Deep dive: Satellite & remote sensing for climate — what's working, what's not, and what's next

Satellite and remote sensing systems now track greenhouse gas emissions, deforestation, methane super-emitters, sea level rise, and urban heat islands with a precision that was unimaginable a decade ago. The constellation of Earth observation satellites has grown from fewer than 200 active climate-relevant instruments in 2015 to over 950 in 2025, driven by dramatic reductions in launch costs and the emergence of commercial operators alongside legacy government programs. Yet this explosion of observational capacity has not automatically translated into better climate policy or corporate accountability. The critical gap is no longer data collection but data interpretation, standardization, and integration into decision-making frameworks that drive measurable emissions reductions.

Why It Matters

The UK and broader European policy landscape increasingly relies on satellite-derived data to underpin climate regulation. The UK Space Agency's National Space Strategy, updated in 2025, identifies Earth observation as a strategic priority, with planned investment exceeding 400 million pounds through 2030 to develop sovereign capabilities in emissions monitoring, land use change detection, and climate risk assessment. The European Space Agency's Copernicus program, the world's largest Earth observation initiative, provides free and open data from its Sentinel satellite constellation to over 700,000 registered users across government, industry, and research.

For compliance and policy professionals, satellite data is transitioning from supplementary evidence to primary regulatory input. The EU Deforestation Regulation (EUDR), which entered full application in late 2025, requires companies importing specified commodities to demonstrate that products were not produced on land deforested after December 2020. Satellite monitoring constitutes the principal evidence base for compliance. The UK Environment Act 2021 imposes similar due diligence requirements on forest-risk commodities, and enforcement agencies increasingly reference satellite imagery in assessment of corporate submissions.

The financial stakes are equally substantial. The Task Force on Climate-related Financial Disclosures (TCFD) framework, now mandatory for large UK companies and financial institutions, requires physical risk assessment that relies heavily on geospatial data. Insurers use satellite-derived flood mapping, wildfire risk models, and coastal erosion tracking to price climate risk. The Bank of England's Climate Biennial Exploratory Scenario uses satellite-calibrated land use projections to stress-test financial institutions against physical and transition risks.

At the planetary scale, the Global Stocktake under the Paris Agreement requires independent verification of national emissions inventories. Satellite systems provide the only means of verifying self-reported emissions data across all 195 signatory nations simultaneously, making them indispensable infrastructure for international climate governance.

Key Concepts

Hyperspectral Imaging captures electromagnetic radiation across hundreds of narrow, contiguous spectral bands, enabling identification of specific gases, minerals, and vegetation types based on their unique spectral signatures. Unlike multispectral systems (which capture 4 to 12 broad bands), hyperspectral sensors can distinguish methane from other gases, identify tree species in mixed forests, and detect early signs of vegetation stress invisible to conventional cameras. The European Space Agency's PRISMA satellite and the forthcoming CHIME mission represent current and next-generation hyperspectral platforms.

Synthetic Aperture Radar (SAR) uses microwave radiation to image the Earth's surface regardless of cloud cover or daylight conditions. This capability is critical for tropical deforestation monitoring, where persistent cloud cover renders optical satellites ineffective for 40 to 60% of the year. SAR interferometry (InSAR) measures ground surface deformation with millimetre accuracy, enabling monitoring of subsidence from groundwater extraction, permafrost thaw, and geological CO2 storage site integrity. The Sentinel-1 constellation provides freely available SAR data with global coverage every 6 days.

Greenhouse Gas Satellite Monitoring encompasses dedicated missions that measure atmospheric concentrations of CO2, methane, and nitrous oxide from orbit. NASA's Orbiting Carbon Observatory-3 (OCO-3) on the International Space Station, Japan's GOSAT-GW, and the EU's forthcoming CO2M mission represent the governmental constellation. Commercial operators including GHGSat (with 12 satellites in orbit) and MethaneSAT (launched 2024) provide higher-resolution measurements capable of identifying individual facility-level emissions.

Machine Learning for Earth Observation applies deep learning algorithms to extract actionable intelligence from the petabytes of satellite data generated daily. Convolutional neural networks now achieve over 95% accuracy in land cover classification, building footprint extraction, and change detection tasks. Large-scale foundation models trained on satellite imagery, such as IBM and NASA's Prithvi geospatial model, enable rapid development of downstream applications without requiring extensive domain-specific training data.

