Hello, I’m Bongjun Lee, author of Technology and Solutions for Demonstrating Carbon Reduction. Today, I’d like to talk with you about digital MRV from the book. Let’s start with a familiar question: “Did it really go down?” Whether you are a farmer adopting carbon reduction technology or an investor looking at a climate tech company, you ultimately want to know the answer.
Shall we picture a scene? After a farm starts feeding low-carbon feed, the methane concentration graph in the barn goes down. That’s encouraging. But can we immediately say, “Methane emissions fell because of the feed”? I’d suggest pausing here to look at the conditions beyond the graph, too. How long were the ventilation fans running that day? Were the number of cattle and feeding times the same?
Did Emissions Fall Just Because Concentration Went Down?
First, let’s distinguish concentration from emissions. Concentration tells us how much methane is mixed into the air. Emissions tell us how much methane was released over a given period. The two figures answer different questions.
Even if the same amount of methane is released, increasing ventilation can mix it into more air and lower its concentration. The indoor air has been diluted. Conversely, even if the concentration is similar, a change in the volume of air moving outside can change the amount of methane emitted. That’s why it is difficult to confirm the feed’s reduction effect from a concentration graph alone.
That does not mean sensor readings are useless. Sensors are important tools for observing changes in the field. The point is that to interpret emissions, we need to connect those readings with information such as airflow, background concentration, and measurement location. This is especially important in naturally ventilated barns, where we first need to consider how to account for wind and air movement.
Digital MRV Keeps an Explanation for Those Numbers
Let’s unpack MRV. It consists of three processes: measurement, reporting, and verification. Data are produced in the field, the method used to calculate them is explained, and the explanation and evidence are checked against defined criteria. In Chapter 7, I discuss this process as a single chain of evidence. The links leading all the way to the final figure matter.
Digital MRV uses sensors, communications, and data processing technologies in this process. The World Bank’s MRV explainer also explains that digital technology can support data collection, processing, and quality management. But the fact that data were collected automatically does not, by itself, confirm the quality of the measurements. We should also consider whether the equipment was installed in an appropriate location and what condition it was in.
Before deciding what to measure, I recommend thinking first about how you intend to use the results. Are you trying to identify an unusual condition on a farm? Compare the effects of feed? Submit reduction volumes to an external party? Different purposes call for different data and verification scopes.
Shall We Also Distinguish What AI Can Help With?
AI comes up often in discussions of climate tech. It can help identify sensor anomalies or estimate values for short gaps in the data. The point to remember is that making good predictions and proving the cause of a change are not the same thing.
Suppose AI predicts changes in methane concentration well. That does not automatically lead to the conclusion that the feed caused those changes. We also need to consider whether the season, ventilation, or livestock conditions changed at the same time. Nor should values filled in by calculation be presented as if they were directly measured. To assess the results, you need to know what was observed and what was estimated.
If you wonder whether AI will work on other farms, it is useful to ask whether its performance has been checked under the conditions on those farms. You can also check the training data, model version, and conditions in which it does not perform well. I emphasize these records in Chapter 8 to explain the scope in which a model can actually be used.
Investors, Look at the Records Behind the Graphs, Too
If you are evaluating a climate tech company, I hope you will consider not only how much data it has, but also what decisions those data can help inform. When a reduction figure appears on screen, trace where it came from. Check which equipment supplied the data, what calibration and calculations were applied, and who changed the data and why, if any changes were made.
This pathway is called data lineage. Other people should be able to review the process as well. If the graph is available but the raw data are not, review may become difficult. That is why you should also check data access rights and handover terms, calibration history, and processing rules.
Even if the same number can be recalculated, the conclusion may still be inaccurate if the input data or comparison conditions are wrong. We need to examine both reproducibility of the calculation and the suitability of the supporting inputs. I think checking whether a company can explain these details is a good place to start a technical review.
Questions to Keep in Mind as You Read the Book
Whose decision are these data meant to support, and what decision is it?
Does the explanation distinguish concentration from emissions?
Are there records to compare livestock, feeding, and ventilation conditions?
Are directly measured values distinguished from values estimated by AI?
Can the raw data and calibration and change histories be reviewed?
Have the verifier’s role and potential conflicts of interest been checked?
Carbon reduction verification is a process of answering these questions one by one. It means looking beyond the fact that sensors have been installed and data have accumulated. It means recording what change was confirmed, how far it can be explained, and what still needs to be checked. Even when there is a verification result, credit issuance or sales revenue must be checked separately against the applicable requirements and contracts.
You do not need to do every calculation yourself after reading the book. Just knowing what to ask when you receive data can make a difference. You can look at a technology’s promise alongside the evidence for it. The next time you see a carbon reduction graph, I hope you will also take a look at how those numbers were produced.
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.