CO₂ and Carbon Monitoring from Satellites: Measuring the Invisible

In 2019, OCO-2 data revealed something unexpected over a region in central Africa: CO₂ concentrations were significantly higher than models predicted, not from fossil fuels but from massive biomass burning during the dry season. The satellite measurements provided the first direct evidence that tropical fire emissions in this region had been systematically underestimated. That discovery — impossible without space-based CO₂ measurements — reshaped carbon cycle models for an entire continent.

Measuring CO₂ from space is arguably the most technically demanding task in Earth observation. The signal you're looking for is a 1-2% variation on top of a large background, through an atmosphere that's constantly changing. But the stakes — verifying national emissions, tracking the global carbon budget, and identifying unknown sources — make it one of the most consequential.

How Satellites Measure CO₂

Unlike trace gases like NO₂ or methane that have strong, distinctive absorption features, CO₂'s spectral signature is subtle and overlaps with water vapor. Satellites use shortwave infrared (SWIR) spectroscopy, analyzing sunlight that has passed through the atmosphere twice — down to the surface and back up to the sensor.

The measurement principle:

1. Sunlight passes through the full atmospheric column to the surface
2. The surface reflects some of this light back through the atmosphere
3. The satellite sensor measures the reflected spectrum at very high spectral resolution
4. CO₂ absorption features near 1.61 μm and 2.06 μm are analyzed
5. The depth and shape of these absorption lines reveal the total CO₂ in the atmospheric column

The result is XCO₂ — the column-averaged dry-air mole fraction of CO₂ — reported in parts per million (ppm).

Current Satellite Missions

OCO-2 (NASA, 2014-present)

The gold standard for space-based CO₂ measurement.

• Resolution: ~2.25 km × 1.29 km footprint
• Coverage: Along-track measurements only (narrow swath, ~10 km wide)
• Precision: ~1 ppm for individual soundings, <0.5 ppm for regional averages
• Revisit: 16-day repeat cycle (same ground track), but only useful in clear-sky conditions

OCO-2 has fundamentally changed our understanding of the carbon cycle, revealing seasonal breathing patterns of terrestrial ecosystems, quantifying urban CO₂ domes, and detecting enhanced concentrations downwind of large power plants.

GOSAT / GOSAT-2 (JAXA, 2009-present)

Japan's Greenhouse gases Observing SATellite was the first dedicated CO₂ monitoring satellite.

• Resolution: ~10.5 km footprint
• Coverage: Discrete point measurements on a grid pattern
• Precision: ~2 ppm for individual soundings
• Advantage: Also measures methane (CH₄), enabling combined greenhouse gas analysis

OCO-3 (NASA, ISS, 2019-present)

Mounted on the International Space Station, OCO-3 adds a unique capability: snapshot area mapping (SAM mode) that can image a 80 × 80 km area in a single overpass.

This mode is specifically designed for mapping CO₂ plumes from cities and industrial facilities, providing spatial context that OCO-2's narrow track cannot.

Upcoming: CO2M (ESA/Copernicus)

The CO₂ Monitoring Mission, planned for 2026, will be the first operational satellite designed specifically to monitor anthropogenic CO₂ emissions at national scales. Key improvements:

• Wide swath: ~250 km (vs. OCO-2's ~10 km)
• Simultaneous NO₂ measurement: NO₂ is co-emitted with CO₂ from combustion and serves as a tracer to separate fossil fuel CO₂ from biogenic CO₂
• Constellation of three satellites (CO2M-A, B, C) for near-daily coverage

The Detection Challenge

Global average atmospheric CO₂ is approximately 424 ppm (2025). The enhancement from a large city might add 2-5 ppm. A single power plant might add 1-3 ppm in its immediate downwind plume. Detecting these signals requires measurement precision better than 0.25% of the background — an extraordinary technical requirement.

Factors that complicate measurement:

• Clouds: Any cloud in the light path invalidates the measurement. Globally, only ~30% of satellite soundings are cloud-free
• Aerosols: Particles scatter light, changing the effective path length and biasing the CO₂ retrieval
• Surface albedo: Dark surfaces (water, dense forest) reflect less light, reducing signal-to-noise ratio
• Topography: Mountains alter the atmospheric column length, requiring precise elevation corrections

Applications

Urban Emissions Monitoring

OCO-2 and OCO-3 have detected CO₂ enhancements over cities including Los Angeles, Beijing, São Paulo, and Mumbai. By combining satellite XCO₂ with wind data and atmospheric transport models, researchers can estimate city-level CO₂ emission rates.

Current accuracy: Urban emission estimates from satellite data agree with inventory-based estimates within 20-30% for well-observed cities. This isn't precise enough to verify Paris Agreement commitments yet, but CO2M aims to reduce this uncertainty to ~10%.

Power Plant and Facility Monitoring

Individual large power plants (>10 Mt CO₂/year) can be detected by OCO-2 as localized XCO₂ enhancements downwind. Commercial satellites like GHGSat and Carbon Mapper are pushing toward facility-level monitoring with higher spatial resolution.

Carbon Cycle Science

The most impactful application to date: understanding how terrestrial ecosystems and oceans absorb CO₂. OCO-2 data has revealed:

• Tropical forests are more variable carbon sinks than previously thought
• El Niño events cause massive CO₂ releases from tropical ecosystems
• The Northern Hemisphere land sink is larger than estimated by ground-based networks
• Seasonal CO₂ patterns track photosynthesis and respiration cycles at continental scales

Verification of National Emissions

The Paris Agreement requires countries to report their greenhouse gas emissions, but reported inventories are based on activity data (fuel consumption, industrial output) and emission factors — not direct measurements. Satellite CO₂ monitoring will eventually provide an independent verification layer.

This is the primary motivation behind ESA's CO2M mission: creating a measurement system that can detect whether a country's actual emissions match its reported numbers.

Complementary: Solar-Induced Fluorescence (SIF)

OCO-2 accidentally became one of the most important tools for measuring photosynthesis from space. Plants emit a faint fluorescent glow (Solar-Induced Fluorescence, SIF) when photosynthesizing, detectable in the same spectral region where OCO-2 measures CO₂.

SIF provides a direct proxy for photosynthetic activity — fundamentally different from vegetation indices like NDVI, which measure greenness (a proxy for potential photosynthesis) rather than actual carbon uptake. The combination of CO₂ and SIF from the same sensor provides unprecedented insight into the carbon cycle: SIF shows where carbon is being absorbed, while XCO₂ shows the net atmospheric effect.

The Road Ahead

The next decade will transform satellite CO₂ monitoring from a scientific research tool to an operational monitoring system:

• CO2M constellation (ESA, 2026+) — first systematic monitoring of national emissions
• MicroCarb (CNES, 2025) — French microsatellite providing additional global XCO₂ measurements
• TanSat-2 (China, planned) — Chinese contribution to the global monitoring constellation
• Commercial sector — GHGSat, Carbon Mapper, and others providing facility-level monitoring

The convergence of these missions means that by 2030, atmospheric CO₂ will be monitored with sufficient coverage and precision to independently verify emission reduction commitments. For a gas that's invisible to the human eye, satellites are making it increasingly difficult to hide.