Deep dive: Earth observation satellites & climate analytics — what's working, what's not, and what's next

The European Space Agency's Copernicus Climate Change Service processes over 80 terabytes of satellite data daily, feeding climate models that now cover 98% of the Earth's land surface with sub-kilometre resolution (Copernicus, 2025). That throughput represents a 20-fold increase from a decade ago and reflects a fundamental shift: earth observation (EO) has moved from a research curiosity to operational infrastructure underpinning climate policy, carbon markets, disaster response, and investment decisions across every sector of the global economy.

Why It Matters

Climate change is no longer a future projection. It is a present-day risk factor embedded in supply chains, insurance portfolios, agricultural yields, and urban infrastructure. Managing that risk requires observation systems capable of measuring atmospheric composition, land-use change, ice mass loss, ocean temperatures, and ecosystem health at scales ranging from individual facilities to entire continents, and at cadences ranging from hourly to decadal.

Ground-based monitoring networks remain essential, but they cover a fraction of the planet's surface. The World Meteorological Organization reports that Africa has only one-eighth of the minimum recommended density of weather and climate observation stations, and large parts of Central Asia, the Amazon basin, and the Arctic have effectively no continuous ground-based monitoring (WMO, 2025). Satellites fill this gap by providing consistent, repeatable, and geographically comprehensive measurements regardless of national infrastructure or political cooperation.

The economic value is significant. A 2024 study by the European Commission estimated that Copernicus data and services generate approximately 5.1 billion euros in annual downstream economic value across sectors including agriculture, energy, insurance, and urban planning (European Commission, 2024). The US National Oceanic and Atmospheric Administration attributes over $30 billion in annual avoided weather-related losses to satellite-enabled forecasting improvements (NOAA, 2025). For climate-focused founders and investors in the UK, satellite-derived analytics represent both a growing market opportunity and an increasingly essential input for regulatory compliance, carbon accounting, and physical risk assessment.

Key Concepts

Radiometric resolution describes the sensitivity of a satellite sensor to differences in the energy it detects. Higher radiometric resolution enables finer discrimination between land cover types, vegetation health states, or atmospheric gas concentrations. The Sentinel-2 satellites, for example, capture data in 13 spectral bands at 12-bit radiometric resolution, allowing detection of subtle changes in crop health or forest canopy density that lower-resolution instruments would miss.

Revisit time is the interval between successive observations of the same location by a satellite or constellation. Revisit time determines how quickly changes can be detected and is critical for applications like deforestation monitoring, flood mapping, and emissions tracking. Single satellites typically revisit every 5 to 16 days, while commercial constellations such as Planet Labs achieve daily global coverage.

Analysis-ready data (ARD) refers to satellite imagery that has been processed to remove atmospheric interference, geometric distortions, and sensor artefacts, making it suitable for direct use in analytical workflows without further preprocessing. The shift toward ARD provision by agencies and commercial providers has dramatically lowered the barrier to entry for non-specialist users, enabling climate analytics startups to focus on insights rather than data engineering.

Climate Essential Variables (ECVs) are the physical, chemical, and biological variables defined by the Global Climate Observing System as essential for systematic observation of Earth's climate. There are currently 54 ECVs spanning atmosphere, ocean, and land domains, and satellite observations contribute to monitoring at least 40 of them (GCOS, 2025).

What's Working

Copernicus and Open Data as a Public Good

The European Union's Copernicus programme remains the gold standard for open-access earth observation, and its impact on climate analytics has been transformative. The Sentinel satellite constellation, comprising six mission families, provides free and open data covering atmospheric composition (Sentinel-5P), land surface characteristics (Sentinel-2), ocean dynamics (Sentinel-3), and radar-based surface deformation (Sentinel-1). By early 2026, the Copernicus Data Space Ecosystem had registered over 680,000 active users across 190 countries, with data access volumes growing at 35% year on year (Copernicus, 2025).

The Copernicus Climate Change Service (C3S), operated by ECMWF, delivers climate reanalysis datasets, seasonal forecasts, and climate projection tools that are used by national meteorological services, reinsurance companies, and infrastructure planners worldwide. The ERA5 reanalysis dataset, which provides hourly estimates of atmospheric variables from 1940 to the present at 31 km resolution, has become the de facto reference dataset for climate research and was cited in over 12,000 peer-reviewed publications in 2025 alone.

