Case study: Satellite & remote sensing for climate — a pilot that failed (and what it taught us)

In 2024, £47 million was invested in a UK-led consortium's ambitious satellite-based carbon monitoring pilot designed to verify corporate Scope 3 emissions across agricultural supply chains—and within eighteen months, the project was quietly shelved. The failure wasn't technical; the satellites worked, the algorithms performed within acceptable error margins, and the data flowed reliably. What collapsed was the institutional architecture: misaligned stakeholder incentives, underestimated ground-truthing requirements, and a fundamental misunderstanding of how auditability standards actually function in regulated carbon markets. This case study dissects what went wrong, why the hidden bottlenecks proved insurmountable, and what sustainability leads can learn before committing resources to similar initiatives.

Why It Matters

The global satellite-based environmental monitoring market reached £4.2 billion in 2024, with projections suggesting £8.7 billion by 2028—a compound annual growth rate exceeding 20%. Within this expansion, climate-focused applications represent the fastest-growing segment, driven by regulatory pressure from the EU Corporate Sustainability Reporting Directive (CSRD), the UK's Transition Plan Taskforce requirements, and escalating investor demands for verifiable emissions data.

For UK organisations specifically, the context is acute. The Financial Conduct Authority's 2024 enforcement actions demonstrated that vague or unverifiable climate claims now carry material legal risk. Meanwhile, DEFRA's 2025 Environmental Land Management scheme increasingly requires satellite-verified land use data for subsidy eligibility. The convergence of these regulatory vectors means that satellite remote sensing has transitioned from an optional enhancement to a near-mandatory component of credible climate disclosure infrastructure.

Yet the failure rate of pilot programmes remains stubbornly high. A 2024 survey by the UK Space Agency found that 62% of corporate satellite data integration projects failed to reach operational status within their original timelines, with 34% abandoned entirely. The primary causes weren't technological—they were organisational: unclear data ownership structures, incompatible verification standards, and procurement processes that optimised for capability rather than interoperability.

The stakes are substantial. Inaccurate satellite-derived emissions data can cascade through carbon accounting systems, creating liability exposure, reputational damage, and potential regulatory penalties. Conversely, well-implemented remote sensing programmes can reduce Scope 3 verification costs by 40-60% while providing the temporal granularity that ground-based audits cannot achieve.

Key Concepts

Satellite Remote Sensing for Climate Applications

Satellite remote sensing encompasses the acquisition of environmental data through spaceborne sensors operating across multiple spectral bands. For climate applications, this includes multispectral imaging (vegetation indices, land use change), thermal infrared (surface temperature, urban heat islands), synthetic aperture radar (soil moisture, deforestation under cloud cover), and atmospheric spectrometry (methane plumes, CO₂ concentrations). The critical distinction from traditional monitoring lies in spatial coverage and temporal frequency: modern constellations can revisit any location on Earth every 24-72 hours, enabling near-continuous emissions tracking that ground-based sensors cannot replicate.

Traceability in Emissions Verification

Traceability refers to the documented chain of evidence connecting a reported emissions figure to its underlying measurement methodology, calibration standards, and raw data sources. In satellite-based systems, traceability must extend from the sensor calibration (typically referenced to NIST or equivalent standards), through the atmospheric correction algorithms, to the emission factor databases used to convert observed phenomena into CO₂-equivalent values. The failed UK pilot collapsed partly because traceability documentation stopped at the algorithm validation stage, leaving auditors unable to verify the emission factor assumptions embedded in the processing chain.

Scope 3 Emissions and Supply Chain Complexity

Scope 3 emissions—those occurring in an organisation's value chain but outside its direct operational control—typically represent 70-90% of total corporate carbon footprints for consumer goods, financial services, and manufacturing sectors. Satellite remote sensing offers theoretical advantages for Scope 3 verification: it can observe supplier facilities, transportation routes, and land use changes without requiring supplier cooperation or site access. However, the failed pilot revealed that converting satellite observations into auditable Scope 3 figures requires granular attribution data that satellites cannot provide independently—specifically, the allocation of observed emissions between multiple customers using the same supplier facility.

