How Earth Signals measures ecological condition.

The idea

Carbon credits price one ecological service. Earth Signals — and the Earth Credit underneath — accounts for all of them: carbon, water, biodiversity, connectivity, soil, and the function of the whole ecosystem. Healthy land does all of those at once. A single number that summarises only one of them quietly misleads.

EC-M-1.1 is the methodology that turns satellite data, soil data, species data, and ground truth into a single Ecological Condition Index — a 0-to-1 number that tells you what the land is, relative to what it would be if healthy.

The six dimensions

Each dimension is scored from 0 (collapsed) to 1 (fully healthy) against a biome-specific baseline. The dimensions are SEEA-EA's ecological-condition typology, operationalised at parcel resolution.

The six scores are combined using a penalised geometric mean (Greco et al. 2019). Translation: one bad dimension lowers the whole score. Excellence in one area can't paper over collapse in another. A forest that scores 0.95 on physical and chemical but 0.20 on compositional is a forest where the vertebrates are gone. The mean treats it that way.

What an ECI looks like in practice

A representative parcel of mixed northern hardwood forest in Vermont under EC-M-1.1, July 2025 to June 2026. ECI dips through the dormant winter months as the compositional and functional dimensions narrow, then climbs back into the growing season. The shaded band is the 95% confidence interval. Past months are immutable — the same chart, queried in 2040, returns the same values.

The threat picture

The most expensive failure of carbon credits has been paying people to protect land that was never going to be lost. Earth Signals computes a separate Threat Multiplier from three independent public data sources:

• Land-use change In the surrounding region. Hansen / GFW canopy time series, annual.
• Development pressure Within a defined neighbourhood radius. Kennedy et al. (2020) Global Human Modification.
• Ecosystem rarity IUCN Red List of Ecosystems and the Nature Needs Half framework (Dinerstein et al. 2017).

Land facing real threat earns full weighting. Land facing no measurable threat earns reduced weighting. The data is public — we read it, we do not assign it.

The cautious math

EC-M-1.1 deliberately reports the cautious end of the measurement's confidence range, not the headline number. Early observations carry wider intervals; later observations, with more ground truth and more validation, narrow. Trust is earned over time, not assumed on day one.

Earth Credit = Verified Acres × Cautious ECI × Threat Multiplier

Earth Signals reports each component — the headline ECI, the confidence interval, the threat multiplier, and the final yield — on every observation. Customers and counterparties can choose the conservatism that suits their use.

Where the measurement stretches — and we say so

Not every parcel has every input the methodology specifies. EC-M-1.1 reports, on every observation, data notes and spec deviations: every place where a substitute was used for a methodology-spec source. Substitutions are bounded — they move the score by less than 10% in either direction, and the engine logs each one explicitly.

The lineage URL on every observation lets any customer drill from the headline number down to the inputs, the substitutions, and the math. There are no opaque boxes.

Standards we build on

• UN SEEA-EA System of Environmental-Economic Accounting — Ecosystem Accounting (2021). The six-dimension condition typology.
• IUCN GET Global Ecosystem Typology and the Red List of Ecosystems.
• GEO BON EBVs Essential Biodiversity Variables.
• Greco et al. (2019) Penalised geometric mean used for the ECI.
• Dinerstein et al. (2017) Ecoregion classification and the Nature Needs Half framework.
• Kennedy et al. (2020) Global Human Modification map.