For UK-based founders, Copernicus represents a massive free data layer that can underpin commercial climate analytics products without licensing costs. Companies such as Cervest (now part of Moody's) built their entire climate risk analytics platform on Copernicus and related open datasets before adding proprietary modelling layers.

Commercial Constellations Enabling Daily Monitoring

Planet Labs operates the largest commercial earth observation constellation, with over 200 satellites providing daily coverage of the entire Earth's land surface at 3 to 5 metre resolution. This cadence has transformed applications that previously relied on weekly or monthly satellite revisits. In the climate domain, Planet's data is used for near-real-time deforestation alerts, agricultural carbon project verification, and post-disaster damage assessment.

The Global Forest Watch platform, operated by the World Resources Institute and powered in part by Planet imagery, detected 4.1 million hectares of tropical primary forest loss in 2024, with alert latency reduced to under 5 days from detection to publication (Global Forest Watch, 2025). This speed enables enforcement agencies and civil society organisations to respond while illegal clearing is still in progress, rather than discovering it months later in annual forest inventories.

In carbon markets, satellite-based monitoring has become essential for credibility. Verra's updated Verified Carbon Standard requires projects involving avoided deforestation (REDD+) to demonstrate additionality and permanence using satellite-derived land cover change data at intervals no greater than annual. Pachama, Sylvera, and other carbon credit rating platforms use Planet, Sentinel-2, and Airbus imagery combined with LiDAR data to independently verify the carbon stocks claimed by forest carbon projects, flagging discrepancies that have led to credit invalidation and programme reforms.

Satellite-Derived Climate Risk Analytics for Finance

The UK's Financial Conduct Authority and the Bank of England were among the first regulators globally to require climate-related financial disclosures aligned with the Task Force on Climate-related Financial Disclosures (TCFD) framework. Meeting these requirements has driven demand for satellite-derived physical climate risk analytics that can assess portfolio exposure to flooding, heat stress, wildfire, and coastal erosion at the asset level.

Companies including Climate X, Jupiter Intelligence, and Moody's (incorporating the former Cervest platform) use satellite observations of land subsidence, sea level rise, vegetation dryness indices, and urban heat island effects as inputs to probabilistic risk models. Climate X, a London-based startup, provides property-level flood and subsidence risk scores derived from Sentinel-1 radar interferometry and Sentinel-2 land cover data, serving UK mortgage lenders and insurers who must quantify physical climate risk across their loan books.

The insurance sector's adoption has been particularly rapid. Swiss Re's CatNet platform integrates satellite-derived exposure data with catastrophe models, and Lloyd's of London now requires syndicates to demonstrate how satellite and geospatial data inform their natural catastrophe risk assessments. A 2025 review by the Prudential Regulation Authority found that 78% of UK insurers reporting under Solvency II were using satellite-derived data in at least one component of their climate risk assessment process (PRA, 2025).

What's Not Working

Fragmented Data Ecosystems and Interoperability Gaps

Despite the volume of available satellite data, integrating observations from multiple sources into coherent analytical workflows remains a significant challenge. Each satellite mission produces data in different formats, coordinate systems, spectral bands, and temporal cadences. Combining Sentinel-2 optical imagery with Planet's SuperDove data, Landsat archives, and commercial radar products requires substantial data engineering investment that smaller analytics teams and startups struggle to resource.

Cloud platforms such as Google Earth Engine, Microsoft Planetary Computer, and the Copernicus Data Space Ecosystem have reduced preprocessing burdens, but they impose their own constraints around computational costs, data export limitations, and vendor lock-in. Founders building climate analytics products must choose between platform dependency and the engineering overhead of self-hosted data pipelines, and neither option is ideal for early-stage companies operating with limited capital.

Standards for analysis-ready data exist (the Committee on Earth Observation Satellites' CEOS-ARD specification), but adoption across commercial providers remains inconsistent. The result is that a significant portion of the climate analytics value chain is consumed by data wrangling rather than insight generation.

