Sage Journals: Discover world-class research

Abstract

Carbon & Place (https://www.carbon.place) is an ongoing research project to produce a free family of web tools intended to explain the spatial variation in per-capita carbon footprints across Great Britain and how they can be reduced. The tools present results via interactive maps using GIS data, small area statistics, surveys, and models to aid planners, policymakers, and communities in understanding their climate impact. Local people can benefit from disaggregated analysis as it can be more personally relevant and account for local circumstances and needs. This paper provides an overview of the project, its open-source website, and analysis pipeline, as well as reporting on its progress to date and future work.

Keywords

housing transport networks geodemographics carbon footprint synthetic population

Introduction

Carbon footprints are a widely used method for attributing responsibility for emissions to groups or individuals (Chen et al., 2021; Pandey et al., 2011). They are commonly calculated for large geographic areas (countries/regions/cities) (Minx et al., 2013; Moran et al., 2018) or groups and individuals (companies/households/individuals) (Ottelin et al., 2019; Steen-Olsen et al., 2016; Weber and Matthews, 2008; Wiedenhofer et al., 2017). However, the intermediate scale of neighbourhoods or communities is far less studied (Heinonen et al., 2020). This is a problem because neighbourhood scale characteristics can significantly affect parts of the carbon footprint, and there is an emerging research agenda around how place-based solutions are more effective at achieving decarbonisation goals (Lai et al., 2025; Middlemiss et al., 2024). For example, building properties and travel patterns closely relate to neighbourhood history and characteristics (Marsden et al., 2020). Furthermore, local governments are often on the front line of delivering decarbonisation policy and, as such, need to understand how and where to target their interventions (Bedford et al., 2023; McMillan et al., 2024).

While there are many papers that establish the relationship between spatial characteristics (such as urbanity) and carbon footprints (Heinonen et al., 2020), only a few attempts have been made to produce neighbourhood level carbon footprints for nations or regions (Dawkins et al., 2024; Jones and Kammen, 2014; Morgan et al., 2021). Thus, there is a gap between the understanding that spatial characteristics are important for understanding carbon footprints, and the creation of carbon footprint datasets that can be used by communities and policymakers.

Carbon & Place seeks to fill this gap for Great Britain with a group of free open-source web tools (https://www.carbon.place) that explore the spatial variation in per-capita consumption-based carbon footprints. These tools are produced from an open-source and reproducible analysis pipeline, and outputs are (or soon will be) published as open data. The project will run from 2024 to 2028, and this paper reports on initial results, ongoing research, and future plans. Carbon & Place expands the scope of earlier work in several ways: by going into more detail on selected topics with multiple tools; by including Scotland and Wales; by considering historical change (2010–present); and by exploring future scenarios.

Methodology

Calculating a consumption-based carbon footprint of a neighbourhood has several challenges. First, everybody has a carbon footprint, and everything we do contributes to our carbon footprint. Thus, the tool is encumbered with an impossibly broad scope of everyone and everything in Great Britain. Second, while the UK has a lot of small area datasets, there is insufficient data to know everything about everyone for obvious privacy and practical reasons. Therefore, a two-tier approach must be taken, whenever possible, neighbourhood specific data is used to calculate parts of the carbon footprint. For parts of the footprint that data is unavailable, we resort to national surveys of household consumption but construct sub-samples of the survey designed to be representative of each neighbourhood.

Thus, a patchwork of data and analysis with varying levels of complexity and spatial detail is assembled into a common unit of the average per-capita consumption carbon footprint for each Lower Super Output Area (LSOA) in Great Britain for each year since 2010.

Input data

Input data comes from a broad range of sources listed in Tables 1 and 2.¹

Table 1.

Selected input data in the form of small area statistics of datasets with a high geographical resolution. These data are used mainly to measure activity or characteristics directly at the neighbourhood scale.

