Measuring Carbon Footprints of Agri‑Food Products

5. Science-based methods

Abstract

Science-based methods are the second building block for carbon footprints. This chapter introduces the IPCC guidelines for quantifying emissions and related national guidance for quantification at the farm level. The chapter also discusses the challenge of continuous improvement of these methods.

5.1. Overview

Reporting standards and guidelines answer the question of what needs to be reported; science-based methods are needed to answer the question of how emissions can be quantified.

Ideally, emissions would be measured directly, but this is rarely practical in food systems (at least with current technology). A wide range of methods exist to estimate emissions. An important question is therefore how to choose between the available methods.

Governments face the same question in drawing up their National Greenhouse Gas Inventory Reports under the UN Framework Convention on Climate Change. Important guidance has been developed in this context by the Intergovernmental Panel on Climate Change (IPCC) (IPCC, 2006[1]). Its 2006 Guidelines consist of five volumes, of which one focuses on Agriculture, Forestry, and Other Land Use. The Guidelines were last refined in 2019 (IPCC, 2019[2]). The Guidelines make a useful distinction between three “tiers” of calculation methods, from least to most detailed.

To illustrate these different tiers, consider the question of how to quantify methane emissions from enteric fermentation in North American dairy farming [adapted from Rotz (2018[3])].

• The simplest approach (known as a “Tier 1” approach) is to multiply the number of dairy cows by a fixed emission factor of 138 kg CH₄ per year, which is the emission factor for North American dairy cattle provided by IPCC (2019[2]).

• A more refined approach (known as “Tier 2”) lets the emission factor vary based on gross energy intake and a methane conversion factor. Countries can develop their own methane conversion factor, but if these are not available, IPCC (2019[2]) provides default values which depend on milk production levels and on feed quantity and quality.

• An even more refined set of approaches are so-called “Tier 3” methods. These consist of a wide range of methods which can provide more accurate estimates, but require more data as input. For example, these could include more detailed statistical models for emission factors, or process-based models. Some of these approaches might be appropriate for research purposes but too cumbersome for other applications (Rotz, 2018[3]).

Specifically for emissions in Agriculture, Forestry and Other Land Use (AFOLU), the IPCC Guidance notes that Tier 3 methods may include process-based models and inventory measurement systems driven by high-resolution activity data and disaggregated at sub-national level, including, for example, comprehensive field sampling. IPCC also provides further guidance on using Tier 3 methods, covering topics such as sampling methods for measurement-based approaches and model selection, calibration, and evaluation for model-based approaches (IPCC, 2019[2]). Box 5.1 discusses such methods in the context of measuring soil organic carbon.

Box 5.1. Soil organic carbon

Three types of Tier 3 approaches exist to measuring and monitoring soil organic carbon. These are direct measurements, remote sensing, and simulation models (Paul et al., 2023[4]). These methods vary in terms of cost and reliability.

• Direct measurement (i.e. soil sampling) is the most reliable approach, but is costly and time-consuming when applied to large areas. One reason is that soil characteristics, including the content of organic carbon, can vary considerably even across a single field. Soils also contain a mix of organic and inorganic components, and SOC levels fluctuate with soil depth. All this makes it harder to collect representative samples, and harder to extrapolate accurately from measurements in specific sites to estimates for broader areas. Moreover, to quantify changes over time requires re-sampling at intervals of at least 3-5 years, as it takes several years for soil management measures to create observable impacts (Paul et al., 2023[4]).

• Remote sensing offers a potentially cost-effective means of monitoring SOC across large areas in the topsoil, but requires specific conditions such as bare soil, low water content, etc. To date, no studies have successfully detected SOC changes at the field scale using remote sensing methods (Paul et al., 2023[4]).

• Simulation models are the most economical and readily available option, and are already widely used for certification schemes (Oldfield et al., 2022[5]). However, currently used simulation models have important shortcomings. A detailed review by Garsia et al. (2023[6]) identified 221 soil organic carbon simulation models. Of these, less than one-third (64) had been validated in line with IPCC guidelines. Of those that had been validated, few were validated for multiple countries and land uses, with large gaps for sub-Saharan Africa and the Middle East.

Given the importance of soil organic carbon sequestration as a potential mitigation option, greater investment is needed in developing and validating soil organic carbon simulation models across a wide range of contexts.

The development of more precise methods to quantify GHG emissions and removals is an active area of research, and further guidance may be useful to help practitioners understand the strengths and weaknesses of different methods. In the United States, for example, the US Department of Agriculture (USDA) in 2014 published a review of relevant methods, updated in 2024 (Hanson, Itle and Edquist, 2024[7]). It informs USDA’s own efforts in estimating GHG fluxes and is used as a basis to update USDA’s estimation tools (COMET-Planner and COMET-Farm, discussed in the next chapter); the review is also a valuable source of information for farmers and other stakeholders.

In terms of the reporting categories of the GHG Protocol’s draft Land Sector and Removals Guidance (Table 4.2), the IPCC Guidance for AFOLU covers land management emissions (both CO₂ and non-CO₂), land management CO₂ removals, and emissions and removals from land use change. So-called “non-land emissions” in agriculture (e.g. combustion of fuels for tractors or heating) and emissions from other segments of the supply chain (e.g. fertiliser production, food processing, and transport) are covered in the IPCC Guidance for other sectors.

