If you built a pilot inventory or a FLAG baseline on the September 2022 draft Land Sector and Removals Guidance, part of that work survives the final text and part of it needs redoing.
2026 is the year the draft became a rulebook. The GHG Protocol published the Land Sector and Removals Standard in January, SBTi revised its FLAG guidance in March, and version 1.1 of the LSR Standard arrived in June alongside a 518-page implementation guidance. Across those documents, 32 requirements were set out, and most of them changed since the draft. The Standard takes effect on 1 January 2027, the last day the 2022 draft is worth anything for target setting, and is due for review by 2030.
This is the second blog in a two-part series, and it assumes you already know what the LSR Standard is, how land emissions split into four subcategories, and how the traceability ladder works. Readers coming to this fresh should start with Part 1, The Land Sector and Removals Standard: a new rulebook for land emissions.
The draft you piloted covered forestry. Version 1 of the Standard covers agriculture and CO2 removal technologies, and Part 1 sets out the background to that decision.
Much of what forest-exposed companies piloted still holds. Deforestation remains land use change. What version 1 leaves for later is carbon accounting on forest land itself.
The draft allowed piloting teams to either classify all their land as managed, or build and consistently apply their own managed and unmanaged split. The final text settles the question for productive agricultural land: every change in land carbon stock there counts as anthropogenic. Forest land and non-productive, non-forest land sit outside version 1’s rules, which is why the choice disappeared rather than being resolved for everyone.
The second provision sits in the Standard but cannot be followed yet. It lets companies with a land management unit boundary include nearby non-productive land, hedgerows, riparian buffers, agroforestry margins and set-aside, where it shares a management plan with the productive land and connects to it directly (Requirement 7). It also requires size and distance thresholds, and the Standard has not set them. If your pilot claimed carbon on any of those, park that work.
The draft applied to companies with land sector activities. The final applies to companies with significant land sector activities, and a decision that yours fall below that line is a disclosure you justify. The GHG Protocol declines to set a number, and its own footnote points at SBTi, which requires a FLAG target once FLAG emissions reach 20 percent of Scopes 1, 2 and 3.
In the draft, physical traceability was mostly a gate on removals. It now determines your Scope 3 spatial boundary for emissions, removals and every other land metric, and you hold that single boundary across every accounting category for a given volume of a given product. Your traceability position governs the whole Scope 3 land inventory, where it used to govern one corner of it. Part 1 sets out the ladder those levels sit on.
Mass balance is the change most likely to alter what a food or agriculture company can claim. The draft grouped it with book and claim and ruled both out. The final admits mass balance as physical traceability where four conditions hold (Requirement 8):
Book and claim remains outside. Since bulk agriculture moves through silos, mills and crushers, this opens sourcing-region accounting to companies who piloted at jurisdiction level and stopped there.
Impact traceability arrived as a named concept. Emissions and removals you can trace to a project or intervention, rather than to physical product, are reported separately from the physical GHG inventory. Supply shed programmes finally have somewhere defined to sit.
Part 1 explains the 20-year assessment period and the linear weighting that runs through it. What changed is the option beside it. The draft let you spread land use change emissions using linear or equal discounting, and the final requires linear amortisation, permitting alternatives only in narrow circumstances you disclose and justify.
The chart below shows how wide the gap between the two methods is at the recent end. Linear amortisation assigns 9.75 percent of a conversion’s emissions to the year of conversion and steps down half a point each year to 0.25 percent in year 20. Equal discounting assigned a flat 5 percent throughout. Anyone who piloted equal discounting with recent conversion in the supply shed should expect the figure to rise by close to double for those years.
The draft asked you to disclose and justify whichever of two allocation methods you used, shared responsibility or product expansion. The final settles on product expansion, with shared responsibility surviving as a justified exception. Product expansion assigns conversion to the crops that expanded onto the converted land, so the number concentrates on expanding commodities where it used to spread across everything grown in the region.
Direct land use change, dLUC, is conversion traced to identified land rather than inferred from regional statistics. In the draft, spatial methods for estimating conversion inside a jurisdiction appeared in a box, flagged as something for pilot testers to consider. In the final they are a full calculation approach, sitting between statistical land use change and land management unit level dLUC. With sourcing region traceability, a crop extent map and jurisdictional land cover maps, you can locate conversion spatially using regional data. For example, a company sourcing cocoa can calculate land use change across every cocoa-growing area in its sourcing country, rather than spreading that country's whole land use change total across everything grown there. The Standard also asks you to take the most accurate calculation approach your data and traceability allow.
The same allocation method applies across emissions, removals and other metrics within a given spatial boundary (Requirement 9). The Scope 3 allocation decision tree grew from three steps to five: before physical allocation, a company now checks whether a law mandates a specific approach, then whether a sector recommendation does. For agriculture, that second check is a ready-made hierarchy, crop rotations, cover crops, intercropping and processing facilities each carry their own allocation rule.