Name and description	Source
Boundaries: OA/LSOA/MSOA for 2011/2021	(ONS, 2024a)
Wards, parishes, local authority	(Digimap, 2014)
Consistencies
Scottish data zones 2011/2022
Postcodes
Ordnance survey open data	(Ordnance Survey, 2014)
Ordnance survey points of interest	(Ordnance Survey, 2024)
OpenStreetMap	(Geofabrik, 2021)
LIDAR digital terrain/surface maps: England	(Environment Agency, 2024)
Wales	(Welsh Government, 2024)
Scotland	(Scottish Government, 2024a)
Census 2011/2021 (multiple datasets)	(ONS, 2011)
2011/2022 LSOA area classifications	(ONS, 2018)
Boundaries: OA/LSOA/MSOA for 2011/2021	(ONS, 2024a)
Wards, parishes, local authority	(Digimap, 2014)
Consistencies
Scottish data zones 2011/2022 postcodes
Mid-year population estimates	(ONS, 2024b)
Mid-year population estimates	(National Records of Scotland, 2021)
Council tax: England/Wales	(VOA, 2024)
Scotland	(Scottish Government, 2024b)
Energy performance certificates	(MHCLG, 2024)
Energy performance certificates	(Energy Savings Trust, 2024)
Gas and electricity consumption LSOA (England/Wales)	(DESNZ, 2024)
Postcode (GB)	(DESNZ, 2024)
Income (MSOA) 2012–2020 (England/Wales)	(ONS, 2023a)
Land ownership by UK/overseas companies	(Land Registry, 2022)
INSPIRE freehold polygons	(Land Registry, 2024)
House prices and transactions (England/Wales)	(O’Brien, 2024)
Propensity to cycle tool	(Goodman et al., 2019)
Public transport frequency	(FOE, 2023; Morgan and Philips, 2023)
Vehicle registrations	(DFT, 2024)

Table 2.

Selected input data in the form of national or aggregate data, which is used to establish general consumption patterns when local data is unavailable.

Name and description	Source
UK consumption-based accounts	(ONS, 2023b)
Emissions factors 2010–present	(BEIS, 2022a)
Living costs and food survey	(ONS, 2024c)
Understanding society	(Institute for Social and Economic Research, 2024)
National travel survey	(DfT, 2023a)
Anonymised MOT tests	(Cairns et al., 2017; DfT, 2023b)
UK flights and emissions 1990–2021	(Morgan et al., 2025)

Analysis

The analysis is performed in the R programming language (R Core Team, 2025) using the targets framework for reproducible research (Landau, 2021) and is published on GitHub (https://github.com/PlaceBasedCarbonCalculator/build/). This paper is too short to fully describe the complete analysis method which includes over 300 stages²; however, Figure 1 provides a highly simplified overview of the major steps in producing the per-capital carbon footprint for each LSOA and year from 2010–2022. A comparable workflow will be employed to update the outputs each year as new input data becomes available. Many of the input datasets are published on a 2–3 year time lag, and several datasets were affected by the COVID-19 pandemic. Therefore, at time of writing there is a larger than usual delay for the most recent year for which a comprehensive analysis can be performed.

Figure 1.

A simplified flowchart of the main analytical steps to producing the per-capita emissions estimates. Input datasets are shown in orange correspond to datasets listed in Tables 1 and 2. For clarity preliminary stages such as data cleaning and preparation are not shown. Datasets with a dashed boundary indicate that precursor analysis accounts for changes in LSOA boundaries and population change over time have been done. In this figure LSOA is used to refer to LSOAs in England and Wales and Data Zones in Scotland, for simplicity the harmonisation of data across the three countries is not shown. Important intermediate datasets are shown in blue, and major outputs are shown in green. Data visualisations available on the website are in purple.

The remainder of this section provides detail on some common techniques which can illustrate the approach to calculating carbon footprints. More detail is provided in the website’s manual and GitHub repositories.

Adjusting for different boundaries, boundary changes, and population change

As the common unit of analysis is the 2021 LSOA (or Data Zone in Scotland), all datasets are adjusted to match this unit via aggregation, disaggregation, or transformation.

Aggregation includes summarising individual point data (e.g. building Energy Performance Certificates) or postcode data using weighted sums. In most cases, a point intersection is used to assign data to the right LSOA. Simple examples include mapping the most common housing type for each LSOA by aggregating EPCs, or producing a pie chart of house type within the report card.

The primary type of transformation is converting 2011 LSOA boundaries to 2021 LSOA boundaries. Fortunately, most LSOAs in England and Wales did not change their borders between Censuses, so no change is needed. However, some LSOAs were split, merged, or underwent a complex change, within Scotland most of the 2011 Data Zones were replaced in the 2022 Census resulting in a more complex transformation.

Only six LSOAs experienced a complex change, and these were small adjustments, such as the example in Figure 2. For this analysis, a one-to-one paring was established between the precursor and successor LSOAs. While this introduces some errors in the analysis, it is likely much smaller than other uncertainties in the analysis and only affects a few locations. Merged LSOAs are simple to account for, as any pre-2021 data can be added to get values for the new LSOAs. When historical averages are reported (e.g. average age), a population-weighted mean is used to merge data across LSOAs.

Figure 2.

An example of the complex, but fairly insignificant boundary changes that affect a small number of LSOAs between 2011 (red) and 2021 (black). In this case 12 houses out of hundreds have moved from one LSOA to its neighbour, representing 1% of the population of the LSOA. Base map from OpenStreetMap.