5.2. A first assessment

Science-based methods for quantifying emissions in food systems are generally well developed. The IPCC Guidance’s Tier 1 and 2 methods provide an internationally accepted baseline, and some countries have developed further guidance on the most appropriate Tier 3 methods. Such country-specific guidance is particularly useful given the important role of local conditions such as soils, climate, or farming approaches in shaping emissions. It is beyond the scope of this report to review the various methods in detail; for example, the USDA review of these methods for the US context alone runs to some 600 pages (Hanson, Itle and Edquist, 2024[7]). But in general, it appears that there exists a foundation of science-based methods covering most of the relevant categories of emissions and removals described in carbon footprint reporting standards. The default methods provided by IPCC (Tier 1 and Tier 2) are designed to be unbiased, i.e. neither systematically above or under the true value; but as these are relatively coarse approximations, there may be considerable non-systematic error in a given application. Tier 3 methods are designed to be more adapted to local circumstances but require more effort (e.g. more precise data). For developing countries (notably sub-Saharan Africa), however, fewer Tier 3 methods have been developed so far. Additional research here would be welcome. As noted above, simulation models for soil organic carbon may similarly be an area where additional research (and especially validation) would be useful.

Scientific insights will continue to evolve, however, and calculation methods can therefore be expected to evolve over time as well. For example, research based on atmospheric measurements suggests that existing calculation methods (which include the Tier 1, 2 and 3 approaches) may in several cases significantly understate true emissions (Miller et al., 2013[8]) (Deng et al., 2022[9]), including methane emissions from animal agriculture (Hayek and Miller, 2021[10]). Improved satellite measurement could similarly provide new insights (Bourke, 2024[11]). These methods can be seen as ‘top-down’ approaches, starting from measured concentrations of GHGs in the atmosphere and tracing this back to emissions sources. By contrast, Tier 1, 2, and 3 methods are typically ‘bottom-up’, as they are often based on measuring or modelling individual farms or farm animals. New research findings may lead to updates over time in the Tier 1, 2 and 3 methods to better match ‘top-down’ estimates.

One practical challenge is in translating these evolving insights into updates of authoritative guidance documents. For example, IPCC Guidance was originally published in 1996, updated in 2006, and subsequently refined in 2019, which means that new methods may take several years before being referenced in updated IPCC Guidance. Similarly, it may take time for improved calculation methods to find their way into practical tools for calculating emissions.

Finally, it is worth noting the important connection here with National Greenhouse Gas Inventory Reports required under the UNFCCC. The national reporting level differs from the product-level view (the focus of this report) in several ways. First, when reporting at a national scale, questions of allocation across products are generally not relevant: it is sufficient to know emissions of dairy cows without having to worry about allocating these emissions across milk, meat, and manure. Second, the product-level view is based on a life cycle perspective, which means that the relevant activities in scope are not constrained by national borders. Thus, emissions from fertiliser production are part of the product-level view for crops even if these emissions occurred in another country. Third, some methods appropriate for national inventories may be either too coarse or too complex for a product-level view. They may be too coarse when some producer-specific differences are not taken into account for national inventories because they would “cancel out” at the national level or when studying these differences would not be a good use of resources (e.g. because the activity is minor for the country as a whole). In other cases, they may be too complex when national inventories rely on sophisticated Tier 3 methods which are not easily reduced to an easy-to-use farm level calculation tool.

Despite these important differences, the science-based methods as originally developed for national inventories form the backbone for measuring emissions at firm-level and product-level, and the methods used in national inventories can be seen as a “default” choice for measurement at these other levels, unless there are good reasons to use alternative methods.

References

[11] Bourke, I. (2024), Methane: the tricky hunt for hidden emissions, BBC, https://www.bbc.com/future/article/20240306-agricultural-methane-is-a-climate-action-blind-spot.

[9] Deng, Z. et al. (2022), “Comparing national greenhouse gas budgets reported in UNFCCC inventories against atmospheric inversions”, Earth System Science Data, Vol. 14/4, pp. 1639-1675, https://doi.org/10.5194/essd-14-1639-2022.

[6] Garsia, A. et al. (2023), “The challenge of selecting an appropriate soil organic carbon simulation model: A comprehensive global review and validation assessment”, Global Change Biology, Vol. 29/20, pp. 5760-5774, https://doi.org/10.1111/gcb.16896.

[7] Hanson, W., C. Itle and K. Edquist (2024), Quantifying greenhouse gas fluxes in agriculture and forestry: Methods for entity-scale inventory, Technical Bulletin Number 1939, 2nd Edition, US Department of Agriculture, Office of the Chief Economist.

[10] Hayek, M. and S. Miller (2021), “Underestimates of methane from intensively raised animals could undermine goals of sustainable development”, Environmental Research Letters, Vol. 16/6, p. 063006, https://doi.org/10.1088/1748-9326/ac02ef.

[2] IPCC (2019), 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories.

[1] IPCC (2006), 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme, Intergovernmental Panel on Climate Change, https://www.ipcc-nggip.iges.or.jp/public/2006gl/.

[8] Miller, S. et al. (2013), “Anthropogenic emissions of methane in the United States”, Proceedings of the National Academy of Sciences, Vol. 110/50, pp. 20018-20022, https://doi.org/10.1073/pnas.1314392110.

[5] Oldfield, E. et al. (2022), “Crediting agricultural soil carbon sequestration”, Science, Vol. 375/6586, pp. 1222-1225, https://doi.org/10.1126/science.abl7991.

[4] Paul, C. et al. (2023), “Carbon farming: Are soil carbon certificates a suitable tool for climate change mitigation?”, Journal of Environmental Management, Vol. 330, p. 117142, https://doi.org/10.1016/j.jenvman.2022.117142.

[3] Rotz, C. (2018), “Modeling greenhouse gas emissions from dairy farms”, Journal of Dairy Science, Vol. 101/7, pp. 6675-6690, https://doi.org/10.3168/jds.2017-13272.