Validating what the ecosystem does naturally
Three ways to replace a modelled estimate with a measurement. The method is the same in each — measure the atmosphere over the thing you are claiming, and record where every number came from.
Carbon markets built on truth, not estimates
A restoration project's value rests on a number almost nobody measures directly. The default is a model: take a land-use class, apply a regional coefficient, and report the product as though it were an observation.
We measure the flux over the site itself, continuously, and publish the uncertainty band alongside the figure. When the measured value disagrees with the model, that disagreement is the finding — not an error to be reconciled away.
- What we measure
- Net ecosystem exchange, hourly, over the project boundary
- How
- Tower-mounted flux sensors, satellite reanalysis, atmospheric inversion
- What you get
- A verified figure, an uncertainty band, and the raw series
- Who it is for
- Project developers, registries, ratings agencies
Real-world emissions data, enterprise-wide clarity
Scope 3 is usually spend multiplied by an industry factor, which means a company that genuinely improves its sourcing sees no change in its reported number. The incentive points the wrong way.
Measuring at the sourcing region replaces the factor with an observation. In practice some regions come out materially cleaner than the average assumed — and some do not, which is the point of measuring.
- What we measure
- Emissions intensity per sourcing region, per season
- How
- Regional inversion, supplier data, ground truth sampling
- What you get
- A restated Scope 3 line your auditor can follow
- Who it is for
- Food, apparel and materials groups with land-linked supply
Disclosure that survives an audit
CSRD, ISSB and the SEC rule all ask the same awkward question: where did this number come from, and can you show the chain? A modelled estimate answers the first half.
Because every figure is a measurement with a recorded provenance, the audit trail is the product rather than a document assembled afterwards.
- What we measure
- The same series, mapped to the disclosure line items
- How
- Provenance recorded per reading; nothing is restated silently
- What you get
- Report-ready tables, plus the chain behind each cell
- Who it is for
- Sustainability and finance teams filing under CSRD or ISSB
Every engagement runs
the same four steps.
Scope the boundary
What is being claimed, over what area, for what period. Written down before anything is instrumented, because it is the thing an auditor will test.
Instrument
Flux towers where the site allows, satellite and inversion where it does not. Coverage and its gaps are recorded, not smoothed over.
Measure and reconcile
Continuous readings, reconciled against the model that would otherwise have been used. Divergence is reported, with its cause where we can identify it.
Publish with provenance
The figure, the uncertainty band and the chain behind them, in a form that goes into a disclosure without being re-typed.
Start with the number you are least sure about.
Most engagements begin with one contested figure and a boundary drawn on a map.
Let's talk