Skip to main content
Methodology · Carbon MRV

Satellite-Based Carbon MRV

Published on April 17, 2026

Most carbon removal claims today rest on self-reported spreadsheets, project-developer attestations, and registries that audit paperwork rather than physical reality. Independent satellite measurement closes that gap - turning sequestration claims into audit-ready, investor-grade credits.

What Carbon MRV Is

MRV stands for Measurement, Reporting, and Verification. It is the quality-control backbone of every serious carbon market - the process that determines whether a tonne of CO2 claimed as removed is actually removed, whether that removal is durable, and whether the credit issued against it represents real climate benefit.

Weak MRV is the single biggest reason the voluntary carbon market has faced credibility issues in recent years. Projects have been over-credited, reversed by wildfire or harvest, or counted twice across registries. Buyers who pay for durable removal end up holding paper that doesn't match what's in the ground.

Why Satellites Change the Math

Satellites provide three things that ground-based MRV cannot deliver at scale:

  • Independence. The sensor does not belong to the project developer, the registry, or the buyer. NASA, ESA, and JAXA operate the missions; commercial providers add higher resolution. Nobody selling credits controls the measurement.
  • Continuity. A forest stand or soil plot is re-measured every few days, not once per five-year reporting cycle. Reversals from fire, harvest, disease, or drought are detected in near real time, not after the fact.
  • Global reach. A single methodology can cover every project on Earth simultaneously - so credits from Kenya, Finland, and Brazil are measured against the same ruler.

Floodlight's MRV Stack

VAULT applies satellite-based MRV across the three carbon removal categories that institutional buyers most commonly underwrite:

Forestry & Afforestation

Above-ground biomass quantified from a harmonized Landsat 4-9, Sentinel-2, and Sentinel-1 SAR stack on Google Earth Engine. A Random Forest model is trained against the ESA CCI Above-Ground Biomass 2018 reference layer using NDVI, NDWI, EVI, and visible/infrared band features. 21-day median composites at 7-day intervals capture phenology and detect canopy disturbance, leakage, and reversal on each overpass. Methodology alignment with Verra VM0047 v1.1, ART TREES, and the Core Carbon Principles is the active compliance track.

Soil Carbon

Soil organic carbon modeled from satellite-derived land-use, tillage, and cover-crop signals, calibrated against in-situ soil samples. Practice change is verified remotely; buyers stop paying for attestations they cannot audit.

Geologic & Subsurface Storage

Surface deformation (InSAR), thermal anomalies, and fugitive methane plumes monitored around CCS injection sites and legacy oil-and-gas infrastructure. Leakage and induced seismicity surface in days, not years.

What the Output Looks Like

Every credit VAULT verifies carries:

  • A quantitative tonnage estimate with an explicit uncertainty range, validated against the ESA CCI biomass reference layer
  • A confidence score tied to satellite coverage, cloud cover, training-sample density, and model R-squared
  • A durability assessment and reversal-risk signal driven by per-overpass canopy and disturbance detection
  • An auditable chain of raw observations, spectral indices, model versions, and calibration data
  • An explicit VM0047 alignment posture: regulatory surplus plus a Stocking Index performance benchmark as the documented additionality path

Registries, corporate buyers, and third-party auditors get the same file. Nobody has to trust the seller.

Who Uses This

  • Corporate buyers hedging net-zero commitments who need credits that survive scrutiny by auditors and NGOs
  • Asset managers underwriting carbon removal as a distinct asset class with quantified downside
  • Project developers who want premium pricing for independently verified credits
  • Registries and rating agencies adding satellite-grade MRV to their certification stack
Satellite-Based Carbon MRV | Floodlight Knowledge Hub