Split LSOAs are the most common type of changed LSOA due to population growth and the most complex to analyse. Three metrics were used to ascertain how historical data should be split between LSOAs. The number of households in 2011 and 2021 according to the census, the number of adults in each year according to the ONS mid-year population estimate (2011 boundaries), and the number of dwellings in each year according to the Valuation Office Agency (VOA) (2021 boundaries). It was assumed that the number of households in any given year was proportional to the number of dwellings and a linear interpolation of the number of adults per household between 2011 and 2021. Thus, for each year, the number of households was estimated, and the ratio of households in a given year was used to split that year’s data (such as electricity consumption) between the new LSOAs.

As shown in Figure 1 by the dashed boarder, many datasets have had to be adjusted prior to the analysis to enable a common unit of analysis across multiple years. The annual estimates of population, dwellings, and households in each LSOA are also used to normalise emissions and other variables to a per-capita or per-household basis. It is important that these estimates are correct, because under/over estimation of the population can significantly distort emissions calculations and give a misleading impression of where emissions are rising or falling (Morgan et al., 2022).

Calculating parts of the carbon footprint when local data is available

For some parts of household carbon footprints (e.g. domestic gas and electricity consumption), long-time series small area statistics exist (see Table 1), subject to boundary transformations (see above). Thus, total emissions can easily be calculated by multiplying consumption (e.g. kWh gas/electric) by published carbon intensities for each year (BEIS, 2022b). Total emissions are divided by mid-year LSOA population to estimate per-person carbon footprints.

For surface transport emissions, a mix of small area data about car ownership and average mileage in some years is combined with disaggregated data to model the spatial variation and emissions rates of vehicles in each LSOA. Public transport emissions are modelled in less detail, but as they are a small part of total emissions, they make a minor difference to the results.

Calculating parts of the carbon footprint when local data is unavailable

Disaggregated consumption data is unavailable for all parts of the carbon footprint, such as food, goods, and services. Local consumption is modelled for these kinds of activities using survey data and synthetic population approaches (Morgan, 2025; Wu et al., 2022). These models assume that the socio-demographic and geodemographic consumption patterns observed in survey data, such as the strong correlation between income and consumption (Owen and Barrett, 2020), are the best predictors of local consumption patterns.

Two years of the Living Costs and Food Survey (LCFS) are used to provide detailed spending data for 12,000 households across the 34 Classification of Individual Consumption According to Purpose (COICOP) (ONS, 2024c). The COICOP categories in the LCFS match the categories in the UK national consumption-based carbon accounts; thus, national carbon emissions can be distributed to LSOAs in proportion to household spending in each area.

To ascertain which subset of the LCFS best represents each LSOA a synthetic population of households has been constructed from the 2021/22 Censuses. The synthetic population was created by Iterative Proportional Fitting (IPF) using the R package mipfp and the five variables listed in Table 3. The multivariate tables listed in Table 4 were used as the input data for the IPF algorithm. The ONS excludes LSOAs where the multivariate tables would be disclosive; hence, some LSOAs only have access to partial data. Two additional intermediate variables (Accommodation Type,³ Household Composition (6 Variables)) were used as they helped differentiate households when other data was missing. The IPF algorithm was seeded with the multivariate national population distribution reducing the creation of impossible or unlikely households (e.g. a couple in a one-person household).⁴

Table 3.

Variables and possible values used to characterise each household in the synthetic population and link small area Census data to the Living Costs and Food Survey.

Variable	Possible values
Tenure	Owned outright; owned with mortgage; social rent; private rent/rent free
Household composition	One person >=66; one person <66; couple without children; couple with dependent children; couple with non-dependent children; lone parent; lone parent with non-dependent children; family >=66; other with children; other without children; other including >=66 or students
Number of cars & vans	0; 1; 2; 3 or more
Household size	1; 2; 3; 4 or more
Output area classification	Rural-urban fringe; affluent rural; rural growth areas; larger towns and cities; university towns and cities; older farming communities; Sparse English and Welsh countryside; ageing coastal living; seaside living; Scottish countryside; ethnically diverse metropolitan living; London cosmopolitan; manufacturing legacy; mining legacy; service economy; Scottish industrial legacy; country living; Northern Ireland countryside; prosperous semi-rural; prosperous towns; industrial and multi-ethnic; urban living; city periphery; expanded areas

Table 4.

Multivariate input tables from the 2021 Census used to construct the synthetic population and the percentage of LSOAs (England/Wales) for which data was available.

