 1500 cities in 46 countries, and disaggregates five sectors: power generation, residential (buildings), industry, ground transportation, and aviation. The goal of this dataset is to improve the timeliness and temporal resolution of city-level emission inventories and includes estimates for both functional urban areas and city administrative areas that are consistent with global and regional totals. Comparisons with other datasets (i.e. CEADs, MEIC, Vulcan, and CDP-ICLEI Track) were performed, and we estimate the overall annual uncertainty range to be ±21.7%. Carbon Monitor Cities is a near-real-time, city-level emission dataset that includes cities around the world, including the first estimates for many cities in low-income countries.
BibTeX
@inproceedings{huo2022carbon,
  author = {Da Huo and Xiaoting Huang and Xinyu Dou and Philippe Ciais and Yun Li and Zhu Deng and Yilong Wang and Duo Cui and Fouzi Benkhelifa and Taochun Sun and Biqing Zhu and Geoffrey Roest and K. R. Gurney and Piyu Ke and Rui Guo and Chenxi Lu and Xiaojuan Lin and Arminel Lovell and Kyra Appleby and Philip L. DeCola and Steven J. Davis and Zhu Liu},
  title = {{Carbon Monitor Cities near-real-time daily estimates of CO2 emissions from 1500 cities worldwide}},
  booktitle = {Scientific},
  year = {2022},
  doi = {10.1038/s41597-022-01657-z},
}
Figure