Every farm in the country carries the same emissions figure per cow — a globally-derived default set in Geneva, unadjusted for UK breeds, diets, or management. The data to change that has been flowing through routine milk recordings for decades. Nobody has used it for this purpose.
The problem
The best-managed farm and the worst carry identical methane figures. There is no measurement infrastructure that can tell them apart.
UK dairy farms report enteric methane using IPCC Tier 1 default emission factors — global averages derived from international livestock datasets with no adjustment for UK conditions. This is not a gap that better farm management can close. It is a data infrastructure problem. Without cow-level or herd-level measurement, there is no basis for differentiation, no route to carbon markets, and no data layer for a breeding programme for reduced methane.
Tier 1Current UK reporting basis
0UK herd-level methane datasets
0.68–0.75MIR prediction R² — McParland 2023
10yr+Published MIR research base
International context
The methodology is proven. The deployment gap is real.
Mid-infrared spectral analysis of routine milk samples as a methane proxy has been validated for over a decade. Belgium deployed it at population scale in 2016. France uses it for on-farm counselling. Canada has built national genomic evaluations from it. No country has crossed into MRV or carbon credit origination. That step remains unbuilt everywhere.
Belgium
~2M records operational since 2016. Research and genetic counselling.
Deployed
France
Routine deployment via milk recording companies. On-farm advisory.
Deployed
Canada
National genomic evaluation for methane efficiency — Lactanet operational.
Deployed
Ireland
Research phase. McParland 2023 dataset. ICBF evaluation in development.
Research
UK
Not deployed. NMR spectral stream exists and is unused for this purpose.
None
Milk Seam
Building the UK deployment layer. MRV and carbon origination as terminal objective.
Target
The programme
A staged build — each step is the prerequisite for the next.
Stages 1–3 are achievable with existing infrastructure and published methodology. Stage 4 depends on corporate sustainability reporting standards currently evolving in the right direction. Stage 5 — carbon origination — cannot be approached without Stages 1–4 in place, and faces structural blockers that are outside this programme's control to resolve.
01
Population signal
→ first UK herd-level methane distribution
Apply published MIR prediction models to NMR/AB Agri spectral data from UKDCN pilot farms. No new hardware. No farm burden. The first population-level view of UK dairy methane variance the industry has ever had.
↓ signal creates the case for validation investment
02
UK validation
→ UK-specific model accuracy confirmed
Published models were built on European and Irish data. Transferability to UK breeds and diets requires paired ground truth measurement. UK data exists in pockets — AFBI Hillsborough is the primary institutional source. Volume and quality to be assessed.
↓ validated model unlocks breeding and management applications
03
Breeding signal
→ UK sire-line methane phenotypes at commercial scale
With enough cow-level predictions across enough herds, genetic patterns emerge. Canada's Lactanet has taken this step. ICBF in Ireland is building toward it. The UK has no equivalent — yet. This stage has clear commercial value independently of carbon markets.
↓ breeding signal generates the phenotypic database MRV requires
04
MRV foundation
→ farm-level emissions signal for sustainability reporting
A validated, nationally-deployed model constitutes the first credible MRV data layer UK dairy has ever had. Processors under Scope 3 reporting obligations can replace Tier 1 defaults with farm-level predicted figures, disclosed with appropriate uncertainty. CSRD direction is supportive but not yet settled for agricultural methane.
↓ MRV foundation is the prerequisite the carbon market cannot bypass
05
Carbon origination
→ verifiable credits from dairy methane variance
The terminal objective. Farms demonstrably below a corrected baseline generate a quantified, auditable emissions advantage monetisable in voluntary carbon markets. This step has not been taken anywhere in the world.
Why Stage 5 is not yet reachable
Three structural blockers remain unresolved: voluntary carbon market acceptance of predicted data; baseline methodology for farms that have always emitted less than Tier 1 assumes; and additionality standards for methane variance driven by genetics rather than intervention. These require VCM governance evolution — the direction is right, the timeline is years. They cannot be bypassed. They can only be approached once Stages 1–4 have built the data infrastructure beneath them.
Business case
Near-term revenue does not depend on carbon markets.
Three revenue models are viable in the near-to-medium term. Each depends only on Stages 1–3. Carbon origination is a conditional long-term position, not a basis for near-term planning.
Near term
Data licensing — breeding organisations
MIR-predicted methane phenotypes as input to national genetic evaluations. Canada's Lactanet model is the proof of concept. BCMS and genomic service providers in the UK are the natural customers.
Near term
Supply chain sustainability data licensing
Processors and retailers under CSRD Scope 3 obligations need farm-level methane data. A predicted figure with disclosed uncertainty and a published methodology is more defensible than a global average. Arla, Müller and major retailers are the natural buyers.
Medium term
Research and innovation funding
Shadow analysis and UK validation work are fundable under Defra and Innovate UK frameworks. UKDCN is itself Defra-funded. A structured research collaboration with AFBI or Teagasc is a realistic near-term development pathway.
Conditional
Carbon credit origination
Dependent on resolution of Stage 5 blockers. Not a basis for near-term financial planning. Remains the terminal value proposition — but will not be reached without the data infrastructure Stages 1–4 provide.
Get in touch
Milk Seam is in its pre-pilot phase. We are building relationships with UKDCN, the academic community working on MIR methane prediction, and NMR/AB Agri on data access. If you work in this space — as a researcher, policy maker, or data holder — we would like to hear from you. We are not pitching. We are assembling the right people around a problem that has been sitting unsolved for a decade.
A full programme assessment — including the three structural blockers to carbon origination in detail, the required data infrastructure, and the specific stakeholder map — is available on request.