A record that feed was purchased and a conclusion that methane emissions fell do not tell the same story. That does not make purchase records useless. In Chapter 7 of Technology and Solutions for Demonstrating Carbon Reduction, I emphasize levels of evidence for emissions reductions as a way to explain this difference. Today, let’s look at the relationship between data and claims in carbon data verification.
A distinction to make first as you read the book
If records are incomplete, there is no need to conclude that no real change occurred. But what external claims can be made with the data currently available is a separate question. It may help to distinguish observations, causal effects, and generalization to other farms. The same data may be sufficient for one purpose and insufficient for another. Evidence needed to confirm participation differs from evidence needed to quantify emissions reductions. Deciding the purpose first also makes it easier to determine which records and reviews merit the cost.
MRV is a system that connects Measurement, Reporting, and Verification. It attaches calculation methods and conditions of application to results estimated from observations and activity records, then checks the claim against established criteria. Checks used for internal management have different purposes and requirements from verification required for external submission. First decide who will use the results and for what decision; then you can design the necessary evidence and verification process.
The basic principle is to make claims only within the scope supported by the evidence. Activity records show whether someone took part; suitable observations explain changes under particular conditions. A study design that addresses confounding is needed to establish a technology’s causal effect, and further evidence covering the relevant scope is needed to generalize findings to other farms. Independent review tests the evidence, but it cannot supply a missing control group or make a sample representative.
Measurement produces data, reporting documents the calculations and interpretation, and verification checks a claim against established criteria. These functions and responsibilities should be distinguished, but it is not necessary for three different organizations to carry them out in every MRV system. The same organization may handle measurement and reporting, while the impartiality and competence of verification, and conflict-of-interest controls, should meet the level required by the applicable standard and scheme.
Before signing a contract, agree on the claim, period, boundaries, and applicable criteria; the assurance level and materiality; access to source data; conflict-of-interest management; and how the deliverables may be used. Where needed, also check the organization’s accreditation or registration scope. Read a verification opinion within that scope, and do not expand it into a blanket certification of the entire company, product performance, or future emissions reductions.
Distinguish verification of claims about emissions or reductions that have already occurred from validation of assumptions and methods supporting future performance. Validation of a business plan does not mean its planned reductions will be achieved or that credits have been issued. Post-project data and the relevant scheme’s subsequent procedures are still needed.
What to look for in the field
Suppose a farm submits receipts and feeding logs. These can support an assessment of purchases and use. If observation data are added, we may also see changes under particular conditions. But further evidence from a comparison design and calculation basis is needed before claiming how much emissions fell because of the feed.
Good data combine quality suited to the purpose with traceability. Even if the same calculation can be reproduced, a shared bias in the inputs can still make the result wrong. Verifiability does not replace accuracy; it allows others to assess accuracy and the scope of application.
Quantifying reductions requires both measurement traceability and traceability of the calculations. Link measurements to a documented calibration chain and uncertainty, and trace activity data, emission factors, and models through their sources, versions, and conditions of application to the final result. Not every reduction figure comes from direct measurement alone. Along with checking whether calculations can be reproduced, reviewers should confirm that the data and methods used are appropriate for the claim.
Here is the key insight: evidence has levels, and each level sets limits on what you can say. Evidence at a lower level is not worthless. Activity records prove participation, which is sufficient for many purposes. The problem is using lower-level evidence to make a higher-level claim.
When you receive a document making an emissions reduction claim—such as research findings, a corporate disclosure, policy materials, or a proposal—check what level of evidence supports the claim and whether its wording goes beyond that level.
Questions to ask in an investment review
When reviewing data, ask what statement the data support rather than how much data were collected. Check whether records establishing participation have been presented as verified emissions reductions. The evidence distinctions proposed in the book are not legal grades; they are practical tools for reducing this kind of confusion.
Non-dispersive infrared (NDIR) uses methane’s absorption of infrared light. Unlike catalytic combustion, it does not depend on oxygen or face catalyst poisoning, but it is not free from the effects of optical contamination, temperature, pressure, or moisture. Do not assume that the principle alone makes a sensor suitable for long-term monitoring on a farm. For each product, check whether its detection limit, repeatability, and drift are adequate relative to the expected concentration and the difference between indoor and outdoor concentrations.
Evaluate AI materials according to their intended use. To predict future operation, consider preserving time order; to deploy at other farms, consider separating data by farm. Also separate the data used to select the model from the data reserved for final performance checks. Compare against a simple baseline model, and examine not only average error but also bias by farm and season and conditions under which the model fails. This is necessary to explain the model’s real scope of use.
A question to leave with before closing the book
I hope you will describe precisely what the data you have now can support saying. If you see where evidence is insufficient, the purpose of the next records and tests will also become clearer. The value of carbon data becomes more concrete when you know the scope of the claims you can make with it.
About the author
Bongjun Lee graduated from the Graduate School of Management of Technology (MOT) at Sogang University. He spent seven years at Infitron Co., Ltd. and 20 years at Honeywell, working for a total of 27 years in gas measurement in service engineering, technical sales, and channel management. He is now the founder of AI Safety Korea Co., Ltd., where he is researching livestock methane monitoring and measurement, reporting, and verification (MRV) systems using sensors, AI, and data analytics.