Calibration and Validation Deficits

Satellite-derived climate variables require rigorous ground-truthing to maintain accuracy over time. Sensor degradation, orbital drift, and changes in atmospheric conditions can introduce systematic biases that propagate into downstream analytics if not corrected. The Global Climate Observing System has identified persistent calibration gaps for several ECVs, particularly soil moisture, above-ground biomass, and ocean colour, where reference networks are sparse or underfunded (GCOS, 2025).

For commercial applications, the validation problem is particularly acute in carbon markets. Satellite estimates of forest carbon stocks rely on allometric models that convert canopy height and density measurements into biomass estimates, but these models are calibrated primarily on temperate and boreal forests. Their accuracy degrades significantly in tropical forests, where species diversity, canopy structure, and wood density vary enormously. Studies have documented errors of 30 to 50% in satellite-derived biomass estimates for tropical forests when compared with ground-based inventory plots (Duncanson et al., 2023).

Latency Between Observation and Decision

While satellite revisit times have improved dramatically, the time between data acquisition and the delivery of actionable insights to decision-makers remains a bottleneck. Raw satellite imagery must be downloaded, processed, quality-controlled, and analysed before it yields useful information. For many climate applications, this pipeline takes days to weeks, which is adequate for trend monitoring but insufficient for real-time operational decisions such as flood response, wildfire suppression, or emissions event detection.

The challenge is compounded by the computational demands of modern climate analytics. Machine learning models trained to detect deforestation, map flood extent, or classify crop types require substantial GPU resources and skilled operators. Automated pipelines exist but are often fragile, requiring manual intervention when cloud cover, sensor anomalies, or algorithm failures disrupt processing.

Downstream Translation and Accessibility

A persistent gap exists between the sophistication of satellite-derived climate data and the capacity of end users to interpret and act on it. Municipal planners, agricultural extension officers, and small business owners rarely have the technical skills to work directly with satellite data products. While platforms like Google Earth Engine have democratised access for researchers, the non-technical user base that most needs climate analytics remains underserved.

In the UK context, local authorities responsible for climate adaptation planning often lack both the staff and the budget to integrate satellite-derived flood risk, heat exposure, and subsidence data into spatial planning decisions. The Geospatial Commission's 2025 review found that only 34% of English local planning authorities were actively using satellite or geospatial data in their climate adaptation strategies (Geospatial Commission, 2025).

Key Players

Established Organisations

• European Space Agency (ESA): Operates the Copernicus Sentinel constellation and funds EO research programmes through the Climate Change Initiative
• NASA: Operates Landsat (jointly with USGS), EMIT, OCO-3, and multiple Earth science missions providing open-access climate data
• EUMETSAT: Operates the Meteosat and MetOp satellite programmes providing meteorological and climate monitoring for Europe and Africa
• UK Space Agency: Funds the National Centre for Earth Observation and supports commercial EO development through the Space for Climate programme

Startups and Growth-Stage Companies

• Planet Labs: Operates the largest commercial EO constellation with daily global coverage at 3 to 5 metre resolution
• Climate X: London-based platform providing asset-level physical climate risk scores using satellite-derived data for financial institutions
• Sylvera: Uses satellite data and machine learning to rate carbon credit quality, with a focus on nature-based solutions
• Pachama: Combines satellite imagery and LiDAR to verify forest carbon projects and detect deforestation risks
• Descartes Labs: Provides a geospatial analytics platform for ingesting and analysing satellite data at scale

Investors and Funders

• Seraphim Space Investment Trust: London-listed trust specialising in space technology investments including EO analytics companies
• UK Research and Innovation (UKRI): Funds EO and climate analytics research through NERC and Innovate UK programmes
• European Investment Bank: Finances EO infrastructure and downstream applications through the InvestEU programme

Action Checklist

• Audit current climate data sources and identify where satellite-derived inputs could replace or supplement ground-based measurements
• Register for Copernicus Data Space Ecosystem access and evaluate Sentinel data products relevant to your sector
• Assess physical climate risk exposure across facilities and supply chain using satellite-derived analytics platforms
• Implement satellite-based monitoring for carbon projects, deforestation-free supply chain commitments, or methane emissions tracking
• Develop internal capacity to interpret and act on satellite-derived climate insights, or partner with specialist analytics providers
• Engage with UK Space Agency programmes and Innovate UK funding calls supporting commercial applications of EO data
• Integrate satellite-derived land use and vegetation data into Scope 3 emissions calculations for agricultural and forestry supply chains
• Evaluate emerging satellite data products from commercial providers for competitive intelligence and market monitoring applications

