Measuring Carbon Footprints of Agri‑Food Products

11. Updating as new scientific insights and techniques become available

Abstract

Scientific insights as well as technologies and practices will continue to evolve. The eighth building block for carbon footprints is therefore that all other building blocks should be able to adapt over time, incorporating new scientific insights and techniques. This chapter discusses the tension between the need for flexibility and the need for stability, and argues that existing initiatives should adopt an explicit process for reviews and updates.

Science and technology are continuously evolving. The building blocks for measuring carbon footprints will therefore need frequent updates as well. But frequent updates could create additional costs and uncertainty. A deliberate approach should strike a balance between the need for change and the need for stability.  

Some examples can illustrate the need for updates.

• Reporting standards. Current reporting standards are a compromise between what is desirable and what is feasible. For example, the IDF guidance for carbon footprints in the dairy sector (Chapter 4, Section 4.1.1) notes that soil carbon sequestration is important but that there is a lack of consensus on methodology. The current version of the guidance therefore recommends reporting sequestration separately rather than making it part of the carbon footprint calculation itself. It makes sense that such guidance would be revised once methods are sufficiently mature. Reporting standards could also be relaxed if it turns out that some requirements are not feasible or not as important as previously thought.

• Science-based methods. Scientific research is continuously refining the most sophisticated (Tier 3) models for quantifying GHG emissions. But even the less sophisticated (Tiers 1 and 2) methods are subject to change. As noted earlier, research using atmospheric inversion techniques suggests that methane emissions from livestock might in some cases be greater than currently thought. If confirmed, such insights should be reflected in emission factors. As another example, if new mitigation options are available to farmers, science-based methods should be developed to quantify their reduction potential.

• Farm-level tools. A smoother connection between farm management software, smart farming equipment, government databases, and the like would make it possible to use more detailed data for farm-level carbon footprint calculations while reducing the reporting burden for farmers. In turn, this means more sophisticated calculation methods could be used. The availability of new mitigation options should also be reflected in farm-level tools, once it is clear how reductions should be quantified.

• Secondary databases. As firms and farmers start to reduce their emissions, LCA databases should be updated to reflect this new reality. For example, lower emissions in fertiliser production should lead to lower life-cycle emissions for wheat, and hence also for bread. Moreover, as science-based methods change, LCA databases should re-calculate emissions, too. As noted earlier there is also a need for further “regionalisation” of LCA databases, using new data and models to create more fine-grained emission factors. More precise estimates could over time replace extrapolations.

Changes in one building block may require changes in several others. For example, an update to the reporting requirements of the GHG Protocol or new science-based methods would trigger changes in farm-level tools and LCA databases and potentially in digital tools used to share data along the supply chain.

11.1. The tension between change and stability

While these examples illustrate the need for updates, there is a tension between change and stability.

On the one hand, if standards, methods, and numbers are not sufficiently updated all three of the levers to reduce emissions in food systems would be weakened.

• Shifting across product categories: If product categories manage to reduce their average emissions this should be reflected in the numbers. Otherwise consumers and supply chain actors would be relying on the wrong signals, and sectors would not be rewarded for reducing their emissions.

• Shifting between suppliers: If suppliers manage to reduce their emissions this should similarly be reflected in the numbers. Otherwise customers would not have the right information to allow them to switch to lower-emission suppliers, and suppliers would not be rewarded for reducing their emissions.

• Reducing emissions through mitigation options: If new mitigation options are not included in calculation methods for carbon footprints, producers would not be rewarded for lowering their emissions. This weakens the incentive to adopt new mitigation techniques – which in turn would weaken the incentive for developing them in the first place.

A lack of regular updates would thus weaken the signals needed to help consumers, producers, and other actors work towards lower emissions in food systems. In addition, a lack of regular updates would mean that any inconsistencies would take a long time to resolve. Discussions over revisions of the building blocks might also become more tense as stakeholders feel that it is “now or never” to introduce or resist a change.

Such a lack of flexibility could be the unintended by-product of rules to create greater reliability. For example, to avoid greenwashing or fragmentation, governments or standard setters might decide to prescribe certain methods and exclude others. While this creates more clarity and comparability today, there is a risk that doing so would effectively halt innovation unless a process exists to revise these decisions.

But there are good reasons to keep changes to a minimum. Frequent changes (especially ones which “cascade” through the different building blocks) create costs, and make it harder to interpret numbers. For example, if farm level tools changed their calculation method every year, it would be difficult for farmers and other actors to understand whether fluctuations in carbon footprint estimates are due to their own actions or to changing methods. The same applies to changes in LCA databases or reporting standards. Ideally, historical data would be recalculated using the new methods, but this is not always feasible and does not fully remove risks of misinterpretation or confusion. Comparisons between products, firms, and countries will be complicated if it is unclear whether numbers were all derived using the same standards and methods. The carbon footprint of fertiliser is an input in the calculation of the carbon footprint of wheat, bread, and so on: in this case there may be confusion over whether all relevant calculations were updated. In short, frequent updates could create uncertainty and a feeling that carbon footprint estimates are arbitrary. These problems would be compounded if there is no clear indication of which version of a standard, method, or database was used, and if it is unclear for how long the current version of a standard, tool, or database will remain valid before another update occurs.