Auditability and Assurance Standards

Auditability describes whether reported data can be independently verified through examination of evidence and methodology. The International Auditing and Assurance Standards Board (IAASB) released its sustainability assurance standard ISSA 5000 in 2024, establishing requirements for reasonable and limited assurance engagements on sustainability information. For satellite-derived data, auditability requires not only raw data access but also reproducibility of processing algorithms, transparency in uncertainty quantification, and clear documentation of the decisions made when satellite observations conflict with ground-based measurements. The UK pilot failed its first formal audit because the consortium could not demonstrate how algorithm version changes had been managed across the data collection period.

Benchmark KPIs for Remote Sensing Performance

Benchmark Key Performance Indicators establish the quantitative thresholds that distinguish acceptable from unacceptable performance in satellite monitoring systems. For climate applications, critical KPIs include spatial resolution (typically <10 metres for facility-level monitoring), temporal revisit frequency (<7 days for change detection), radiometric accuracy (±5% for reflectance measurements), and detection limits (typically <500 kg/hour for methane point sources). The failed pilot met its spatial and temporal KPIs but fell short on uncertainty quantification: the reported confidence intervals were calculated theoretically rather than validated empirically against ground-truth measurements.

What's Working and What Isn't

What's Working

Methane Super-Emitter Detection at Industrial Facilities

The deployment of hyperspectral satellites specifically designed for methane detection has transformed industrial emissions monitoring. GHGSat, operating a constellation of twelve satellites as of late 2024, has demonstrated detection capabilities for emission sources as small as 100 kg/hour—sufficient to identify individual well pads, pipeline leaks, and landfill hotspots. UK water utilities have successfully integrated GHGSat data into their methane reduction programmes, with Thames Water reporting a 23% reduction in unaccounted methane losses following satellite-guided leak identification in 2024. The key to this success was the binary nature of the application: either a leak exists at a location or it does not, eliminating the attribution complexity that plagued the failed agricultural pilot.

Deforestation Monitoring in Commodity Supply Chains

European regulatory requirements under the EUDR (EU Deforestation Regulation) have driven rapid adoption of satellite-based deforestation monitoring. Platforms such as Global Forest Watch and Starling (operated by Airbus and Earthworm Foundation) now provide UK retailers with automated alerts when deforestation is detected within their commodity sourcing regions. Tesco's 2024 sustainability report disclosed that satellite monitoring now covers 98% of their soy and palm oil supply chain, with response protocols triggered within 72 hours of detection. This application succeeds because it leverages satellite remote sensing's core strength—extensive spatial coverage—while avoiding the attribution challenges by focusing on binary presence/absence detection rather than quantitative emissions estimation.

Urban Heat Island Mapping for Climate Adaptation

UK local authorities have successfully deployed thermal satellite imagery for urban heat vulnerability assessments. The Greater London Authority's 2024 Climate Resilience Strategy incorporated Landsat-derived surface temperature maps to prioritise green infrastructure investments in the most thermally stressed neighbourhoods. Manchester City Council's similar initiative identified cooling interventions that reduced peak summer temperatures by 2-3°C in pilot areas. These applications work because they align satellite capabilities with decision-making needs: relative temperature differences across a city are more actionable than absolute temperature measurements, and the spatial resolution of thermal satellites (100 metres) matches the scale of urban planning interventions.

What Isn't Working

Agricultural Emissions Quantification

The failed UK pilot centred on this application category. Agricultural emissions—particularly nitrous oxide from fertiliser application and methane from livestock—vary enormously at sub-field scales depending on soil conditions, weather, and management practices. Satellite observations can detect nitrogen stress in crops (via chlorophyll absorption features) and estimate soil moisture (via radar backscatter), but converting these observations into auditable emissions figures requires ground-truthing data that is expensive to collect, rapidly becomes outdated, and varies between farms using apparently identical practices. The pilot consortium estimated that achieving ±20% accuracy in field-level N₂O emissions would require soil sampling at 50-metre intervals—a density that eliminated any cost advantage over traditional inventory methods.

