Understanding datasets on greenhouse gas emissions

Reducing emissions of greenhouse gases (GHG) is the key to mitigating climate change. It is essential to have reliable data on human-induced GHG emissions to analyse the trend of historical emissions, develop policies to reduce emissions and track the progress towards climate goals.

There are several datasets for GHG emissions and they serve different objectives. For example, there are “production-based” emissions and “demand-based” emissions datasets. The former cover GHG emissions generated from production activities in a country’s economy, whereas the latter (also called “consumption-based emissions” or “carbon footprints”) capture GHG emissions generated along global production chains to meet the demand of a country (Figure 1).

This blog article discusses production-based emissions datasets. A new OECD Working Paper by the OECD’s Statistics and Data Directorate and the Environment Directorate looks in more detail at how such datasets are compiled and provides an overview of official data availability.

GHG emissions inventories and Air Emission Accounts (AEAs) are two of the main datasets for production-based emissions:

• National inventories are compiled by countries and reported to the United Nations Framework Convention on Climate Change (UNFCCC). They are used to define Nationally Determined Contributions (NDCs) under the 2015 Paris Agreement and to track countries’ progress in implementing NDCs.
• AEAs are also compiled by countries, but in this case as part of work to build the accounts of the System of Environmental Economic Accounting (SEEA)¹. They are collected by the European statistical agency Eurostat, the OECD, and the UN Statistical Division. AEA data is suitable for integrating with economic data, for example for calculating emissions intensity per unit of GDP, projecting future emissions trajectories through economic modelling, and calculating carbon footprints.

The inventories and AEAs are compiled using different accounting principles, classification systems and scope of emissions.

Emissions inventories are compiled according to the guidelines of the Intergovernmental Panel on Climate Change (IPCC)² and are based on the territory principle. This means they cover GHG emissions generated by entities within a specific geographical area. By contrast, AEAs are based on the same accounting principles as the international System of National Accounts (SNA), following the residence principle. A resident company is one that has a predominant economic interest in a country. For example, CO₂ emissions from domestic flights in country A by an airline owned by a company based in country B should be attributed to country A if the territory principle is followed and to country B if the residence principle is followed.

Emissions sources are broken down according to different classification systems in the inventories and the AEAs. For the inventories, emissions sources are classified according to the IPCC categories which are based on physical emission processes. On the other hand, AEAs are disaggregated by International Standard Industrial Classification of All Economic Activities (ISIC) and activities of households. Treatment of road transport provides an important contrast: whereas emissions from fuel sold on the territory are categorised as “road transportation” in the inventories (regardless of who may be driving the vehicle), in the AEAs, emissions from road transport are allocated to different industries depending on the principal activity for which the vehicle is being used – for example, if the vehicle is used by a supermarket to transport goods, it would be classified to Section G “Wholesale and retail trade” in ISIC Revision 5.

It is also important to note that the scope of the emissions covered differs between inventories and AEAs. Inventories cover most GHG emissions, but they do not include emissions from international transport or CO₂ emissions from biomass in national totals. AEAs do not include emissions or removals from Land Use, Land Use Change and Forestry (LULUCF), although this may change in future.

Because of these differences, emissions estimates in the inventories and AEAs are not the same. The differences can be explained by introducing bridging items to ensure consistency. Taking total residence-based emissions (excluding CO₂ emissions from biomass) as the starting point, emissions by residents fuelling outside the territory and all international air and maritime transport by resident units are deducted, and emissions by non-residents fuelling on the territory for land transport as well as domestic air and maritime transport are added. From this, we can derive total territory-based emissions excluding LULUCF. In some cases, there are considerable differences. An example is Denmark, which has a large international maritime industry. As Figure 2 shows, most of the difference between the country’s residence- and territory-based emissions estimates is explained by emissions taking place abroad by Danish-owned ships (water transport).

Note: Residents means companies and households that are resident in the country. Non-residents means companies and households that are resident abroad.

Source: OECD Data Explorer, Air emission accounts

The OECD provides official production-based emissions reported by countries and emissions estimates based on official data. These datasets are available from the following links:

• Air emission accounts
• GHG inventories
• Agricultural greenhouse gases emissions

Country coverage, available years, timeliness, sectoral disaggregation, and gases differ across these datasets. The new OECD Working Paper includes an analysis of data availability for each OECD dataset, as well as for the International Energy Agency (IEA) datasets on CO₂ emissions from fuel combustion as part of its database on GHG emissions from energy.

The different GHG emissions datasets complement each other. The OECD has been working to fill the gaps in data availability in coordination with other international organisations, and also to improve the consistency between datasets. Several Directorates of the OECD are currently working together to develop a single, internally consistent GHG emissions database in which it will be possible to reconcile different estimates of the same variable (such as emissions levels) by adjusting for known differences between the datasets, such as the territory principle versus the residence principle. This will support analytical work including tracking progress towards climate change mitigation targets and measuring carbon footprints, as well as contributing to the OECD’s flagship initiative known as the Inclusive Forum on Carbon Mitigation Approaches (IFCMA).

1. The SEEA Central Framework is available at https://seea.un.org/content/seea-central-framework
2. The 2006 IPCC Guidelines for National GHG Inventories are available at https://www.ipcc-nggip.iges.or.jp/public/2006gl/ and its 2019 refinement at https://www.ipcc-nggip.iges.or.jp/public/2019rf/