Cities and towns play a key role in reducing global greenhouse gas emissions, contributing approximately 75 % of global CO2 emissions. With the world’s urban population projected to grow from 55 % in 2018 to 70 % in 2050, their role is even more important. Although activity-based (“bottom-up”) estimates are commonly used to estimate emissions, they often lack independent verification. Recent study, led by DY Ahn et al. (2025), uses satellite observations of CO2 from Orbiting Carbon Observatory-3 (OCO-3) to estimate CO2 emissions for 54 global cities using a top-down approach. This study highlights the potential of satellite data to bridge the gaps between top-down and bottom-up emission estimates, thereby increasing the robustness and transparency of emissions monitoring.

The global analysis, which includes 54 cities in 27 countries, was contributed by the computationally efficient Cross-Sectional Flux (CSF) method. OCO-3, operated by NASA since 2019, is located on the International Space Station (ISS) and, thanks to its “Snapshot Area Mapping (SAM)” capability, is able to collect high-density measurements of total CO2 column (XCO2) with high spatial resolution. To identify urban emissions, the CSF method integrates NO2 observations from the TROPOMI satellite and trajectory simulations from the HYSPLIT model. TROPOMI provides high-spatial resolution NO2 measurements that are used to detect urban NO2 emissions. HYSPLIT simulates forward trajectories, increasing the probability of detecting urban emissions and improving the accuracy of CSF estimates by dynamically extracting wind speeds along the emission path. The team used OCO-3 SAM data collected between September 2019 and November 2023, representing 4.2 years of data. For 54 locations, 2,381 Gaussian curves were successfully estimated from 434 OCO-3 SAM observations.

The results of the satellite analysis show that although top-down (satellite) global emission estimates agree with two widely used bottom-up datasets (EDGAR and ODIAC) within 7 %, this agreement is due to the cancellation of large errors in opposite directions at the level of individual cities. The study revealed significant regional discrepancies:

• Bottom-up estimates tend to overestimate emissions for cities in Central East Asia and South and West AsiaIn Central East Asia, all four bottom-up estimates showed positive errors compared to satellite estimates. In South and West Asia, EDGAR and ODIAC also overestimated emissions.
• Conversely, bottom-up estimates tend to underestimate emissions in Africa, East and Southeast Asia and Oceania, Europe and North AmericaIn Africa, satellite estimates fell outside the 1 sigma range of all four bottom-up estimates.

In addition to comparing emissions, the satellite socio-economic analysis provided interesting insights:

• High-income cities they tend to have less carbon-intensive economiesFor example, North American cities emit 0.1 kg of CO2 per USD of economic output, while African cities emit 0.5 kg of CO2 per USD. This relationship points to decoupling CO2 emissions from economic growth in regions such as North America, Europe, and East and Southeast Asia and Oceania.
• Per capita emissions decrease as population size increasesCities with fewer than 5 million inhabitants emit an average of 7.7 tCO2 per person, while cities with more than 20 million inhabitants emit only 1.8 tCO2 per person. This trend confirms previous studies and suggests that larger cities may be more efficient in terms of energy consumption per capita.

Although the CSF method provides consistent and scalable emission estimates for many cities, it has its limitations. It cannot attribute emissions to specific source sectors (e.g. fossil fuels vs. biofuels) or track scope 2 or 3 emissions (beyond the geographic boundaries of the city). Therefore, it is crucial to effectively support global cities in achieving their zero-emission goals. integration of different data sets, including satellite data, high-resolution models, local inventories and bottom-up data. However, satellite observations can monitor Scope 1 CO2 emissions (within the geographic boundaries of a city), which account for approximately two-thirds of total greenhouse gas emissions in C40 cities. This study confirms the growing role of satellite data in verifying urban CO2 emissions and supporting mitigation efforts for global cities. Spring