Variable 1	Variable 2	Variable 3	Coverage
Accommodation type	Tenure		100%
Household composition (15 variables)	Tenure		100%
Household composition (6 variables)	Tenure		100%
Household size	Tenure		100%
Household composition (6 variables)	Cars & vans		100%
Cars & vans	Tenure		100%
Cars & vans	Household size		99.9%
Household composition (6 variables)	Household size		99.8%
Cars & vans	Accommodation type		99.1%
Household composition (6 variables)	Accommodation type		99.0%
Accommodation type	Household size		99.0%
Cars & vans	Tenure	Household size	98.5%
Household composition (6 variables)	Tenure	Cars & vans	91.5%
Household composition (15 variables)	Accommodation type		58.0%

In addition to the five variables in Table 3, household income is used to improve the quality of the match. For income data a normal distribution is created based on the average income and confidence intervals published by the ONS (ONS, 2023a). The normal distribution is used to weight the selection of households ensuring that selected households are within the 99% confidence intervals for household incomes in each LSOA.

For all variables, a match is found by maximising a similarity score. Each pair of values is weighted in terms of how similar they are (e.g. Detached to Semi-Detached = 0.8). For 56.9% of households, it is possible to obtain a perfect match across all six variables, and only 2.2% of households get a match score of less than 95%, indicating that in most cases a single small compromise has been made.⁵ The lowest score of 80% affects just one of the 24.6 million households in the synthetic population.

Flights are a significant part of some people’s carbon footprint and are treated separately to other types of transport and consumption. Flights are a difficult part of the carbon footprint to account for as there are no local datasets that record how much people fly. Even national datasets such as the National Travel Survey do a poor job of capturing flights as infrequent trips, such as holidays, are often missed in surveys that focus on daily or weekly travel (Wadud et al., 2024). Furthermore, official emissions calculations of international flights can be misleading. For example, standard practice is to only report flights leaving a country as being part of national carbon emissions and to calculate only CO₂ emissions rather than CO₂ equivalent emissions used for other parts of the carbon footprint. This is incompatible with household consumption emissions which should account for the total impact off all flights taken by the household (Büchs and Schnepf, 2013). So more detailed passenger origin-destination data is used to estimate total emissions (Morgan et al., 2025). The LCFS records spending on flights, but it only covers a brief period (3 months) so instead the total number of flights in the past 12 months is used as the metric to apportion emissions to LSOAs. As the LCFS distinguishes between domestic and international flights those can be apportioned separately. Unfortunately, the LCFS does not distinguish between short-haul and long-haul flights which is a significant omission from an emissions perspective.

Future scenarios

The ultimate goal of the Carbon & Place analysis is to produce local decarbonisation scenarios for each LSOA in Great Britain. This work will build on other projects currently ongoing within the Place and Futures themes in Energy Demand Research Centre. The general approach to producing these local scenarios is to downscale the Positive Low Energy Futures (PLEF) scenarios (Brand et al., 2022). The PLEF are a range of national scenarios for the UK to achieve net-zero by 2050. The aim is to produce local decarbonisation scenarios that are consistent with PLEF that recognise the pre-existing differences between neighbourhoods and how that will affect the speed and method that neighbourhoods achieve net-zero. Thus, it should be possible to say at the neighbourhood scale what changes need to occur, when they need to be finished, and whether the neighbourhood is currently on-track to achieve net-zero. This final output can then be used by communities, planners, and policymakers to design, deliver, and evaluate net-zero policies.

Validation

Validation of the final results is challenging as no directly comparable small area emissions statistics exist to compare against. Thus, most validation checks have focused on internal consistency, for example, when national emissions have been downscaled to LSOAs, we check the sum of all LSOAs matches the national total.

It is also possible to cross check variables in the Synthetic Population to ensure that the characteristics match what is observed in the real population. Figure 3 shows the synthetic population fairly accurately reflects the income distribution across the country (R² = 0.96); however, it does tend to underestimate incomes in the wealthiest areas due to a shortage of high-income households in the LCFS.

Figure 3.

Observed and modelled average household income for MSOAs in England and Wales in 2020. The red line indicates perfect match, the blue line is trendline of the modelled data.

Figure 4 shows the high correlation between the number of households recorded within each LSOA with specific combinations of characteristics in the Census and the Synthetic Population. It is not possible to perfectly reconstruct the Census due to the ONS adding a small amount of random error into the published results to protect privacy. This results in small disagreements between the inputs listed in Table 4 in both the size of individual group and the total population of LSOAs. When there was disagreement in the total population of the LSOA, the IPF algorithm was set to target the median population of the input datasets; thus, small errors are unavoidable.

Figure 4.

An example of an internal validity check of the Synthetic Population. In this case observed (Census 2021) and modelled (Synthetic Population) number of households in each LSOA are compared. Each point represents a unique combination of LSOA, household composition, tenure, and number of cars. Which is known for most LSOAs (as described in Table 4). For clarity, the graph has been limited to values less than 300 households excluding 130 out of 1,741,076 points. This shows there is a strong correlation between the Synthetic Population and the Census.

