Back to overview

About Carbon-Eye

Carbon-Eye is a geospatial decision-support system that estimates above-ground biomass and carbon stock across Kenyan counties - combining satellite and environmental data with machine-learning ensembles in Google Earth Engine, as interactive maps and county statistics.

Mission

Put satellite-based carbon and biomass information directly in the hands of the people who act on it — county environment teams, conservation and restoration organisations, researchers, and early-stage carbon-project teams — without requiring a bespoke remote-sensing project every time. Select an area, get a validated estimate, understand exactly how reliable it is.

Vision

A future where any county, community or programme — in Kenya first, and beyond it over time — can see where its carbon is stored and where it's being lost, on demand and in plain language, so the scarce, expensive work of field verification gets pointed at exactly the places that need it most.

47
Kenyan counties
10 m
Sentinel-2 resolution
3
Models in the ensemble
9
Earth-observation datasets

What we do

Carbon-Eye runs a live analysis pipeline entirely inside Google Earth Engine: satellite and environmental data go in, three independently trained and validated machine-learning models come out, combined into a single ensemble estimate with its own reported accuracy. Nothing is pre-computed or illustrative — every run is a fresh analysis of the exact area and year requested. See the full methodology for the data sources, models and validation approach, or the guide for a walkthrough of the platform itself.

Who it's for

County environment & forestry teams
Conservation & restoration organisations
Researchers & students
Carbon-project teams (early screening)
Land-use planners & NGOs

Including organisations like Kenya Forest Service, KEFRI, NGOs, and students & academic researchers.

How we think about responsible use

Carbon-Eye is a screening and prioritisation tool, not a certified carbon registry. Before any estimate informs a crediting, investment or policy decision, we recommend:

  • Validate with local field data
  • Assess uncertainty
  • Check land tenure and baselines
  • Check leakage and permanence
  • Follow the relevant methodology

How it's built

Carbon-Eye is a small, focused stack rather than a large platform — each layer does one job and nothing runs that isn't needed to turn a county and a year into a validated carbon estimate.

Analysis engine

Google Earth Engine - every predictor stack, model run and map tile is computed live, not pre-baked.

Backend API

FastAPI (Python), deployed on Google Cloud Run, orchestrating Earth Engine jobs and serving results.

Frontend

Next.js (React, TypeScript), deployed on Vercel, rendering the interactive maps, charts and county explorer.

Storage & auth

Firestore for run history, Clerk for account and session management.

Built on open standards

Every dataset and constant behind Carbon-Eye is a published, citable standard — nothing proprietary or opaque in the pipeline itself.

Google Earth EngineCopernicus / ESANASAJAXAIPCC carbon fraction (0.47)

Suggested citation

Carbon-Eye (2026). Above-ground biomass & carbon stock across Kenyan counties - a geospatial decision-support system.

Open Carbon-EyeRead the guide