When an ESG analyst asks us whether our deforestation data can feed into their Scope 3 emissions reporting, the honest answer is: it can help, but the frameworks you're reporting against were not built with satellite monitoring in mind, and that gap is where a lot of the current disclosure quality falls apart.
Deforestation is one of the largest contributors to land-use change emissions globally, and for companies with commodity supply chains — soy, palm oil, beef, timber, cocoa — the clearing events happening at the origin end of those chains are, in principle, Scope 3 Category 11 (use of sold products) or Category 1 (purchased goods and services) emissions attributable to the company. In practice, most corporate Scope 3 inventories treat deforestation-linked emissions as either zero, estimated from commodity-level emissions factors, or not reported at all. The frameworks aren't forcing better practice. Here's why.
What TCFD Actually Requires
The Task Force on Climate-related Financial Disclosures (TCFD) recommends disclosure of climate-related risks and opportunities, including physical risks from land-use change and transition risks from tightening forest-related regulation. TCFD does not specify a deforestation monitoring methodology. It does not require a spatial resolution. It does not set a cadence for monitoring supply chain parcels. What it requires is that material risks be disclosed and that companies describe how they identify, assess, and manage those risks.
A company can satisfy TCFD's deforestation-related disclosure with a qualitative statement about their supplier engagement program and a reference to an industry-level forest risk materiality assessment. That's compliant. It's not the same as knowing whether any parcel in your soy supply chain was cleared after your deforestation commitment date.
The gap between "TCFD-compliant" and "actually knowing your deforestation exposure" is enormous.
GRI 304 and the Biodiversity Disclosure Standard
GRI Standard 304 (Biodiversity) requires disclosure of significant impacts on biodiversity, including through supply chain land use. GRI 304-2 asks for "significant direct and indirect impacts on biodiversity" — which in principle should include deforestation-linked land clearing in agricultural supply chains. In practice, most GRI 304 disclosures describe impacts at the level of commodity category ("we source palm oil from high-risk regions") rather than at the level of individual parcels or clearing events.
GRI released an updated biodiversity standard in 2024 that strengthens the language around supply chain disclosure, but does not mandate parcel-level satellite monitoring. The standard asks for location-specific impacts but leaves the identification methodology to the company. This creates a situation where two companies can both report against GRI 304, one using country-level supply chain risk scores and one using parcel-level satellite time series, and both disclosures look comparable in a CSR report.
SBTi FLAG Guidance: Closer, But Still Imprecise
The Science Based Targets initiative's Forest, Land and Agriculture (FLAG) guidance, finalized in 2022, is the framework most directly relevant to deforestation-linked emissions quantification. FLAG requires companies in agricultural supply chains to set targets covering land-use change emissions as part of their science-based targets, and to report progress annually against a base year.
This is genuinely more demanding than TCFD or GRI. FLAG methodology requires companies to quantify their land-use change emissions using recognized accounting methods — typically the GHG Protocol Land Sector and Removals Guidance or country-level IPCC carbon stock factors applied to supply chain volumes. Companies must disaggregate FLAG emissions from their Scope 1/2/3 fossil fuel emissions.
Where FLAG still falls short, from a satellite evidence perspective: the quantification methodology allows companies to use national or subnational deforestation rate statistics combined with commodity production volumes to estimate emissions, rather than tracing actual parcel-level deforestation attributable to their specific supply chain. This is understandable — parcel-level traceability is genuinely difficult for some commodities — but it means the emissions figure in a FLAG-compliant report may bear little relationship to what's actually happening on specific parcels supplying that company.
We're not saying FLAG is wrong for allowing statistical estimation. The accounting has to work at scale, and not every company has the supply chain traceability infrastructure to go parcel by parcel. But as satellite monitoring becomes cheaper and parcel-level data becomes more accessible, the gap between "what FLAG allows" and "what's technically knowable" is going to become harder to defend to investors and regulators.
The Monitoring Cadence Problem
Even when companies do invest in forest monitoring tools, the dominant practice is annual or semi-annual review. This is driven by reporting cycles — ESG disclosures are typically annual, sustainability committees review risks quarterly at best, and the systems that feed those committees are often updated on the same cadence.
The problem is that deforestation doesn't happen on annual cycles. A clearing event in the Brazilian dry season (June-September) is detectable within days by Sentinel-2 or Sentinel-1. If a company's monitoring system runs annually, a clearing event that occurred in July on a supplier parcel won't appear in the data until the next annual update — by which point the biomass is gone, the land may be under soy, and there is no remediation path.
Annual monitoring is appropriate for annual reporting. It is not appropriate for supply chain risk management if the goal is to intervene before deforestation happens, or to engage with suppliers before clearing events occur. The disconnect between reporting cadence and monitoring cadence is where most supply chain forest commitments fail in practice.
What Parcel-Level Data Actually Enables
Consider a mid-size food company sourcing soy from multiple trading desks with upstream exposure to Brazilian origin. Their FLAG emissions estimate is built from USDA production statistics and Brazil's annual MapBiomas deforestation figures allocated proportionally to their sourcing volumes. That's the current state of art for most companies outside the largest global commodity traders.
With parcel-level satellite monitoring, that same company could instead maintain a polygon dataset of the specific farms in their supply chain (to the extent traceability permits), run change detection on those polygons against a 2020 forest baseline, and directly quantify what area has been cleared in scope of their supply chain since their commitment date. The resulting emissions figure is a direct count, not a statistical allocation. It's also actionable: when a polygon shows clearing activity, the company can immediately flag that supplier relationship for review rather than discovering the issue a year later at the next annual audit.
The legal standard is moving this direction. EUDR's geolocation and no-deforestation-after-2020 requirements (discussed separately in our EUDR post) effectively mandate a version of parcel-level monitoring for operators placing certain commodities on the EU market. Companies that build this infrastructure for EUDR compliance will have the foundation for significantly higher-quality FLAG reporting as a byproduct.
Closing the Gap
The frameworks are not the ceiling — they're the floor. TCFD, GRI 304, and SBTi FLAG define minimum disclosure expectations, not best practice. Companies whose Scope 3 strategies are genuinely trying to reduce deforestation-linked emissions rather than simply disclose them are already doing things the frameworks don't require: parcel-level traceability, near-real-time monitoring, direct supplier engagement triggered by satellite alerts.
For ESG practitioners, the honest question to ask about your current deforestation disclosure is: would this data change if deforestation on your supply chain parcels tripled next year? If your reporting methodology insulates the disclosed figure from actual on-the-ground clearing events, then you're reporting the structure of your estimation method, not your actual deforestation exposure. Those are very different things, and the difference matters increasingly to sophisticated investors and regulators who understand the distinction.