Real-Time Carbon Credit Verification

Multiple startups have attempted to use satellite data for continuous verification of nature-based carbon credits, promising to replace expensive periodic site audits. These efforts have consistently underperformed. The fundamental challenge is that carbon stock changes in forests and soils occur at rates that are below satellite detection thresholds over short time periods. A forest sequestering 5 tonnes of CO₂ per hectare annually changes its spectral signature by amounts indistinguishable from normal seasonal variation. The UK Woodland Carbon Code considered and rejected a satellite-first verification approach in 2024 after technical trials showed that reliable stock change detection required multi-year time series—too slow for the annual credit issuance cycles that markets demand.

Scope 3 Attribution Across Shared Infrastructure

The failed pilot's most intractable problem involved attribution. When a satellite observes emissions from a fertiliser manufacturing plant, multiple customers purchase that facility's output. Allocating the observed emissions to specific corporate Scope 3 inventories requires knowledge of commercial relationships and purchasing volumes that cannot be derived from satellite imagery. The consortium attempted to solve this through data-sharing agreements with suppliers, but encountered legal barriers (commercial confidentiality), technical barriers (incompatible data formats), and incentive barriers (suppliers had no motivation to facilitate scrutiny of their operations). By month fourteen, fewer than 30% of target facilities had signed data-sharing agreements, and the project team concluded that the attribution problem was "structurally unsolvable" without regulatory mandates requiring disclosure.

Key Players

Established Leaders

Planet Labs operates the world's largest commercial Earth observation constellation, with over 200 satellites providing daily global coverage at 3-5 metre resolution. Their monitoring services are widely used for commodity traceability and land use change detection, with UK clients including major retailers and financial institutions.

Airbus Defence and Space provides high-resolution optical and radar imagery through its Pléiades Neo and TerraSAR-X satellites. Their Starling platform, developed in partnership with Earthworm Foundation, is a leading tool for deforestation-free supply chain verification used by UK commodity traders.

Maxar Technologies offers very high-resolution imagery (30 cm) suitable for facility-level emissions monitoring and infrastructure assessment. Their analytics platforms support climate risk assessment for UK insurers and real estate investors.

European Space Agency (ESA) operates the Copernicus Sentinel constellation, providing free and open access to medium-resolution multispectral, radar, and atmospheric data. The Sentinel-5P satellite is particularly important for UK atmospheric monitoring, providing daily NO₂ and methane observations.

NASA Jet Propulsion Laboratory leads development of advanced atmospheric sensing instruments, including the EMIT hyperspectral imager and upcoming Carbon Mapper satellites. Their open data policies and algorithm transparency set benchmarks that commercial providers increasingly follow.

Emerging Startups

Satellite Vu (UK-based) develops thermal infrared satellites specifically designed for building energy performance monitoring. Their constellation, launching through 2025, will enable city-scale thermal efficiency assessments with 3.5-metre resolution.

GHGSat (Canada, with UK operations) operates the leading commercial methane detection constellation. Their sensors achieve 25-metre resolution for methane plume mapping, enabling identification of individual emission sources at industrial facilities.

Pixxel (India) is deploying hyperspectral satellites capable of distinguishing hundreds of spectral bands, enabling more precise identification of vegetation stress, mineral composition, and pollution sources than conventional multispectral systems.

Muon Space (US) is developing a constellation specifically for greenhouse gas monitoring, combining multiple sensing modalities to improve accuracy and reduce cloud interference challenges.

Isardsat (Spain, with UK partnerships) specialises in radar data processing for soil moisture and crop monitoring, providing analytics services that complement optical satellite observations for agricultural applications.

Key Investors & Funders

UK Space Agency administers grants and contracts for UK-based satellite companies and data integration projects. Their 2024-2027 National Space Strategy includes £180 million allocated to Earth observation applications, with climate monitoring as a priority area.

European Space Agency (ESA) Climate Change Initiative funds research and development of climate data records derived from satellite observations, with UK institutions as major participants despite Brexit.

Breakthrough Energy Ventures has invested in multiple satellite-based emissions monitoring startups, including significant funding for methane detection constellation development.

Lowercarbon Capital provides venture funding specifically for climate technology companies, with portfolio investments in satellite analytics platforms and carbon monitoring systems.

DEFRA Environmental Research Programme funds applied research on satellite data integration for UK agricultural and environmental policy, including trials of satellite-based subsidy verification systems.

