Why the methane tonne you buy should be a measurement, not an estimate.
Most carbon credits are, at their core, an estimate wearing the costume of a fact. A project developer takes an activity — a hectare of forest, a herd of cattle, a leaking well — applies an emission factor drawn from a published table, multiplies by an assumed baseline, and produces a number. The number is precise to several decimal places. It is also, very often, wrong. Not fraudulently wrong. Wrong the way any figure is wrong when it is derived from a coefficient rather than an observation.
For methane, that gap between estimate and observation is not a rounding error. It is the whole game. Methane intensity varies enormously from well to well and basin to basin — a function of geology, equipment age, pressure, and operating history that no generic factor can capture. Factor-based approaches, the kind that sit under most methane-related credits, have been documented to underestimate measured emissions by 1.6 to 3 times depending on the basin and the operation. A credit priced off a broad average is, definitionally, mispriced. Sometimes it flatters the seller. Sometimes it flatters the atmosphere. It rarely tells the buyer the truth.
The LMCX Carbon Avoidance Token begins from the opposite premise: if you cannot measure it, you should not tokenize it. The methane abatement behind each CAT is quantified to OGMP 2.0 Level 5 — the highest tier of the United Nations’ oil-and-gas methane framework. Level 5 is demanding by design. It requires simultaneous source-level direct measurement and site-level top-down atmospheric reconciliation, with the two independent data streams checked against each other on a continuous basis. In the LMCX architecture, that means Kuva optical-gas-imaging at the source and DrillSage satellite spectrometry overhead, reconciled rather than assumed. The credit is issued against what was observed, not against what a table predicted.
This is a higher bar than it sounds. As of 2024, only 31 companies in the OGMP membership had achieved Level 5 status across more than half of their operated production — worldwide. It is not a checkbox; it is a measurement discipline that most of the industry has not reached. When we say the reference asset behind a CAT meets Level 5 at the individual-asset level, we are describing a standard that the scarcity of the market itself confirms. The supply that follows from measuring this rigorously is naturally scarce, because the measurement is naturally hard.
It is worth being concrete about what “reconciled” means, because it is the part that most distinguishes measurement from estimation. A single instrument, however good, can be argued with. Optical-gas-imaging at the wellhead sees the source directly but can miss intermittent or diffuse emissions. A satellite pass sees the whole site from above but at coarser resolution. The point of running both is that neither is asked to be the sole witness. When the bottom-up source measurement and the top-down atmospheric observation are required to agree — continuously, not once — a whole class of errors that survives in single-method programs simply cannot. An emission the ground instrument understated shows up in the atmospheric column. A plume the satellite attributed to the wrong source is corrected by the ground data. Reconciliation is not redundancy for its own sake; it is the mechanism by which a measurement earns the right to be treated as a fact rather than a reading.
That discipline has a commercial consequence worth naming. Because Level 5 measurement is expensive, slow, and operationally demanding, the volume of methane abatement that can be quantified to this standard is structurally limited — not by a marketing decision to appear scarce, but by the sheer difficulty of the measurement itself. Scarcity that comes from a supply cap is a promise. Scarcity that comes from a standard almost no one can meet is a moat. The CAT’s supply is constrained by the second kind, and we would rather have a smaller quantity of tonnes that were genuinely observed than a larger quantity that were merely assumed.
We want to be precise about what this claim is and is not. Measurement is the foundation of a credible credit — it is not the entirety of credit quality. A perfectly measured tonne still has to clear questions of additionality, baseline, and permanence before it deserves an institution’s balance sheet, and we address those elsewhere, on their own terms. What measurement does is remove the first and most common failure mode: the quiet substitution of an assumption for a fact. You cannot build integrity on top of a number you never actually observed. Everything downstream — the accounting, the verification, the legal wrapper — inherits the error in the measurement if the measurement was a guess.
There is a reason this matters more to institutions than to anyone else. A corporate buyer retiring credits for a voluntary claim can absorb some measurement uncertainty as reputational risk. A bank underwriting an instrument, a treasury holding an asset, or an allocator reporting financed emissions cannot. Their auditors will ask how the underlying tonne was quantified, and “we applied the standard factor” is an answer that invites a markdown. “It was directly measured to OGMP Level 5 and reconciled against satellite data” is an answer that survives the room. We built the CAT for that room.
None of this makes measurement glamorous. It is instrumentation, calibration, reconciliation, and documentation — the unshowy work that happens before anything is issued. But it is the difference between an environmental commodity and an environmental estimate. The market has spent a decade discovering, credit class by credit class, what happens when that difference is ignored: the quiet re-rating, the retracted claim, the buyer who thought they had bought a tonne and had actually bought a coefficient.
Measured, not modeled, is not a slogan. It is a sequence. Measure first, to a standard the market itself can interrogate. Then, and only then, build the credit. A methane tonne is either something you observed or something you assumed. We think institutions have earned the right to buy the first kind.