Measurement and verification
Every number on this site has a method behind it, and the method is published alongside the number. Anything we cannot measure honestly, we do not claim.
Why measurement matters
Buyers of land based carbon have been burned repeatedly, and the sector's response has mostly been better marketing. We think the correct response is to make the method inspectable, because a buyer who can check the working does not have to trust the seller.
There is a commercial argument for this as well as an ethical one. When most claims in a market cannot be verified, a claim that can be verified is worth more. The work described here is not overhead on the business. It is a substantial part of what we are selling.
What we measure and how
- Area and boundaryAt establishment
- Block boundary recorded as coordinates and held as a geospatial file, so that every later measurement refers to the same defined area rather than an approximate one.
- Planted count and speciesAt establishment
- Recorded per block by species with planting dates. This is the denominator for every survival figure that follows, so it is recorded before planting rather than reconstructed after.
- SurvivalScheduled revisits
- Counted on fixed sample plots located by coordinates, revisited on a set schedule including at least one count after a full dry season. Results are reported with the sample size and the resulting margin of error.
- GrowthScheduled revisits
- Height and stem diameter on the same sample plots, which is what allows biomass and therefore carbon to be estimated rather than assumed.
- Canopy changeContinuous
- Vegetation index time series from satellite imagery over the recorded boundary, giving an independent check on the field counts and catching decline between visits.
- Biochar carbonPer batch
- Feedstock mass in, char mass out, and the process conditions of the batch, since the stability of the carbon depends on how it was produced.
- SoilBaseline and periodic
- Organic carbon and basic soil chemistry sampled at establishment to create a baseline, then resampled periodically, because a claim of improved soil is meaningless without the starting point.
Sample plots are fixed and revisited, never reselected, because choosing new plots each time allows healthy areas to be favoured without anyone intending to cheat. And a survival count taken before a site has been through a dry season tells you very little, so the count that matters is the one after it.
What we publish, and what we will not do
Per project, on the project page:
- Boundary, area, species mix and planting dates
- Planted count, and survival at every scheduled measurement with its date
- The measurement method and sample size used for that count
- Events that affected the block, including fire, grazing loss, drought and replanting
Three things we commit to not doing, because each is how credibility is usually lost:
- No quiet removal
- A project page that reports poor results stays online. Pages are not deleted because the numbers became inconvenient.
- No single point estimates
- Sampled figures are published with their uncertainty. A survival rate quoted to a decimal place from a small sample is a false precision, and we treat it as one.
- No self verification presented as verification
- Our own measurement is described as our own measurement. Where an independent party has checked something, we name the party and what they checked. We do not blur the two.
Measurement
- What does MRV mean?
- Measurement, reporting and verification. Three separate steps.
- Measure using a method stated in advance.
- Report it in a form others can inspect.
- Have it checked by someone independent of us.
- How is tree survival actually counted?
- On fixed sample plots, not by walking the site and estimating.
- Plots are marked by coordinates and revisited, never reselected.
- Each plot has a planted count recorded before planting.
- The count that matters is after a full dry season.
- Results are published with sample size and margin of error.
- Why do carbon credits have a credibility problem?
- Because many issued credits turned out to be worth less than claimed. Three usual causes.
- Protecting land that was never actually going to be cleared.
- Assuming survival rates nobody ever verified.
- Choosing a baseline that maximises the credit.