Examples

Example 1: The Failed UK Agricultural Emissions Consortium (2023-2025)

A consortium comprising a major UK food retailer, two agricultural cooperatives, a satellite analytics company, and three academic institutions received £47 million in public and private funding to develop satellite-based Scope 3 emissions verification for agricultural supply chains. The pilot covered 2,400 farms across East Anglia and the Midlands, integrating Sentinel-2 optical imagery, Sentinel-1 radar data, and commercial high-resolution thermal observations.

Initial results were promising: crop health indices correlated with fertiliser application records at r²=0.67, and soil moisture estimates achieved ±15% accuracy compared to ground sensors. However, converting these observations to auditable emissions figures proved impossible within budget. The consortium discovered that fertiliser emissions depend critically on application timing, soil temperature, and microbial activity—variables that satellites cannot directly observe. Proxy models introduced uncertainty levels exceeding 50%, rendering the resulting figures unsuitable for carbon accounting under ISSB standards. The project was terminated in month eighteen, with published lessons learned emphasising the gap between technical capability and assurance requirements.

Example 2: Tesco's Satellite-Verified Deforestation Monitoring

Tesco implemented the Starling satellite monitoring platform across their soy and palm oil supply chains in 2022, achieving 98% coverage by 2024. The system combines Sentinel-2 optical imagery with commercial very-high-resolution data, processed through AI models trained to detect forest loss events.

The implementation succeeded where the agricultural consortium failed because it focused on detection rather than quantification. The platform triggers alerts when deforestation is detected within defined sourcing regions, prompting human investigation rather than generating automated carbon figures. This binary approach eliminated the uncertainty quantification challenges that plagued emissions estimation. Tesco reports that the system has enabled them to identify and address 47 deforestation incidents since implementation, with supplier engagement initiated within 72 hours of detection in all cases.

Example 3: Thames Water Methane Monitoring Programme

Thames Water partnered with GHGSat in 2023 to deploy satellite-based methane monitoring across their sewage treatment infrastructure. The programme covers 23 major treatment works, with monthly satellite observations supplemented by targeted ground-based surveys when elevated methane concentrations are detected.

The implementation achieved measurable results: a 23% reduction in unaccounted methane emissions within eighteen months, attributed to faster identification and repair of biogas leaks. Success factors included the binary nature of leak detection (present/absent rather than precise quantification), the relatively controlled environment of treatment facilities (simpler than distributed agricultural sources), and strong internal incentives (methane leaks represent both emissions liability and lost biogas revenue). However, the programme faced challenges integrating satellite data into existing asset management systems, requiring twelve months of IT development before operational deployment.

Action Checklist

• Audit your existing Scope 3 data quality before considering satellite integration—satellite data cannot improve fundamentally flawed ground-truth inputs
• Map your use case against the binary detection vs. continuous quantification distinction—satellite remote sensing excels at the former but struggles with the latter
• Assess supplier willingness to participate in data-sharing agreements before committing to attribution-dependent monitoring approaches
• Verify that proposed satellite data products meet the uncertainty quantification requirements of your target assurance standard (ISAE 3000, ISSA 5000, or equivalent)
• Budget for ground-truthing infrastructure—the failed UK pilot underestimated this cost by 340%
• Establish clear data ownership and liability allocation in contracts with satellite analytics providers before pilots commence
• Engage auditors early in the design phase to ensure monitoring systems produce evidence that meets their verification requirements
• Plan for algorithm versioning and change management—auditors require reproducibility across the full data collection period
• Set realistic timelines: the failed pilot's original eighteen-month schedule should have been thirty-six months based on comparable implementations
• Consider phased deployment starting with high-confidence applications (methane detection, deforestation monitoring) before attempting complex quantification use cases

FAQ

Q: What spatial resolution is required for corporate emissions monitoring?