Under-represented groups such as students, care home residents, and prisoners are also less likely to be accurately modelled by the method. How to best distribute survey respondents and address under-represented groups is an ongoing area of research.

One aspect that is more comparable between the small area data and the Synthetic Population is the consumption of gas and electricity. As we have good data on domestic gas and electricity consumption it is not necessary to use the Synthetic Population to predict gas and electricity consumption. However, the spending data exists in the LCFS so an estimate of energy bills can be made and is shown in Figure 5.

Figure 5.

A comparison of estimated average annual household gas and electricity bills for each LSOA in 2021, based on aggregated metre readings and the Synthetic Population.

In some ways Figure 5 is an unfair test for the Synthetic Population as we are using it to make an absolute prediction of energy bills while for the other types of consumption only the relative consumption is needed to apportion emissions. As the Synthetic Population was not designed to predict domestic energy use it lacks access to relevant variables such as housing type and size that are known to correlate with energy consumption.⁶ Furthermore, the comparison data based on metre readings only provided energy consumption in kWh and we have had to make price assumptions about unit rates and standing charges. Nevertheless, the Synthetic Population almost perfectly predicts the average energy bill in 2021 (£1,197 vs £1,253) and demonstrates a weak but statistically significant (p < 0.001) positive correlation with observed energy consumption across different LSOAs.

Figure 6 tests the Synthetic Population against a range of other variables that are in both the Census and the LCFS. These variables were not used to constrain the Synthetic Population as they are less relevant to carbon emissions and have practical issues such as higher rates of missing data or inconsistent categorisation and measurement across the two datasets. Nevertheless, that the Synthetic Population can produce statistically significant (p < 0.001) predictions across all these variables demonstrates the high degree of collinearity between geodemographic characteristics and supports the hypothesis that synthetic populations can identify the spatial distribution of carbon footprints.

Figure 6.

Comparisons between the Synthetic Population and the 2021 Census. Top left: The average number of rooms per household in each LSOA, note that Synthetic Population is based on self-reported data from the LCFS 2018/19, while the Census 2021 used official data from the Valuation Office Agency. Top right: The percentage of household reference persons in four common types of employment status for each LSOA (employee – red, self-employed – blue, unemployed – green, retired – orange). Bottom left: The percentage of household reference persons with three common ethnicities for each LSOA (white – red, Asian – blue, black – green). Bottom right: The percentage of households in the Synthetic Population from rural or urban areas in each LSOA. A dark red shows an LSOA is urban and most of the respondents from the LCFS selected to represent it lived in an urban area. Conversely a light green shows that the LSOA is rural and only a minority of the selected LCFS households live in a rural area.

Results

Website

The website’s primary function is to provide access to the analysis and data in a format that is accessible to non-technical users. As such Carbon & Place differs from some open science projects in that data is not the only end product, visualisation, and communication are integral parts of the process. This section outlines some of the conceptual features of the website and technical details.⁷ Maplibre, Tippecanoe, and PMtiles are the key technologies that enable the creation of the interactive maps (Fisher, 2024; Maplibre, 2024; Protomaps, 2024).⁸

A custom 2 m terrain and surface model was built using open LIDAR data⁹ and combined with OS and OSM building outlines to build a national 3D building stock model. OSM buildings tend to be more detailed than OS Open Data but have limited coverage. The building outlines were also segmented with the Land Registry INSPIRE polygons. This helped split up large buildings when, for example, a row of terraced homes was represented by a single polygon.

The creation of a custom buildings layer enables the use of dasymetric mapping (Figure 7) which highlight that the data refers to the buildings and/or occupants rather than the areas and aids in navigation by making the roads and buildings more visible.

Figure 7.

Example of the 3D buildings built from open data and used to produce dasymetric maps. In this case neighbourhood per-capita carbon footprints from blue (below average) to red (above average) are shown for part of London (from Stratford looking towards the City of London). For ease of understanding by non-technical users, emissions are converted into grades (A+ to F−) which corresponds to percentile bands. With the national median between C−/D+ and A+/F− corresponding to the best/worst 1% of neighbourhoods.

The main neighbourhood maps are supplemented with other optional map layers (e.g. points representing individual building Energy Performance Certificates) and the pop-up reports (Figure 8). Reports are displayed when clicking on any neighbourhood and are used to convey more detailed information on a specific area.

Figure 8.

Examples from the pop-up report. Top: graph shows changes in the number of dwellings for the neighbourhood selected by the user. Bottom: A range of demographic data is summarised into a community photo illustrating the mix of people who live in the selected neighbourhood. These examples show simple visualisations that makes existing open data more accessible to non-technical users.

