Satellite-based Analysis of CO2 Emissions in Global Cities: Regional, Economic, and Demographic Attributes

Cities play a crucial role in reducing global greenhouse gas emissions. While activity-based (“bottom up”) emission estimates are widely used for global cities, they often lack independent verification. In this study, we use remotely-sensed CO2 observations from the Orbiting Carbon Observatory-3 (OCO-3) to “top-down” estimate CO¬2 emissions for 54 global cities. This global-scale analysis is enabled by a computationally efficient, low latency Cross-Sectional Flux (CSF) approach, which constrains urban plume shape using NO2 observations from TROPOMI and trajectory simulations from HYSPLIT. Our satellite-based emission estimates for 54 global cities agrees within 7% to two widely used bottom-up datasets but reveal regional discrepancies. Bottom-up estimates tend to overestimate emissions for cities in Central East Asia and South and West Asia, while underestimating emissions in Africa, East and Southeast Asia & Oceania, Europe, and North America. Additionally, our satellite-based socioeconomic analysis shows that 1) high-income cities tend to have less carbon-intensive economies: North American cities emit 0.1 kg CO2 per USD of economic output, while African cities emit 0.5 kg CO2 per USD, and 2) per capita emissions decrease with increasing population size, from 7.7 tCO2/person for cities under 5 million residents to 1.8 tCO2/person for cities over 20 million residents. This study highlights the potential of satellite data to bridge gaps between top-down and bottom-up emission estimates, enhancing the robustness and transparency of emissions monitoring. Our findings emphasize the growing role of satellite data in verifying urban CO2 emissions and supporting efforts to mitigate emissions for global cities.