Three subcategories now count:
Companies raising livestock or buying animal products account for conversion on grazing land and on the cropland growing feed, across the animal’s life cycle.
The draft treated reservoir emissions as land management non-CO2 emissions. The final treats them as land use change, and pulls in three groups:
A change in the land use change calculation approach, or in the data and methods inside it, requires you to recalculate your base year (Requirement 10). Every item above is a change of approach or data for someone. Doing that work once, in 2026, costs less than doing it again in 2028.
The draft asked for at least one land tracking metric, chosen from indirect land use change emissions, carbon opportunity cost, or land occupation. The final settles on agricultural land occupation in hectares, in Scope 1 and Scope 3, for each Scope 3 category, which Part 1 covers as a metric in its own right. Carbon opportunity cost of land use survives as a recommendation, as does breaking hectares down by land use category.
The gain is comparability, since every company reporting under the Standard now reports the same unit. The cost falls on pilots that chose one of the other two metrics, which need hectares as well.
In the draft, any action that could significantly increase emissions or reduce removals beyond your Scope 1, 2 and 3 boundary had to be estimated and reported separately, with the company deciding what qualified and quantifying it through intervention accounting methods.
The final replaces that with a closed test, and narrows the subject matter to one thing: food and feed production displaced by your own actions. Leakage applies when one of three activities occurs in your operations or value chain and you report a climate benefit as a result (Requirement 13):
A climate benefit means reduced emissions, increased removals, or reduced emissions measured against a non-biogenic alternative. Changes made to improve the long-term sustainability of food production are carved out. Where the test is met, the method is fixed at carbon opportunity cost, which Part 1 defines.
Two consequences follow for anyone who piloted the draft. Work you did on impacts beyond your boundary that had nothing to do with displaced food falls away. And any leakage you did quantify came out of a different method, so those figures need rebuilding on carbon opportunity cost.
The draft left open how much traceability a removal claim demands. The published Standard answers it: land management removals can be reported with physical traceability to a sourcing region, a land management unit or a harvested area. Sourcing-region reporters meet a defined set of safeguards, covering the boundary, attributable productive lands, allocation and sampling variability (Requirement 20). Combined with mass balance, that brings removals reporting within reach of companies buying commodity crops through first collection points.
Data quality eased in two places. The draft required the net carbon stock increase to be statistically significant before you could report it. That became a recommendation to use sampling approaches producing statistically significant estimates. The requirement to hold measurements representative of the base year likewise became a recommendation. Pilots that stalled on sampling density have a lower bar to clear.
Part 1 covers the allocation rules that stop two companies claiming the same tonne, and the final Standard names who those rules apply against: companies at the same or a similar tier of your value chain.
Product carbon storage settled the draft's second open question. It sits outside the physical GHG inventory as its own accounting category, with the product carbon pool narrowed to the use phase including recycling and reuse. Carbon in landfill and in end-of-life waste moved into a separate waste carbon storage category with conditions of its own.
Part 1 sets out how disaggregated reporting works and what a headline net figure has to contain. The draft-to-final change is a piece of flexibility alongside it: where an accounting category is a small share of your inventory and disaggregated data is unavailable, you may report it aggregated by Scope and Scope 3 category, with disclosure and justification (Requirement 31).
Land emissions also lost their Scope 2 line. The draft placed some land emissions, land metrics and removals in Scope 2. The final routes the land impacts of purchased energy into Scope 3.
SBTi published FLAG guidance version 1.2 on 19 March 2026, an urgent revision to two criteria: C1, which sets who must hold a FLAG target, and C4, the no-deforestation commitment. The edits reflect the launch of the Standard. Three points matter for anyone holding or building a FLAG target:
Most disclosure regimes point at the GHG Protocol, so conformance with the Standard flows into whatever framework you report under.
Before rebuilding anything, find out which of these changes actually touch your commodities. Our complimentary 1-on-1 LSR workshop is a practical gap analysis of your current FLAG reporting against the LSR requirements, covering your data infrastructure, supplier engagement approach and assurance timeline.
Linear amortisation and product expansion are mandatory now, and both push the number the same way where conversion is recent. Terrascope calculates land use change over the required 20-year lookback period, broken down by commodity and scope, with an audit-ready trail behind every figure.
What you can prove now sets the spatial boundary for your whole Scope 3 land inventory. Terrascope’s supplier engagement module runs structured data campaigns with templates tailored by supplier type, and traces every submission through to your footprint.
Since these changes move land emissions, the 20 percent test is worth recomputing rather than carrying over. Terrascope reviews your industry, operational and value chain model to determine FLAG exposure and the right reporting pathway. Princes Group, an international food and drink conglomerate, secured SBTi-validated FLAG targets in 3.5 months.