A: Resolution requirements vary dramatically by application. Methane super-emitter detection at industrial facilities operates effectively at 25-50 metre resolution because emission plumes spread across observable areas. Deforestation monitoring works at 10-30 metre resolution, sufficient to detect most commercial clearing activities. However, field-level agricultural emissions estimation would theoretically require sub-metre resolution to capture the spatial variability in fertiliser application and soil conditions—resolutions that remain prohibitively expensive for continuous monitoring. The failed UK pilot attempted to compensate for 10-metre resolution data with modeling, but the introduced uncertainties exceeded acceptable thresholds. Sustainability leads should match resolution requirements to specific use cases rather than assuming higher resolution automatically produces better outcomes.

Q: How do satellite-derived emissions figures integrate with existing carbon accounting frameworks?

A: Integration remains the primary barrier to operational deployment. The GHG Protocol and ISSB standards were developed for activity-based and financial accounting data, not satellite observations. Satellite data enters corporate inventories either as a validation layer (comparing satellite observations to reported figures) or as an input to emission factor estimation (using satellite-derived activity data with standard emission factors). Neither approach is fully standardised. The Science Based Targets initiative published guidance in 2024 on using satellite data for Scope 3 verification, but auditors report inconsistent interpretation. Organisations deploying satellite monitoring should document their methodological choices extensively and engage auditors before finalising approaches.

Q: What are the typical costs for satellite-based emissions monitoring programmes?

A: Costs span several orders of magnitude depending on scope and approach. Subscription to existing monitoring platforms (Global Forest Watch, Starling) typically costs £50,000-£150,000 annually for large supply chain coverage. Custom monitoring programmes using commercial satellite tasking cost £200,000-£500,000 per year for regional coverage. Comprehensive programmes including ground-truthing infrastructure, algorithm development, and IT integration—like the failed UK pilot—can exceed £10 million over multi-year implementation periods. The failed pilot's budget was considered adequate at project approval but proved insufficient when ground-truthing requirements expanded. Sustainability leads should budget conservatively and build contingency for scope expansion.

Q: How should organisations evaluate satellite data providers?

A: Evaluation should prioritise auditability over technical capability. Many providers can demonstrate impressive technical performance in controlled conditions but cannot produce the documentation chain required for third-party assurance. Key evaluation criteria include: transparency of algorithm methodology (are processing algorithms published or proprietary?), uncertainty quantification approach (theoretical error propagation vs. empirical validation against ground truth), data format compatibility with existing systems, contractual provisions for data access and audit support, and track record of successful integration with carbon accounting frameworks. Request references from organisations that have successfully achieved third-party assurance on satellite-derived data rather than relying on capability demonstrations.

Q: What role do regulatory mandates play in satellite monitoring adoption?

A: Regulatory requirements are accelerating adoption but creating fragmented standards. The EU Deforestation Regulation mandates geolocation data for certain commodities, effectively requiring satellite monitoring for supply chain due diligence. The UK's Environmental Land Management scheme is progressively incorporating satellite-verified land use data into subsidy eligibility. However, carbon accounting regulations (SEC Climate Disclosure, CSRD, ISSB) do not specifically require satellite data, creating ambiguity about acceptable verification methodologies. The failed UK pilot was partly motivated by anticipated regulatory requirements that did not materialise as expected. Organisations should monitor regulatory developments but avoid over-investing in capabilities that may not align with eventual requirements.

Sources

• UK Space Agency (2024). "Earth Observation Applications Survey: Corporate Integration Success Factors." London: UK Space Agency Publications.

• International Auditing and Assurance Standards Board (2024). "ISSA 5000: General Requirements for Sustainability Assurance Engagements." New York: IFAC.

• European Commission (2023). "Regulation (EU) 2023/1115 on deforestation-free supply chains." Official Journal of the European Union.

• Transition Plan Taskforce (2024). "Disclosure Framework: Guidance on Climate Transition Plans." London: TPT Secretariat.

• GHGSat Inc. (2024). "Pulse: Global Methane Emissions Report 2024." Montreal: GHGSat Publications.

• Science Based Targets initiative (2024). "Guidance on Using Remote Sensing Data for Supply Chain Emissions Verification." CDP/UNGC/WRI/WWF.

• Financial Conduct Authority (2024). "Primary Market Bulletin 45: Climate-related disclosure requirements enforcement approach." London: FCA.

• DEFRA (2024). "Environmental Land Management Scheme: Satellite-Based Verification Technical Specifications." London: Crown Copyright.