Sustainability Atlas
Climate Tech & Data·14 min read··...

Interview: the builder's playbook for Satellite & remote sensing for climate — hard-earned lessons

A practitioner conversation: what surprised them, what failed, and what they'd do differently. Focus on implementation trade-offs, stakeholder incentives, and the hidden bottlenecks.

The satellite remote sensing market reached $48 billion in 2025 and is projected to grow to $123 billion by 2033, with environmental monitoring emerging as the fastest-growing segment at 13.3% CAGR. Yet behind these numbers lies a more complex reality: GHGSat's constellation of 16 methane-monitoring satellites detected over 20,000 emissions events in 2024 alone—equivalent to 534 million tonnes of CO₂ equivalent—while MethaneSAT went dark after just 15 months in orbit. We spoke with practitioners building the infrastructure for climate transparency to understand what's working, what's failing, and what investors in emerging markets need to know before deploying capital.

Why It Matters

Satellite-based climate monitoring has transitioned from scientific curiosity to regulatory necessity. The U.S. Inflation Reduction Act imposes methane fees of $900 per metric ton in 2024, rising to $1,500 by 2026. The EU's methane regulations mandate monitoring, reporting, and verification (MRV) for oil, gas, and coal operations—including imports. For emerging markets, these rules reshape export economics: compliance now determines market access.

The stakes are quantifiable. GHGSat's December 2024 publication in Science analysed 32,000 images from 3,114 oil, gas, and coal facilities worldwide, finding 8.3 million tonnes of methane escaping annually. Coal mines showed persistent leaks at 48% of monitored sites. The data names names: Turkmenistan, the United States, Russia, Mexico, and Kazakhstan lead in oil and gas emissions; China and Russia dominate coal.

For investors targeting emerging market infrastructure, satellite MRV offers dual value: enabling compliance with developed-market regulations and providing the transparency layer that carbon markets increasingly demand. Verra's November 2024 partnership with Pachama for digital forest carbon measurement signals that the world's largest carbon program recognises satellite-based verification as essential infrastructure. Gold Standard's February 2025 pilot with CarbonFarm in India—the first satellite-verified rice carbon project—demonstrates emerging market deployment is already underway.

Key Concepts

Understanding the MRV Stack

Satellite-based MRV combines multiple technologies into an integrated verification system. The foundation is optical and hyperspectral imaging: satellites like Sentinel-2 and Landsat-8/9 capture surface conditions at resolutions from 10 to 30 metres, enabling land-use classification and vegetation monitoring. Specialised instruments add greenhouse gas detection: NASA's OCO-2 and OCO-3 measure atmospheric CO₂ concentrations, while purpose-built methane satellites like GHGSat's constellation achieve 25-metre resolution capable of pinpointing individual equipment leaks.

Machine learning transforms raw imagery into actionable data. Pachama's platform, for example, processes satellite imagery to estimate forest biomass and carbon storage, then monitors changes over time to detect deforestation or quantify sequestration. The AI layer addresses a fundamental constraint: satellites generate petabytes of data, far exceeding human analytical capacity.

Ground-truth validation completes the system. IoT sensors, drone surveys, and strategic soil sampling calibrate satellite-derived estimates against physical measurements. This hybrid approach—combining wide-area satellite coverage with targeted ground verification—represents the current state of the art for credible carbon accounting.

Model Risk and Uncertainty Quantification

Deep learning models have advanced climate variable estimation dramatically, but uncertainty quantification remains immature. Recent research identifies three uncertainty sources: data acquisition errors (sensor calibration, atmospheric interference), model design and training uncertainty (architecture choices, training data limitations), and inference uncertainty (applying models to new contexts).

Most commercial platforms emphasise model architecture while underinvesting in data acquisition and inference uncertainties. For investors, this creates due diligence requirements: demand documentation of uncertainty bounds, ask how models are recalibrated across geographies, and verify that representation errors are addressed when comparing satellite-derived observations to ground-truth measurements.

Data Interoperability Challenges

Each satellite mission has unique characteristics—sensitivity thresholds, orbital patterns, performance evolution over time. Simple time-series concatenation across missions creates artificial trends and data discontinuities. The European Space Agency's Fundamental Climate Data Records (FCDRs) attempt to harmonise multi-mission data, but commercial platforms vary widely in their calibration rigour.