FAQ

Q: What satellite data is freely available for climate analytics? A: The Copernicus Sentinel missions provide free and open data covering optical imagery (Sentinel-2), radar (Sentinel-1), atmospheric composition (Sentinel-5P), and ocean monitoring (Sentinel-3). NASA's Landsat archive (50+ years of imagery), EMIT methane and mineral data, and MODIS/VIIRS products are also freely available. The Copernicus Climate Change Service provides processed climate reanalysis datasets and projections at no cost. These open datasets form the foundation for most commercial climate analytics products.

Q: How are UK regulators using satellite data in climate policy? A: The Financial Conduct Authority and Bank of England require TCFD-aligned climate risk disclosures, which increasingly rely on satellite-derived physical risk data. The Environment Agency uses satellite imagery for flood extent mapping and catchment monitoring. Defra's Environmental Land Management schemes use satellite data to verify farmer compliance with environmental commitments. The Geospatial Commission is actively promoting greater use of EO data across government decision-making, though local authority adoption remains uneven.

Q: What are the main limitations of satellite-based climate monitoring? A: Cloud cover remains a persistent challenge for optical sensors, particularly in tropical regions where it can obscure 60 to 80% of observations during wet seasons. Spatial resolution limits the detection of small-scale features such as individual field boundaries or small emission sources. Calibration and validation against ground-truth data are essential but often underfunded. Data latency between acquisition and actionable insight delivery ranges from days to weeks for most applications. Finally, translating satellite data into formats accessible to non-specialist decision-makers remains an ongoing challenge.

Q: What emerging missions will improve EO capabilities for climate? A: The Copernicus CO2M mission (scheduled for 2026) will measure anthropogenic CO2 emissions at the national and city level for the first time from space. ESA's BIOMASS mission (launched 2025) uses P-band radar to measure forest biomass in tropical regions where optical sensors struggle. The NISAR mission (NASA/ISRO joint venture) will provide systematic radar observations of land surface deformation, ice dynamics, and ecosystem change. Commercial constellations continue expanding, with Planet, Satellogic, and others planning hyperspectral and thermal infrared missions that will add new measurement capabilities.

Q: How should founders approach building products on satellite data? A: Start with open datasets from Copernicus and NASA to validate product-market fit without data licensing costs. Use cloud-native platforms (Google Earth Engine, Microsoft Planetary Computer) for prototyping to minimise infrastructure investment. Focus on the translation layer between raw data and decision-maker needs, as this is where the greatest value capture opportunity exists. Plan for data interoperability challenges from the outset by building flexible data ingestion pipelines. Consider UK Space Agency grants and Innovate UK programmes as non-dilutive funding sources for EO-based product development.

Sources

• Copernicus. (2025). Copernicus Data Space Ecosystem: Annual Usage Report 2025. European Commission.
• European Commission. (2024). Copernicus Market Report: Downstream Economic Value Assessment. Brussels: EC Directorate-General for Defence Industry and Space.
• World Meteorological Organization. (2025). State of Climate Services 2025: Observation Gaps and Infrastructure Needs. Geneva: WMO.
• NOAA. (2025). Satellite Data and Services: Economic Impact Assessment. Washington, DC: National Oceanic and Atmospheric Administration.
• Global Forest Watch. (2025). 2024 Tree Cover Loss Data and Analysis. Washington, DC: World Resources Institute.
• GCOS. (2025). The Status of the Global Climate Observing System 2025. Geneva: Global Climate Observing System.
• Duncanson, L. et al. (2023). Aboveground biomass density models for NASA's Global Ecosystem Dynamics Investigation (GEDI). Remote Sensing of Environment, 298, 113845.
• Prudential Regulation Authority. (2025). Climate-Related Financial Risk: Satellite and Geospatial Data Use in UK Insurance. London: Bank of England.
• Geospatial Commission. (2025). Geospatial Data in Local Authority Climate Adaptation: Usage Review. London: Cabinet Office.