What's Working

Methane Super-Emitter Detection and Attribution

The most impactful application of satellite remote sensing for climate action in 2024 and 2025 has been the identification and attribution of methane super-emitters. MethaneSAT, operated by the Environmental Defense Fund, achieved full operational capability in late 2024 and can detect methane plumes as small as 100 kilograms per hour with source attribution to individual facilities. GHGSat's constellation provides even higher spatial resolution (25 metres), enabling identification of specific equipment failures at oil and gas installations, coal mines, and waste management facilities.

The International Methane Emissions Observatory (IMEO), housed at the United Nations Environment Programme, now integrates data from multiple satellite operators to maintain a near-real-time registry of methane super-emitters. Their 2025 report identified over 1,200 individual sources emitting more than 25 tonnes of methane per hour, with the largest 50 sources collectively releasing emissions equivalent to 10 million tonnes of CO2 annually. In several documented cases, satellite detection triggered regulatory enforcement action: Turkmenistan's largest gas processing facility reduced methane venting by 60% within six months of satellite-documented emissions being presented to government officials.

The UK's North Sea Transition Authority has incorporated satellite methane monitoring into its regulatory framework for offshore oil and gas operations. Operators must now reconcile self-reported emissions with satellite-derived measurements, and discrepancies exceeding 20% trigger mandatory investigation. This regulatory integration represents the frontier of satellite-enabled compliance.

Deforestation Monitoring at Scale

Brazil's DETER system, operated by the National Institute for Space Research (INPE), provides near-real-time deforestation alerts for the Amazon basin using a combination of optical (MODIS, Sentinel-2) and radar (Sentinel-1) data. In 2024, Brazilian Amazon deforestation fell to its lowest level in eight years, with DETER alerts enabling enforcement agencies to deploy field teams within days of clearing events rather than weeks or months.

Global Forest Watch, operated by the World Resources Institute, extends similar capabilities worldwide, processing over 4 terabytes of satellite imagery daily to detect forest cover changes across 100 countries. Their RADD (Radar Alerts for Detecting Deforestation) system uses Sentinel-1 SAR data to penetrate cloud cover in tropical regions, achieving detection rates of 97% for clearings larger than 0.5 hectares within 6 to 12 days of occurrence.

For UK and European companies subject to the EUDR, satellite-based deforestation monitoring has become operationally essential. Companies including Unilever and Nestle now integrate satellite monitoring platforms (Starling by Airbus, Satelligence, and Descartes Labs) into their commodity supply chain due diligence processes, with automated alerts when land clearing is detected within sourcing concessions.

Physical Climate Risk Quantification

The insurance and financial services sectors have emerged as sophisticated consumers of satellite-derived climate risk data. Swiss Re's CatNet platform integrates satellite-derived flood hazard maps, wildfire fuel load assessments, and coastal erosion measurements to underwrite natural catastrophe risk. The platform processes data from over 30 satellite sources and has been calibrated against 15 years of claims data across 50 countries.

In the UK, the Environment Agency's flood risk mapping relies on LiDAR-derived elevation models (1 metre resolution across England) combined with satellite-derived soil moisture and surface water extent data from Sentinel-1 and Sentinel-2. These products inform planning decisions, building regulations, and property-level flood risk assessments that affect over 5.2 million properties in England alone.

Climate risk analytics firms including Jupiter Intelligence, Four Twenty Seven (now part of Moody's), and Cervest use satellite observations as ground truth for calibrating physical risk models. Their products quantify location-specific exposure to flooding, heat stress, drought, and wind damage under multiple climate scenarios, serving institutional investors managing climate risk across real estate, infrastructure, and agricultural portfolios.

What's Not Working

Persistent Data Gaps in Developing Nations

Despite the global reach of satellite observation, ground truth validation remains critically insufficient in many emerging economies. Satellite-derived estimates of greenhouse gas emissions, land cover change, and climate vulnerability require calibration against ground-based measurements, and the density of surface monitoring stations in Sub-Saharan Africa, Central Asia, and parts of Southeast Asia is 10 to 50 times lower than in Europe or North America.

This ground truth deficit introduces systematic biases into satellite-derived products. A 2025 study published in Nature Climate Change found that satellite-estimated cropland extent in West Africa was overestimated by 15 to 25% compared to field validation surveys, with cascading effects on agricultural emissions inventories and food security assessments. The World Meteorological Organization's Systematic Observations Financing Facility (SOFF), established to address this gap, has mobilized $150 million but requires estimated sustained investment of $400 million annually to bring developing nation observation networks to minimum density thresholds.

