Cities around the world emit far more methane than official inventories suggest. And rather than decreasing, emissions are rising. This is what emerges from a study published on 13 April in Proceedings of the National Academy of Sciences (PNAS), led by the University of Michigan and funded by NASA and the National Institute of Standards and Technology (NIST). For the first time, a research team has used satellite observations to quantify and monitor urban methane emissions on a global scale.
The researchers – Erica Whiting, a PhD student in climate and space sciences, and Eric Kort, a professor in the same department and the study’s corresponding author – analysed data collected by the TROPOMI tool aboard the European Copernicus Sentinel-5P satellite, launched in 2017 to track air pollution and climate change. TROPOMI measures sunlight reflected by the atmosphere at different wavelengths, with each providing information on the concentration of a specific gas, at a spatial resolution sufficient to distinguish individual urban areas.
Cities account for 10% of global anthropogenic methane emissions
The analysis covered 92 metropolitan areas, comprising a total population of 1.18 billion people. In 2023, the combined methane emissions from these cities accounted for around 10% of global anthropogenic methane emissions – a figure that, according to the researchers, makes total urban emissions almost four times higher than those of the so-called “ultra-emitters” – individual oil and gas infrastructure facilities that release large quantities of methane, such as leaking wells, platforms and pipelines, on which studies and reduction policies have so far focused.
“Cities have the motivation and the power to reduce greenhouse gas emissions and are therefore significant opportunities for impactful reduction,” said Erica Whiting. “But until now, no method existed to quantify and monitor urban methane emissions globally and, consequently, no observation-based method to evaluate reduction strategies.”
The study tracked emissions from 72 cities between 2019 and 2023, including 51 belonging to the C40 network and 21 outside of it. The C40 is a coalition of 97 major cities that have committed to achieving net-zero emissions by 2050 and cutting greenhouse gas emissions, including methane, by half by 2030.
The results are discouraging for those who place their trust in the voluntary commitments of city authorities. Annual methane emissions fell in 2020, due in part to the economic slowdown linked to the pandemic, only to rise steadily thereafter. Between 2020 and 2023, C40 cities recorded a 10% increase, a figure substantially in line with the 12% rise seen in non-member cities. In absolute terms, C40 cities will have to contend with an additional 2 teragrammes of methane emissions per year, equivalent to around 30% of their reduction target. “To reduce greenhouse gas emissions and formulate effective policies, municipalities need to know how much they are emitting and from which sources. But when it comes to methane, there is still a great deal of uncertainty,” said Eric Kort.
The gap between inventories and reality
One of the most significant findings of the study is the discrepancy between “bottom-up” estimates – calculated by adding up emissions reported source by source – and satellite measurements. While inventories indicate a rise in urban methane emissions of between 1.7% and 3.7% since 2020, TROPOMI data show a global increase of 6% since 2019, with varying regional trends: many European cities, in a counter-trend, have recorded a decline. This discrepancy suggests that some urban methane sources are overlooked or underestimated in official records.
Urban methane sources are diverse and often elusive: leaks from natural gas distribution networks, landfill sites, and wastewater treatment plants. However, TROPOMI’s resolution is not yet sufficient to precisely identify emission points within a city. “We are looking at higher-resolution satellite measurements to distinguish the contribution of large localised sources,” explained Kort. “These satellites cannot necessarily tell us how much an entire city emits, but they could tell us what individual landfills or plants are doing.”
C40’s response: “New data, not necessarily new emissions”
Renewable Matter has reported the reaction of the C40 network via Caterina Sarfatti, the organisation’s Managing Director for Inclusion and Global Leadership. Sarfatti welcomes the study with an open mind but also with a reservation about the methodology: “We are pleased that a new and more accurate technology is bringing further data to light. However, as we were not consulted during the research, which is somewhat unusual, we were unable to carry out a peer review.”
The C40’s hypothesis – which Sarfatti cautiously leaves open precisely on account of the lack of dialogue with the researchers – is that satellites are detecting emissions previously invisible to traditional inventories, and this does not automatically mean that cities are emitting more methane than before: “We are seeing additional, more specific and richer data, but this does not necessarily mean that cities are emitting more methane overall.” Regardless of the methodology, Sarfatti confirms that methane reduction is one of the C40 network’s key goals and that work on landfill emissions has been ongoing for years. “Methane emissions are around 87 times more potent than CO₂ in the first twenty years. They are the drivers of global warming in the short term. So we are pleased that there is this attention and focus.”
The network, explains Sarfatti, is primarily investing in the exchange of best practices between cities. In Europe, Milan and Vienna have been rivals for years for the top spot in organic waste collection. But it is the Global South that is showing the most interesting developments: Mexico City has improved its organic waste collection by 50%, while in Ghana, after the closure of landfills in Accra, C40 is working to re-employ informal workers in cooperatives focused on the ecological transition. “Waste management is the main front,” summarises Sarfatti. “And the challenge is to achieve this without leaving anyone behind.”
High-resolution data is required to manage emissions
The conclusion drawn from the comparison between the Michigan study and the C40 response is clear, despite the two parties phrasing it differently: without a monitoring system based on direct observation, urban climate policies risk resting on shaky ground. Whether it is a case of underestimated emissions or technology revealing something that was previously invisible, the bottom line is the same: cities need better data to take effective action.
The problem, after all, is not limited to methane. “On a similar topic – which is high-resolution emissions estimation, rather than using average data that risks being skewed, as happened with methane – we have just published an article in Nature Sustainability,” explains Umberto Fugiglando, head of research strategy and partnerships at MIT’s Senseable City Lab.
The study, authored by Songhua Hu, Paolo Santi, Carlo Ratti and other researchers from the laboratory, introduces a framework that integrates traffic camera imagery and mobile phone data to estimate vehicle emissions on a street-by-street, hour-by-hour basis. Applied to Manhattan, the method demonstrated that omitting high-resolution variables – such as traffic lights, speed variations or fleet composition – introduces errors averaging between -49% and +25% in emissions estimates. The same tool was used to assess the effects of Manhattan’s congestion pricing scheme: in the first eight weeks, traffic fell by 10% and emissions by 16–22%.
“Data is primarily used to measure and help make more informed decisions,” emphasises Fugiglando, who, at the MAUTO in Turin as part of the Biennale OFF programme, explained how MIT is also working on the same principle in Milan: by using data from the black boxes of tens of thousands of cars provided by UnipolTech, the team demonstrated that the introduction of 30 km/h zones leads to an increase in journey time of between 2 and 30 seconds per average trip but reduces accidents and fatalities. “In environmental terms, we did not find any major differences: emissions do not rise or fall significantly. It is clear that if this then leads to a redesign of the city that favours alternative mobility, there would be a real decrease in emissions.”
It is the same principle that emerges from the study published in PNAS: real measurements, not self-reported estimates, are needed to guide decisions. And the technology – satellite, sensor and algorithmic – is already available. All that remains to be built is the political will to translate it into binding strategies. As Fugiglando points out, citizens should be able to access this data through a genuine “democratic process, so that they can see these options and decide which future they prefer”.
Cover: View of the city of Beppu, Japan; Envato image