Does supply (of data on emissions at the urban level) create demand?

Cities are providing significant climate leadership[1], but face difficulties in measuring and reporting emissions with different methods often used. Some current metrics also come with a longer time lag and/or a level of local detail might not be sufficient for local needs. Novel emissions data, including from earth observation or local sensor networks could potentially offer a game-changing technical solution, but are not widely used or adopted. This links to long running debates about how policymakers use research and data, and points to a key challenge for both natural and social scientists.

Resolving this science-policy gap could transform the public communication about emissions targets and policies at subnational level. In this blog we expand on these themes and suggest a way forward which puts the needs of city policymakers at the heart.

Lack of consensus in subnational emissions reporting

At the national level, measurement of emissions is clear and consistent. The conventional measure of emissions inventories has methodological agreement under UNFCCC using the collective science knowledge gathered via the IPCC with verification of these increasingly undertaken.

In the UK, there are local authority territorial estimates of emissions published annually, which use an inventory measure. These provide annual figures on GHGs, such as CO₂ emissions, are released around 18 months after the end of the year to which they relate and with no detail within the local boundaries.

However, there is no consensus on the metric at subnational or city level, with a range of metrics potentially used. Different measures are available and reflect different perspectives on the emission-causing activities which a city is responsible for, whether purely those which occur in a defined geography (“territorial”, or Scope 1, emissions), or those which take place beyond the geographical boundary, perhaps globally, due to consumption of individuals and government within a defined area (consumption-based, or Scope 3, emissions). This was explored in a recent publication as part of the Glasgow Environmental Monitoring of Indoor and Outdoor Air (GEMINOA) project. The upshot is that local decision-makers face a complex landscape of data with complex interpretative implications in terms of what the data does and does not shine a light on.

Technical advances and what is feasible now

At a subnational level, however, there have been considerable technical developments, which offer the possibility to enhance the understanding of urban emissions. Since the first study – published in 2017 and looking at the city of Indianapolis – there have been developments using measures of the concentration of greenhouse gases in the atmosphere. For example, BEACO2N for San Francisco, with a highly dense network of low cost sensors and analysis and modelling (including of weather patterns), can estimate emissions from sources within a specific boundary.  A similarly highly dense network of sensors are used both in Glasgow and Waterloo (Canada), with different approaches in California and across European cities.

Supply does not automatically create demand

While at the scientific frontier, to date, however, there is limited use of these data by policymakers, possibly suggesting that data availability is not sufficient for its use in policy[2]. We discuss some potential reasons for this “gap”.

First, these data are analytically complex and so require expertise both to produce and interpret. The regionalisation of national datasets by comparison has a recognised process and can be readily produced from national level inventories and data on road travel, energy consumption currently available at the national and local level. Such metrics are unable to capture specifics in the vehicle fleet within a specific geography as such data is not readily available at subnational levels.

Second, the resources to produce subnational emissions metrics can be expensive, although costs are coming down rapidly with advances in sensor technologies and modelling. AI use certainly offers possibilities to speed up the computational processes in estimating emissions from concentration measures.

Third, and perhaps most critical, the alignment between existing measures and targets is critical. Targets have been set for a specific measure – employing a particularly methodology – and with clear (and shared) understanding between policy and data teams, internal and external, about the features and its recent trajectory. A measure of the emissions that arise within a specific geography captures some but not all of activities included in emissions inventories. These will align well on elements termed “Scope 1”: those emissions occurring within the area linked to energy consumption such as petrol and diesel for transport use. However, the emissions associated with the generation of electricity which is consumed (but not produced) within a city are Scope 2. These are typically included in emission inventories – and city targets – but would not feature in atmospheric-derived emissions metric.

The factors that explain recent changes in Glasgow’s total CO2 emissions were examined in an FAI report in 2024, and identified the significant role played by improvements in the emissions associated with decarbonisation of the GB electricity grid, which lie well beyond the geography of Glasgow[3]

Fourth, the complexity of calculation makes it potentially challenging to communicate with a public audience, who are an important audience; this is increasingly the case as actions to reduce emissions impact on personal choices. Establishing awareness of measures around net zero and being transparent about the need for and consequences of policy, has been strongly argued as a necessary people-centred approach. Perhaps more local, more spatially explicit environmental data offer an opportunity to directly see how such figures relate to their own experiences. This points to a need for continued, consistent and transparent evaluation of policy impacts.

Finally, as the amount of information available at the urban scale expands with new technologies making available “data assets” which were previously unknown, the appropriate form of data governance – that is the rights, responsibilities and accountabilities for data – becomes critical. With unclear boundaries between the holders of large-scale data, the use of that data in urban governance and the communities to whom that data relates, it is important that questions are posed about what data is gathered, how it is managed and what opportunities users have to interact with data. Some attempts have been made to design real strategies to provide participatory approaches, which can be context-relevant and build data literacy and data democracy.

Ideally, studies that generate new science-based solutions should start with conversation with users/citizens/policymakers. Otherwise, researchers would keep developing things with a focus on their science/technological advancement and trying to find applications later by repurposing their research outcome (then we have further “gaps”). Mostly importantly, citizens & city policymakers need to define together what to achieve (which will reflect the specific urban context) and develop with researchers what solutions are possibly available to provide context-relevant solutions.

 

[1] At the time of writing some 326 cities currently have net zero targets, and are implementing a range of aligned policies.

[2] There could be other uses such as measuring particular local areas below the city-wide scale, as well as interest in these data from users ranging from local community groups, health providers and academic audiences, however we focus on their use by policymakers.

[3]  Emissions from electricity are increasingly set to decline as the GB grid continues to decarbonise. Looking forward, as vehicle electrification gathers pace, it could be the case that emissions within a particular geography decline further, and so cover an increasingly small portion of the missions which are “counted” within an overall total.