Fragmented monitoring systems compound the challenge. Many countries rely on self-reported emissions inventories with inconsistent methodologies. Integrating satellite-derived observations with national MRV systems requires not just technical compatibility but institutional alignment—a process measured in years, not months.

What's Working

GHGSat's Commercial Scale-Up

GHGSat has achieved what many thought impossible: a commercially viable, high-resolution methane monitoring constellation serving major energy producers. Operating 16 satellites by December 2025—the largest dedicated methane fleet globally—the company monitors over 4 million industrial facilities across 110 countries with daily revisit capability.

The business model works because it solves a specific problem for paying customers. Saudi Aramco, ExxonMobil, Chevron, Total, and Petrobras use GHGSat data to identify and repair leaks before they become regulatory liabilities. The company's tip-and-cue collaboration with ESA's Sentinel-5P and Sentinel-2 satellites demonstrates how commercial and government systems can create layered monitoring architectures.

The December 2024 Science publication provided third-party validation of GHGSat's methodology while generating the reputational capital that attracts additional commercial and government clients. This integration of research credibility with commercial deployment represents a replicable model for climate data companies.

Carbon Mapper's Tanager-1 Satellite

Launched in August 2024 and operational by summer 2025, Planet Labs' Tanager-1 satellite delivers hyperspectral imaging capable of detecting both methane and CO₂ plumes simultaneously. The October 2024 first detections demonstrated the system's capabilities: a 2.5-mile methane plume from a Karachi landfill (1,200 kg CH₄/hour), a 2-mile CO₂ plume from South Africa's Kendal power plant, and a 7,100 kg CH₄/hour leak in the Texas Permian Basin.

What distinguishes Carbon Mapper is its data access model. The coalition—including the nonprofit Carbon Mapper, NASA's Jet Propulsion Laboratory, Planet Labs, RMI, and Arizona State University—releases data publicly through a free portal with a 30-day delay. This transparency approach builds trust in satellite-derived emissions data while creating pressure on emitters to address identified leaks.

For emerging markets, the model offers valuable precedent: philanthropic funding from the High Tide Foundation, Bloomberg Philanthropies, and the Grantham Foundation de-risked development costs, enabling a public-good data layer that commercial and government users can build upon.

Verra-Pachama Digital MRV Partnership

In November 2024, Verra—the world's largest carbon program—partnered with Pachama to pilot digital forest carbon measurement. The initiative aims to "deliver standardized, high-integrity, digital carbon accounting at a fraction of the time and cost" of traditional ground-based verification.

This partnership validates satellite-based MRV for voluntary carbon markets. Credit issuance currently takes 1.5–2 years; digital MRV platforms like Rainbow (Riverse) claim 20x faster issuance and 3x lower costs. For investors, the Verra endorsement signals that satellite-verified credits will increasingly command market acceptance—and potentially premium pricing as integrity concerns reshape buyer preferences.

What's Not Working

MethaneSAT's Operational Failure

The Environmental Defense Fund's MethaneSAT launched in March 2024 with significant expectations: basin-scale methane detection with 2 parts-per-billion precision. After 15 months of operations that generated valuable regional data, contact was lost in June 2025. The failure highlights the fundamental fragility of space-based infrastructure—a single satellite represents a single point of failure.

For investors, the lesson is redundancy. GHGSat's 16-satellite constellation can absorb individual spacecraft failures; single-asset missions cannot. The MethaneSAT experience also illustrates why commercial ventures increasingly prefer constellation architectures despite higher initial capital requirements.

Uncertainty Quantification Gaps

Commercial MRV platforms emphasise detection capabilities while underreporting uncertainty bounds. Most deep learning applications for climate variable estimation lack rigorous treatment of aleatoric (data-inherent) and epistemic (model-inherent) uncertainties. This creates verification risks: carbon credits based on satellite-derived estimates may face challenges if underlying uncertainty is later quantified and found to exceed acceptable thresholds.

The problem is particularly acute for soil carbon, where satellite proxies estimate rather than directly measure subsurface conditions. ESA's SatMRV project uses Sentinel-2, Sentinel-1, and Landsat imagery with machine learning for soil organic carbon estimation, but the approach requires ground-truth calibration that many deployments skip.

Data Access and Continuity Risks

U.S. administration efforts to terminate NASA's Orbiting Carbon Observatory missions (OCO) threaten long-term climate data continuity. These missions cost approximately $15 million annually to maintain versus $750 million-plus in initial investment—an extreme cost-benefit asymmetry. The policy uncertainty creates risk for commercial platforms that depend on government data for calibration and validation.