The Carbon & Place website is broken into multiple tools which focus on different topics (e.g. Transport and Accessibility, Retrofit, Land Ownership) but share a common back end. The website is updated every 6 months with new features and updated data, and updates will continue until at least the end of EDRC in 2028.

Bulk data

The website also provides bulk data downloads from the data page (https://www.carbon.place/data). Both finalised outputs and experimental outputs from Beta tools are provided as open data. Experimental datasets are typically work in progress and subject to revision before final release. Additional information about the dataset and methods are provided in the website manual (https://www.carbon.place/manual) and the associated GitHub repositories (https://github.com/PlaceBasedCarbonCalculator).

Conclusion

This paper has given a brief overview of the Carbon & Place project and tools. They are intended to provide an in-depth and locally specific understanding of carbon footprints and decarbonisation pathways across Great Britain. While the challenges to calculating local carbon footprints are significant the tools demonstrate that it is possible, and the thousands of monthly uses suggest that it is a useful exercise for communities and policy makers. The use of a broad range of open data combined with the targets reproducible analysis framework provide transparency around methods and assumptions which enables policy makers and researchers to build their own work on top of the current and future results.

Footnotes

ORCID iD

Malcolm Morgan

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: the authors gratefully acknowledge support from UK Research and Innovation through the Centre for Research into Energy Demand Solutions; grant reference number EP/R 035288/1 and Energy Demand Research Centre; grant reference number EP/Y010078/1.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The tools described in this paper are freely accessible at https://www.carbon.place and the underlying code is published at data is available via the website and GitHub repos. Long term archiving will be done via UK Data Service at the end of the project.

Notes

Author biography

Dr Malcolm Morgan is a Senior Research Fellow in Transport and Spatial Analysis at the University of Leeds, Institute for Transport Studies. His research focuses on the spatial distribution of energy use and carbon emissions, with a particular focus on transport and housing. He works on understanding both why energy use changes between neighbourhoods and whether this is due to choice or circumstances. His research often emphasizes producing practical tools to aid in policy-making such as, Propensity to Cycle Tool (www.pct.bike), Place-Based Carbon Calculator (www.carbon.place), Cycling Infrastructure Prioritisation Toolkit (). He mostly uses quantitative approaches focusing on open source GIS and big data analysis and visualization.

References

Bedford

Catney

Robinson

(2023) Going Down the local: the challenges of place-based net zero governance. Journal of the British Academy 11: 125–156.

BEIS (2022a) Final UK greenhouse gas emissions national statistics: 1990 to 2020. https://www.gov.uk/government/statistics/final-uk-greenhouse-gas-emissions-national-statistics-1990-to-2022 (accessed 19 11 25)

BEIS (2022b) Government conversion factors for company reporting of greenhouse gas emissions. https://www.gov.uk/government/collections/government-conversion-factors-for-company-reporting (accessed 8 16 22).

Brand

Anable

Marsden

(2022) The role of energy demand reduction in achieving net-zero in the UK: transport and mobility. In: ECEEE 2022 Summer Study proceedings, June 2022, Hyeres, France. https://www.researchgate.net/publication/360189100_The_role_of_energy_demand_reduction_in_achieving_net-zero_in_the_UK_Transport_and_mobility

Büchs

Schnepf

(2013) Who emits most? Associations between socio-economic factors and UK households’ home energy, transport, indirect and total CO2 emissions. Ecological Economics 90: 114–123.

Cairns

Anable

Catterton

, et al. (2017) Motoring along: the lives of cars seen through licensing and test data. RAC Foundation. https://www.racfoundation.org/research/environment/motoring-along-the-lives-of-cars-seen-through-licensing-and-test-data

Chen

Zhang

Han

(2021) Where has carbon footprint research gone? Ecological Indicators 120: 106882.

Dawkins

Rahmati-Abkenar

Axelsson

, et al. (2024) The carbon footprints of consumption of goods and services in Sweden at municipal and postcode level and policy interventions. Sustainable Production and Consumption 52: 63–79.

DESNZ (2024) Sub-national gas consumption data. DESNZ. https://www.gov.uk/government/collections/sub-national-gas-consumption-data

10.

DfT (2023a) National Travel Survey. DfT.

11.

DfT (2023b) Anonymised MOT Tests and Results. DfT. https://www.data.gov.uk/dataset/e3939ef8-30c7-4ca8-9c7c-ad9475cc9b2f/anonymised-mot-tests-and-results (accessed 3.29.23).

12.

DFT (2024) Vehicle Licensing Statistics Data Tables. DfT. https://www.gov.uk/government/statistical-data-sets/vehicle-licensing-statistics-data-tables (accessed 11.20.24).

13.

Digimap (2014) Digimap help. https://digimap.edina.ac.uk/webhelp/digimapsupport/about.htm (accessed 8 26 14).