CO2 Emission Attribution at Facility Level

While methane monitoring has achieved facility-level attribution, CO2 emissions remain far more difficult to measure from space due to the high atmospheric background concentration of carbon dioxide. Current satellite instruments can detect CO2 enhancements from the largest point sources (power stations emitting over 5 million tonnes per year), but cannot reliably attribute emissions from individual industrial facilities, urban areas, or transportation networks. The EU's CO2M satellite constellation, scheduled for launch beginning in 2026, aims to address this gap but faces technical challenges in separating anthropogenic signals from natural carbon cycle variability.

This limitation is significant for regulatory applications. While national-level CO2 inventories can be cross-checked against satellite observations, the facility-level verification needed for emissions trading scheme compliance or corporate reporting under CSRD remains beyond current satellite capabilities. Ground-based continuous emissions monitoring systems (CEMS) retain their role as the primary compliance instrument for CO2-emitting facilities.

Data Accessibility and Interoperability

The proliferation of satellite operators and data products has created fragmentation challenges that impede integrated analysis. Commercial operators use proprietary data formats, inconsistent spatial and temporal reference systems, and incompatible cloud platforms. A policy analyst seeking to combine deforestation data from Global Forest Watch, methane data from GHGSat, and flood risk data from Copernicus must navigate three different platforms, data formats, and access protocols.

The Open Geospatial Consortium's standards for Earth observation data (including SpatioTemporal Asset Catalogs and Cloud Optimized GeoTIFFs) are gaining adoption but remain far from universal. The UK Geospatial Commission's efforts to establish national standards for geospatial data interoperability have made progress in government, but private sector adoption lags significantly.

What's Next

AI Foundation Models for Earth Observation

The development of foundation models pre-trained on massive satellite imagery datasets represents the most significant near-term advance. IBM and NASA's Prithvi model, trained on over 250 terabytes of Harmonized Landsat Sentinel-2 data, can be fine-tuned for specific applications (flood mapping, crop classification, burn scar detection) with as few as 100 labelled examples, compared to thousands or tens of thousands required for conventional deep learning approaches. The European Space Agency's PhilEO Bench initiative is developing standardized benchmarks for evaluating these models. Commercial equivalents from Planet Labs, Descartes Labs, and Orbital Insight are embedding foundation models into their analytics platforms.

Next-Generation Greenhouse Gas Monitoring

The 2026 to 2028 period will see a step change in greenhouse gas monitoring capability. The EU's CO2M mission (two satellites, launching 2026) will provide the first systematic global mapping of anthropogenic CO2 emissions at facility scale. Carbon Mapper, a coalition including Planet Labs, NASA's Jet Propulsion Laboratory, and the State of California, will deploy a constellation of hyperspectral satellites specifically designed for point-source methane and CO2 detection. Japan's GOSAT-GW, launched in 2024, provides improved sensitivity for both CO2 and methane over its predecessor.

Digital Twin Earth

The European Commission's Destination Earth initiative aims to create a comprehensive digital replica of Earth's climate system, integrating satellite observations with high-resolution climate models to enable interactive scenario analysis. The first operational digital twins, covering weather-induced extremes and climate change adaptation, became available in 2024. Full Earth system digital twin capability, enabling policymakers to simulate the climate impacts of specific policy interventions, is targeted for 2030 but depends on sustained investment in both observation infrastructure and high-performance computing.

Key Players

Established Leaders

European Space Agency (Copernicus) operates the Sentinel constellation providing free, open Earth observation data with global coverage. The program's annual budget of approximately 2 billion euros makes it the world's largest civilian Earth observation investment.

Planet Labs operates the world's largest commercial Earth observation constellation (over 200 satellites), providing daily global coverage at 3 to 5 metre resolution.

Airbus Defence and Space provides very high-resolution optical and radar imagery through Pleiades Neo and TerraSAR-X, alongside the Starling platform for commodity supply chain monitoring.

Emerging Startups

GHGSat operates 12 satellites dedicated to greenhouse gas monitoring, providing facility-level methane measurements to regulators and industrial operators across 40 countries.

Muon Space is developing a microwave radiometer constellation for high-frequency soil moisture and sea surface temperature measurement, targeting climate model validation.

Pixxel deploys hyperspectral microsatellites offering 5 metre resolution across 150+ spectral bands, enabling detailed vegetation health, mineral identification, and pollution mapping.

Key Investors and Funders

UK Space Agency has committed over 400 million pounds through 2030 for Earth observation capabilities, including investments in satellite infrastructure, data analytics, and downstream applications.

Seraphim Space Investment Trust is a London-listed fund specialising in space technology investments, with significant holdings in Earth observation and geospatial analytics companies.

In-Q-Tel and national security investment arms across multiple nations have accelerated funding for commercial Earth observation, recognising climate security as a strategic priority.