11.2. A first assessment

At the moment, there is no deliberate approach to updating the various building blocks. In fact, many initiatives do not have a pre-defined process or timeline for updates. One notable exception is provided by the ISO standards: all ISO standards must undergo a review at the latest every five years. As the overview of LCA databases showed, LCA databases are also commonly updated, although databases differ in how often this happens. But most initiatives reviewed here are updated infrequently. The IPCC guidance, which is a benchmark for science-based methods, was last updated in 2019; its previous version dated from 2006. The USDA guidance on methods was updated in 2024, with its previous version dating from 2014. The assessment on farm level tools in the United Kingdom (discussed in Chapter 6) concluded that many tools were not aligned on the most recent reporting standards and guidance. Strikingly, the GHG Protocol standards which underpin much emissions reporting do not carry version numbers, and for some of the older standards (e.g. the Agriculture Guidance) even the publication date is not provided in the standards document.

To reconcile the need for change and the need for stability, a deliberate approach to updating the building blocks is needed. Important steps can be taken by initiatives in each building block separately, whereas others require coordination.

Reporting standards, overviews of science-based methods, farm-level tools, and secondary databases could all adopt an explicit policy about how changes will be made. This policy could stipulate a regular timeline for review, and could define how the scientific community, stakeholders, and other initiatives will be involved. Clear version numbers would allow users to specify which version was used, while change logs (i.e. overviews of changes made between two versions) would help users understand how results may differ because of the updates. Third-party verification of carbon footprints could then also mention the version numbers of standards and tools used.

It would be helpful if documents which take stock of science-based methods were updated more frequently. One possibility is to update chapters or sections of chapters separately rather than updating an entire document (as is currently done). Another possibility is to separate discussions of the “state of the art” (which could be updated more frequently) and “recommended” methods (which could be based on feasibility and the need for stability and could hence be updated less frequently). There is a particular need for a process to decide when there is sufficient evidence for a new mitigation technique (e.g. a new practice or new technological solution) to be included in models, and by how much it reduces emissions.

Farm-level tools could adopt clear version numbers and change logs. The tools could clearly state which standards (and which version) they are consistent with, and which version of the relevant scientific guidance they follow. This information could also be part of the output of the tool. Farm-level tools could consider adopting a standardised data format such as the HESTIA format (Chapter 8). Such a standardised data format would make it easier for farmers to store historical activity data, so that historical baselines can be easily recalculated when methods are updated.

These practices could easily be adopted by individual initiatives, but some coordination may be required to align the different processes. For example, it might be helpful for the various initiatives (standards, tools, database providers) to agree on a regular update cycle (e.g. every five years) and on a clear sequence for updating the various building blocks within that cycle (e.g. reporting standards reviewed and updated in year 1, farm-level tools and LCA databases updated in year 2).

It could be useful to have a forum where various initiatives can gather to discuss possible updates. Involving stakeholders in these discussions is important, as these can then communicate to their members what is changing and why (e.g. farm organisations could inform farmers about why estimated carbon footprint numbers will be revised).

One advantage of being explicit about future updates is that it sends a clear message to all stakeholders that the building blocks for carbon footprints will keep evolving but will do so on a predictable timeline. New insights, new techniques, or new concerns of stakeholders could be taken on board in future iterations. This makes innovation and adaptation possible. At the same time, a clear timeline would provide clarity to users about when new changes might be expected.

An explicit process for updates would also help acknowledge that at least initially carbon footprints in food systems may come with considerable error margins, but that stakeholders can work together to reduce these errors over time. As part of the updating process, stakeholders could assess the reliability of carbon footprint estimates, identify the main sources of divergence, and work to improve estimates. For example, an analysis such as that done for farm level tools in the United Kingdom (RSK Adas, 2023[1]) could be done regularly and could form the basis for a discussion with tool providers on how to improve accuracy. As another example, a regular analysis could assess data quality in secondary databases and identify the main activities or products where further research is needed. It could also be used to track progress in scaling up carbon footprints, assessing the main barriers, and developing action plans.

Such a “continuous improvement” approach is common in the fields of quality management and environmental management and underlies modern software development techniques.1 The approach could be a useful way to organise collaboration on the building blocks needed for reliable and widespread carbon footprints in food systems.

References

[3] Ries, E. (2011), The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses, Crown Business.

[1] RSK Adas (2023), Harmonisation of Carbon Accounting Tools for Agriculture, UK Department for Environment, Food, and Rural Affairs, https://sciencesearch.defra.gov.uk/ProjectDetails?ProjectId=20967%0A.

[2] Sutherland, J. (2014), Scrum: The Art of Doing Twice the Work in Half the Time, Crown Currency.

Note

← 1.  See, for example, Sutherland (2014[2]) and Ries (2011[3]).