14.

Energy Savings Trust (2024) Scottish energy performance certificate register. https://www.scottishepcregister.org.uk/CustomerFacingPortal/DataExtract (accessed 11 20 24).

15.

Environment Agency (2024) National LIDAR programme. https://www.data.gov.uk/dataset/f0db0249-f17b-4036-9e65-309148c97ce4/national-lidar-programme

16.

Fisher

(2024) Tippecanoe. https://github.com/felt/tippecanoe

17.

FOE (2023) How Britain’s bus services have drastically declined. https://policy.friendsoftheearth.uk/insight/how-britains-bus-services-have-drastically-declined (accessed 11 28 23).

18.

Geofabrik (2021) Europe. https://download.geofabrik.de/europe.html

19.

Goodman

Rojas

Woodcock

, et al. (2019) Scenarios of cycling to school in England, and associated health and carbon impacts: application of the ‘Propensity to Cycle Tool’. Journal of Transport & Health 12: 263–278.

20.

Heinonen

Ottelin

Ala-Mantila

, et al. (2020) Spatial consumption-based carbon footprint assessments - a review of recent developments in the field. Journal of Cleaner Production 256: 120335.

21.

Institute for Social and Economic Research (2024). Understanding Society. Institute for Social and Economic Research.

22.

Jones

Kammen

(2014) Spatial distribution of U.S. household carbon footprints reveals suburbanization undermines greenhouse gas benefits of urban population density. Environmental Science & Technology 48: 895–902.

23.

Lai

H-L

Devine-Wright

Hamilton

, et al. (2025) A place-based, just transition framework can guide industrial decarbonisation with a social licence. Energy Research & Social Science 121: 103967.

24.

Land Registry (2022) UK companies that own property in England and Wales. https://www.gov.uk/guidance/hm-land-registry-uk-companies-that-own-property-in-england-and-wales (accessed 6.1.24).

25.

Land Registry (2024) INSPIRE index polygons spatial data. https://www.gov.uk/guidance/inspire-index-polygons-spatial-data

26.

Landau

(2021) The targets R package: a dynamic make-like function-oriented pipeline toolkit for reproducibility and high-performance computing. Journal of Open Source Software 6: 2959.

27.

Maplibre (2024) MapLibre GL JS. https://github.com/maplibre/maplibre-gl-js

28.

Marsden

Anderson

Büscher

, et al. (2020) Place-Based Solutions for Transport Decarbonisation, Submission to the Department for Transport’s Consultation on the Transport Decarbonisation Plan. Leeds.

29.

McMillan

Barnes

Nolden

, et al. (2024) Local, place-based governance for net zero: a review and research agenda. Journal of the British Academy 12. doi: 10.5871/jba/012.a47

30.

MHCLG (2024) Energy performance of buildings data: England and Wales. https://epc.opendatacommunities.org/ (accessed 11 20 24).

31.

Middlemiss

Snell

Theminimulle

, et al. (2024) Place‐based and people‐centred: principles for a socially inclusive net zero transition. Geo: Geography and Environment 11: e00157.

32.

Minx

Baiocchi

Wiedmann

, et al. (2013) Carbon footprints of cities and other human settlements in the UK. Environmental Research Letters 8: 035039.

33.

Moran

Kanemoto

Jiborn

, et al. (2018) Carbon footprints of 13,000 cities. Environmental Research Letters 13: 064041.

34.

Morgan

(2025) A synthetic population for small area household consumption and emissions in England and Wales. In:

GISRUK

2025, 23-25 Apr 2025, Bristol, UK. https://eprints.whiterose.ac.uk/id/eprint/229962/

35.

Morgan

Philips

(2023) Long-time series, small-area statistics for car ownership and level of public transport service. In: 2023 UTSG Conference: papers & extended abstracts. UTSG 55th Annual Conference, 10-12 Jul 2023, Cardiff, UK. Universities' Transport Study Group (UTSG) https://eprints.whiterose.ac.uk/id/eprint/220131/

36.

Morgan

Anable

Lucas

(2021) A place-based carbon calculator for England. In: 9th annual GIS research UK conference (GISRUK), Cardiff, 13th-16th April 2021, 1–5.

37.

Morgan

Morton

Monsuur

, et al. (2022) Understanding Change in Car Use Over Time (UnCCUT): End of Project Report. Leeds. https://decarbon8.org.uk/wp-content/uploads/sites/59/2023/01/Understanding-Change-in-Car-Use-over-Time-Morgan-et-al-2022.pdf

38.

Morgan

Wadud

Cairns

(2025) Can rail reduce British aviation emissions? Transportation Research Part D: Transport and Environment 138: 104513.

39.