Separately, data access barriers limit research utility. While NOAA and select NASA data are public, other agencies classify data—including sea ice coverage—behind security barriers. For emerging market applications, dependency on U.S. government data sources creates geopolitical exposure that warrants diversification toward European (Copernicus) and commercial alternatives.

Infrastructure and Debris Constraints

Climate change is cooling the upper atmosphere, causing it to contract and reducing its ability to clear orbital debris. MIT research projects this could reduce low-Earth orbit satellite capacity by 33–82% by 2100. While this timeline extends beyond typical investment horizons, the trend increases collision risk and insurance costs for all LEO operators.

More immediately, emerging market deployment faces infrastructure gaps. High-resolution climate data requires ground receiving stations, cloud processing capacity, and trained analysts—resources concentrated in developed markets. Effective emerging market deployment requires investment in local infrastructure alongside satellite capabilities.

Key Players

Established Leaders

GHGSat — Montreal-based operator of the world's largest methane-monitoring satellite constellation (16 satellites). Monitors 4+ million facilities across 110 countries. Clients include Saudi Aramco, ExxonMobil, Chevron, Total, and Petrobras.

Planet Labs — San Francisco-based Earth observation company operating 200+ satellites. Built and operates Tanager-1 for Carbon Mapper coalition. Won $20 million NASA contract for climate and environmental research data.

Maxar Technologies — Westminster, Colorado-based provider of high-resolution satellite imagery. WorldView Legion constellation expansion in February 2025 enhances Earth imaging capabilities for climate applications.

European Space Agency (ESA) — Operates Copernicus Sentinel constellation providing free, open-access Earth observation data. Sentinel-2, Sentinel-3, and Sentinel-5P support multiple climate monitoring applications.

Emerging Startups

Pachama — San Francisco-based forest carbon MRV platform using satellite imagery and machine learning. Acquired by Carbon Direct in November 2025 following Verra digital MRV partnership.

CarbonFarm — India-based agricultural MRV startup. Piloting first Gold Standard satellite-verified rice carbon project (February 2025) using machine learning for methane tracking.

Terradot — Enhanced rock weathering startup using satellite-based MRV for basalt weathering verification. Raised $54 million Series A in December 2024 from Acre Venture Partners, Microsoft, and Cisco.

Agreena — Copenhagen-based soil carbon platform connecting farmers with carbon markets using satellite imagery and strategic soil sampling.

Key Investors & Funders

Breakthrough Energy Ventures — Bill Gates-backed fund with $2+ billion under management. Portfolio includes Pachama (acquired 2025), Heirloom, and other carbon removal companies.

Bloomberg Philanthropies — Founding funder of Carbon Mapper coalition. Supports public-access emissions data infrastructure.

High Tide Foundation — Philanthropic anchor investor in Carbon Mapper, enabling public-good data layer for methane and CO₂ monitoring.

European Investment Bank — Provides financing for European climate tech companies. Supports ESA Copernicus commercialisation initiatives.

NASA Jet Propulsion Laboratory — Designed Tanager-1 imaging spectrometer. Partners with commercial operators on technology development and mission design.

Action Checklist

  1. Map regulatory exposure: Identify which portfolio companies face U.S. methane fees or EU MRV requirements. Quantify compliance costs under current and projected regulations ($900/ton in 2024 rising to $1,500/ton by 2026 for methane).

  2. Evaluate constellation resilience: For investments in satellite-based MRV, prefer multi-satellite constellations over single-asset missions. MethaneSAT's 15-month operational life illustrates single-point-of-failure risk.

  3. Demand uncertainty documentation: Require MRV platform providers to document uncertainty bounds, calibration methodologies, and ground-truth validation processes. Underquantified uncertainty creates downstream verification risk.

  4. Assess data source diversification: Evaluate dependency on U.S. government data sources. Policy uncertainty around NASA missions warrants building European (Copernicus) and commercial data redundancy.

  5. Prioritise public data layers: Investments that build on public-access data (Carbon Mapper, Copernicus) face lower proprietary data risk than those dependent on closed commercial feeds.

  6. Invest in local infrastructure: Emerging market satellite MRV deployment requires ground receiving stations, cloud processing, and analytical capacity. Budget for infrastructure alongside space segment investments.

