Skip to content
Shyenova

Remote Sensing

Environmental MRV

Carbon claims are becoming auditable. Most measurement built for reporting will not survive that.

Environmental reporting has spent a decade being narrative and is now becoming quantitative and contestable. Restoration commitments, carbon credits, deforestation-free sourcing claims and land degradation targets are increasingly subject to verification by parties with an interest in finding them overstated.

Satellite observation is the only practical way to measure most of this repeatably over large areas, and it is also the way a claim will be checked by anyone who wants to dispute it. That argues for building the measurement to withstand scrutiny from the start, with stated uncertainty, retained provenance and a method somebody else could reproduce.

How the work runs

Four phases, and what comes out of each.

Durations are indicative and assume reasonable data availability. Where data is poor we say so in the first fortnight rather than modelling on sand.
01

1–2 weeks

Establish what has to be defensible

Reporting obligations differ enormously in how much rigour they demand. A voluntary disclosure, a certification standard and a carbon methodology impose entirely different requirements on measurement, uncertainty and documentation. We start from the standard you are reporting against and work backwards, because designing for the wrong one is expensive in both directions.

Outputs

  • Requirements mapped from the applicable standard or methodology
  • Measurement boundary, baseline date and reporting unit defined
  • Uncertainty and documentation requirements stated up front
02

3–5 weeks

Build the baseline and the reference

Almost every environmental claim is a claim about change against a counterfactual, which means the baseline carries as much weight as the measurement. We establish historical condition from the archive, which is often the only way to reconstruct a baseline for a date that has already passed, and define the reference against which change will be assessed.

Outputs

  • Historical baseline reconstructed from the satellite archive
  • Reference or control areas where the methodology requires them
  • Provenance record for every scene and processing step
03

4–6 weeks

Measure with stated uncertainty

Forest cover, canopy condition and above-ground biomass are estimated from optical and radar observation calibrated against field plots where they exist. Biomass in particular carries substantial uncertainty and we report it as a range rather than a point, a single figure implies a precision the method does not have, and an auditor will say so.

Outputs

  • Cover, condition and biomass estimates with confidence intervals
  • Field plot calibration and validation where available
  • Change quantified against the baseline with uncertainty propagated
04

2–3 weeks

Make it reproducible by somebody else

The test of environmental measurement is whether an independent party could repeat it and land in the same place. We deliver the method, the code, the input scene list and the validation data as a package, and run the reporting cycle once with your team so the process is genuinely theirs.

Outputs

  • Reproducible pipeline with code, inputs and provenance
  • Reporting pack aligned to the applicable standard
  • One supervised reporting cycle run by your team

What good looks like

The ranges we design toward.

These are the improvement ranges these levers are known to produce where the underlying practice has not been examined recently. They are targets to test against your baseline, not promises.

±15–25%

Biomass uncertainty

Honest range for satellite above-ground biomass with limited field calibration. Anyone quoting better without plot data is guessing.

90–95%

Forest cover accuracy

Typical accuracy for forest versus non-forest classification, which is a far easier problem than biomass.

Annual → quarterly

Reporting cadence

Achievable shift once measurement is a pipeline rather than a manual exercise.

Deliverables

What you are left holding.

All of it handed over, documented, and with a session for whoever inherits it.
  1. 01Baseline reconstructed from archive imagery, with provenance for every input
  2. 02Cover, condition and biomass estimates reported with confidence intervals rather than false precision
  3. 03Change quantified against baseline with uncertainty propagated through
  4. 04Reproducible pipeline delivered as code, inputs and documentation
  5. 05Reporting pack mapped to the standard you are actually reporting against

What an engagement looks like

A representative shape.

Illustrative. This is the shape of work in remote sensing, not an account of a specific client. We do not publish named results, and we would be wary of a firm that publishes them without the client in the room.

16 weeks

  1. 01Situation

    An agri-processor was committing to procurement volumes months before it had any reliable view of planted acreage, relying on dealer and field-officer reporting whose historical error nobody had measured.

  2. 02What we did

    Built a Sentinel-1 and Sentinel-2 time series across the sourcing districts, trained crop classification on collected ground truth, and ran the satellite estimate in parallel with the incumbent reporting for a full season.

  3. 03Where it landed

    The satellite estimate was available several weeks earlier and diverged from field reporting by a consistent margin in two districts, which ground checks confirmed. The parallel season is what earned the commercial team's trust; the technology alone would not have.

Where this connects

Rarely a self-contained problem.

These are the adjacencies that come up most often on this kind of work. Naming them is usually cheaper than discovering them halfway through.

Questions

The ones we get asked.

Can you measure carbon directly from a satellite?

No. Satellites measure reflectance and radar backscatter, from which biomass is estimated using models calibrated on field measurements, and carbon is then inferred from biomass with a conversion factor. Every one of those steps adds uncertainty. Anyone offering satellite carbon measurement without discussing field calibration is selling you a model output dressed as an observation.

Why report a range instead of a number?

Because a range is what the method supports, and because a point estimate is far more easily attacked. An auditor who is given ±20% and a documented method has been told the truth about the measurement's limits; one who is given a single figure will start looking for why it is wrong, and will find something. The range is the more defensible deliverable, not the weaker one.

We need this for a specific certification standard. Can you work to it?

Yes, and it should be established in the first fortnight rather than discovered later. Standards differ substantially in permitted methods, required uncertainty, baseline rules and documentation. Building generically and retrofitting to a standard afterwards usually means redoing the baseline, which is the most expensive part.

How does this connect to our supply chain reporting?

Directly, and it is usually where the difficulty is. Deforestation-free or land-use claims about a supply base require knowing where the supply base physically is, which for multi-tier agricultural sourcing frequently nobody does. Mapping that is often the real project, and it overlaps heavily with multi-tier supplier mapping.

Start a conversation

Talk to us about environmental mrv.

Most useful first conversations are thirty minutes and involve no proposal. Describe the problem you are looking at and we will tell you honestly whether it is one we can help with.

Or write directly to hello@shyenova.com