National Records of Scotland (2021) Mid-2020 small area population estimates for 2011 data zones. https://www.nrscotland.gov.uk/statistics-and-data/statistics/statistics-by-theme/population/population-estimates/small-area-population-estimates-2011-data-zone-based/mid-2020

40.

ONS (2011) 2011 census data on nomis. https://www.nomisweb.co.uk/census/2011 (accessed 5 15 20).

41.

ONS (2018) About the area classifications. https://www.ons.gov.uk/methodology/geography/geographicalproducts/areaclassifications/2011areaclassifications/abouttheareaclassifications (accessed 4 20 20).

42.

ONS (2023a) Income estimates for small areas, England and Wales statistical bulletins. https://www.ons.gov.uk/peoplepopulationandcommunity/personalandhouseholdfinances/incomeandwealth/bulletins/smallareamodelbasedincomeestimates/previousreleases (accessed 11 20 24).

43.

ONS (2023b) UK environmental accounts: 2023. https://www.ons.gov.uk/economy/environmentalaccounts/bulletins/ukenvironmentalaccounts/2023

44.

ONS (2024a). Open geography portal. https://geoportal.statistics.gov.uk/

45.

ONS (2024b) Lower layer super output area population estimates. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/lowersuperoutputareamidyearpopulationestimatesnationalstatistics (accessed 11 20 24).

46.

ONS (2024c) Living Costs and Food Survey. [data series]. 4th Release. UK Data Service. SN: 2000028, https://doi.org/10.5255/UKDA-Series-2000028

47.

Ordnance Survey (2014) OS OpenData products. https://www.ordnancesurvey.co.uk/business-and-government/products/opendata-products-grid.html (accessed 8 26 14).

48.

Ordnance Survey (2024) Points of interest. https://www.ordnancesurvey.co.uk/products/points-of-interest (accessed 11 20 24).

49.

Ottelin

Heinonen

Nässén

, et al. (2019) Household carbon footprint patterns by the degree of urbanisation in Europe. Environmental Research Letters 14: 1. doi: 10.1088/1748-9326/ab443d

50.

Owen

Barrett

(2020) Reducing inequality resulting from UK low-carbon policy. Climate Policy 20: 1193–1208.

51.

O’Brien

(2024) Dwelling ages and prices. CDRC. https://data.cdrc.ac.uk/dataset/dwelling-ages-and-prices (accessed 11 20 24).

52.

Pandey

Agrawal

Pandey

(2011) Carbon footprint: current methods of estimation. Environmental Monitoring and Assessment 178: 135–160.

53.

Protomaps (2024) PMTiles. https://github.com/protomaps/PMTiles

54.

R Core Team (2025) R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing. Available at: https://www.R-project.org.

55.

Scottish Government (2024a) Scottish remote sensing portal. https://remotesensingdata.gov.scot/data#/list (accessed 11 20 24).

56.

Scottish Government (2024b) Dwellings by council tax band detailed. https://statistics.gov.scot/data/dwellings-by-council-tax-band-detailed-current-geographic-boundaries (accessed 11 20 24).

57.

Steen-Olsen

Wood

Hertwich

(2016) The carbon footprint of Norwegian household consumption 1999–2012. Journal of Industrial Ecology 20: 582–592.

58.

VOA (2024) Council Tax: stock of properties. https://www.gov.uk/government/statistics/council-tax-stock-of-properties-2024 (accessed 11 20 24).

59.

Wadud

Adeel

Anable

(2024) Understanding the large role of long-distance travel in carbon emissions from passenger travel. Nature Energy 9: 1129. Available at: https://doi.org/10.1038/s41560-024-01561-3

60.

Weber

Matthews

(2008) Quantifying the global and distributional aspects of American household carbon footprint. Ecological Economics 66: 379–391.

61.

Welsh Government (2024) LiDAR viewer. https://datamap.gov.wales/maps/lidar-viewer/ (accessed 11 20 24).

62.

Wiedenhofer

Guan

Liu

, et al. (2017) Unequal household carbon footprints in China. Nature Climate Change 7: 75–80.

63.

Heppenstall

Meier

, et al. (2022) A synthetic population dataset for estimating small area health and socio-economic outcomes in Great Britain. Scientific Data 9: 1–11.

Carbon & Place: Data and tools to understand the spatial variation in carbon footprints

Abstract

Keywords

Introduction

Methodology

Input data

Analysis

Adjusting for different boundaries, boundary changes, and population change

Calculating parts of the carbon footprint when local data is available

Calculating parts of the carbon footprint when local data is unavailable

Future scenarios

Validation

Results

Website

Bulk data

Conclusion

Footnotes

ORCID iD

Funding

Declaration of conflicting interests

Data Availability Statement

Notes

Author biography

References