  7. Track standards evolution: Monitor Verra, Gold Standard, and ICVCM developments on digital MRV acceptance. Credit verification standards are evolving rapidly toward satellite-based approaches.

  8. Build stakeholder coalitions: Effective satellite MRV requires alignment across operators, regulators, credit buyers, and emitters. Single-stakeholder commercial models face adoption friction that coalition approaches can overcome.

FAQ

Q: How accurate is satellite-based methane detection compared to ground-based measurement?

A: Commercial satellite systems achieve 25–30 metre resolution and can detect emissions above approximately 100 kg CH₄/hour. GHGSat's 25-metre resolution enables identification of individual equipment leaks, while Carbon Mapper's Tanager-1 achieves 30-metre resolution with 50-metre source localisation. Ground-based sensors remain more precise for continuous monitoring of known sources, but satellites provide the only practical method for wide-area surveillance across thousands of facilities. The hybrid approach—satellites for detection, ground sensors for quantification—represents current best practice. For carbon credit verification, satellite detection identifies where to focus ground-truth efforts rather than replacing them entirely.

Q: What is the investment case for satellite MRV in emerging markets specifically?

A: Emerging markets face three converging pressures: developed-market regulations increasingly apply to imports (EU methane rules cover imported oil and gas), carbon market integrity requirements are tightening globally, and national climate commitments under Paris Agreement require credible MRV systems. Satellite-based approaches offer cost advantages over deploying ground-based sensor networks across dispersed facilities. The February 2025 CarbonFarm pilot in India demonstrates that satellite-verified agricultural carbon projects can achieve Gold Standard certification, opening revenue streams for smallholder farmers. Investment opportunity exists in both the technology layer (local analytics capacity) and the application layer (platforms connecting satellite data to market access).

Q: How should investors evaluate model risk in satellite-derived climate data?

A: Model risk assessment requires examining three uncertainty sources: data acquisition (sensor calibration, atmospheric interference, orbital parameters), model architecture (training data representativeness, algorithm robustness), and inference (generalisability across geographies and conditions). Request documentation of uncertainty quantification methodologies. Ask whether platforms use satellite simulators to compare observations against models. Verify that ground-truth validation occurs in the specific geographies and conditions relevant to the investment thesis. Be particularly cautious with soil carbon estimates, which rely on proxy relationships between surface observables and subsurface conditions—these require more extensive ground-truth calibration than above-ground biomass estimates.

Q: What regulatory developments should investors monitor for satellite climate data?

A: Key regulatory developments include: U.S. EPA methane fee implementation and potential Supreme Court challenges; EU methane regulation phase-in timelines and third-country compliance requirements; Verra and Gold Standard acceptance of digital MRV methodologies for credit verification; ICVCM Core Carbon Principles adoption by major credit buyers; and national MRV system development in major emerging market emitters. The trajectory is clearly toward satellite-based verification, but the pace varies by jurisdiction and application. Agricultural carbon (rice, soil) is moving faster than industrial emissions in some markets, while methane detection has achieved broader acceptance than CO₂ quantification.

Q: What is the realistic timeline for satellite MRV to become standard for carbon credit verification?

A: For forest carbon, satellite-based MRV is already standard practice—Verra's November 2024 Pachama partnership formalised what leading project developers had adopted informally. For agricultural carbon, the February 2025 Gold Standard pilot indicates 12–18 months to methodology approval and broader deployment. For industrial emissions and methane, regulatory acceptance is further advanced: EU methane regulations effectively mandate satellite-compatible monitoring for covered facilities. The constraint is less technological than institutional—aligning verification standards, building evaluator capacity, and establishing precedent through pilot projects. Investors should expect 2–3 year timelines from pilot to scale in most emerging market applications, with faster adoption in jurisdictions that have already adopted developed-market regulatory frameworks.

Sources

The satellite remote sensing market for climate applications represents a maturing sector where commercial viability has been proven and regulatory tailwinds are accelerating. For investors focused on emerging markets, the opportunity lies not in replicating developed-market infrastructure but in building the local capacity—analytical platforms, ground-truth networks, institutional relationships—that transforms global satellite coverage into locally actionable climate intelligence. The technology stack exists; the question is who builds the last mile.

Topics

satelliteremoteverificationmodel riskadditionalityMRVdata interoperability

Related Articles