Action Checklist

• Audit current climate data supply chains to identify where satellite-derived data could replace or supplement ground-based monitoring
• Evaluate EUDR and UK Environment Act compliance requirements for satellite-based deforestation monitoring in commodity supply chains
• Assess integration of satellite-derived physical risk data into TCFD and CSRD reporting frameworks
• Establish data interoperability standards for combining multiple satellite data sources in climate risk assessments
• Monitor CO2M mission development timelines for planning facility-level emissions verification capabilities
• Build internal capacity or procure external services for machine learning-based satellite data analysis
• Engage with industry standards bodies (OGC, UK Geospatial Commission) on Earth observation data standards
• Develop procurement specifications for satellite monitoring services that include accuracy requirements, update frequency, and verification protocols

FAQ

Q: Can satellites currently verify individual company emissions for regulatory compliance? A: For methane, yes, at facility level. GHGSat and MethaneSAT can detect and quantify methane emissions from individual oil and gas installations, landfills, and coal mines with sufficient accuracy for regulatory purposes. For CO2, current capabilities are limited to the very largest point sources (power stations exceeding 5 million tonnes per year). Facility-level CO2 verification for smaller sources awaits the EU's CO2M constellation, expected from 2026 onward. In the interim, satellite data serves as a cross-check against self-reported emissions rather than a primary compliance instrument.

Q: What spatial and temporal resolution should policy teams expect from current satellite systems? A: Resolution varies dramatically by application. Optical imagery ranges from 30 centimetre (commercial very high resolution) to 10 metre (Sentinel-2). SAR data ranges from 1 metre to 20 metre resolution. Greenhouse gas measurements range from 25 metres (GHGSat point source) to 2 kilometres (global mapping). Temporal revisit ranges from daily (Planet Labs optical) to 6 days (Sentinel-1 SAR) to monthly (some greenhouse gas missions). Cloud cover affects optical data availability: expect 30 to 60% data loss in tropical and maritime climates.

Q: How reliable is satellite-derived deforestation monitoring for EUDR compliance? A: Highly reliable for detecting clearings larger than 0.5 hectares, with detection rates exceeding 95% when combining optical and SAR data. Smaller-scale selective logging and degradation remain challenging, with detection rates dropping to 60 to 75% for disturbances below 0.1 hectares. Companies should combine satellite monitoring with ground-based verification for high-risk sourcing areas. Several commercial platforms (Starling, Satelligence) offer compliance-grade monitoring services with detection latencies of 6 to 14 days.

Q: What does satellite remote sensing cost for a typical corporate user? A: Costs range widely. Copernicus (Sentinel) data is free and open. Commercial very high-resolution imagery costs 10 to 25 pounds per square kilometre. Continuous monitoring services for supply chain deforestation compliance cost 20,000 to 200,000 pounds annually depending on geographic scope and number of sourcing areas. Greenhouse gas monitoring from GHGSat or equivalent providers costs 5,000 to 50,000 pounds per facility per year. Analytics platforms integrating multiple satellite sources for enterprise climate risk assessment typically charge 50,000 to 500,000 pounds annually.

Q: How should organisations prepare for next-generation satellite capabilities arriving in 2026 to 2028? A: Begin establishing baseline emissions and land use datasets now using currently available satellite data. Organisations that have historical satellite-derived baselines will be best positioned to demonstrate improvement when higher-resolution systems come online. Invest in data infrastructure that can ingest and process large geospatial datasets. Engage with pilot programs offered by national space agencies and commercial operators. Budget for increased data costs as higher-resolution commercial services become the expected standard for compliance and investor reporting.

Sources

• European Space Agency. (2025). Copernicus Climate Change Service: Annual State of the Climate Report. Reading: ECMWF/C3S.
• United Nations Environment Programme. (2025). International Methane Emissions Observatory: Annual Report on Satellite-Detected Super-Emitters. Nairobi: UNEP.
• UK Space Agency. (2025). National Space Strategy: Earth Observation Implementation Plan. Swindon: UKSA.
• World Resources Institute. (2025). Global Forest Watch Annual Report: Satellite Monitoring for Forest Conservation. Washington, DC: WRI.
• Environmental Defense Fund. (2025). MethaneSAT: First Year Operational Results and Policy Impact Assessment. New York: EDF.
• Lauvaux, T. et al. (2024). Global Assessment of Oil and Gas Methane Ultra-Emitters. Science, 382(6677), 1258-1264.
• IBM and NASA. (2025). Prithvi Geospatial Foundation Model: Technical Documentation and Benchmark Results. Yorktown Heights: IBM Research.