Cropland Management

1.0 Introduction

This Module (Independent components of Isometric Certified Protocols which are transferable between and applicable to different Protocols.) provides the requirements and procedures for the quantification of net carbon dioxide equivalent (CO₂e (The amount of CO₂ emissions that would cause the same integrated radiative forcing or temperature change, over a given time horizon, as an emitted amount of GHG or a mixture of GHGs. One common metric of CO₂e is the 100-year Global Warming Potential.)) removal (The term used to represent the CO₂ taken out of the atmosphere as a result of a CDR process.) removal from the atmosphere via improvements to soil organic carbon (SOC (Soil Organic Carbon)) stocks in cropland systems. This Module sits under the Improved Soil Management Protocol (A document that describes how to quantitatively assess the net amount of CO₂ removed by a process. To Isometric, a Protocol is specific to a Project Proponent's process and comprised of Modules representing the Carbon Fluxes involved in the CDR process. A Protocol measures the full carbon impact of a process against the Baseline of it not occurring.) and must be read in conjunction with it.

Cropland soils represent one of the largest opportunities for terrestrial carbon sequestration¹. Globally, agricultural soils have lost an estimated 133 Pg of organic carbon since the onset of widespread cultivation, and a substantial fraction of this historical loss is recoverable through improved management practices². Recent estimates suggest that improved cropland management alone could sequester between 0.28 and 1.85 Pg CO₂ yr⁻¹ globally, placing cropland SOC enhancement among the most scalable nature-based climate solutions available³.

This Module is process-agnostic with respect to the specific management practices employed. Eligible project (An activity or process or group of activities or processes that alter the condition of a Baseline and leads to Removals or Reductions.) activities encompass any practice or combination of practices that results in a demonstrable net increase in SOC stocks against a counterfactual (An assessment of what would have happened in the absence of a particular intervention – i.e., assuming the Baseline scenario.)baseline, (A set of data describing pre-intervention or control conditions to be used as a reference scenario for comparison.) provided all eligibility and safeguarding requirements set out in this Module and the parent Protocol are met. This approach reflects the wide diversity of cropping systems, soil types, and climatic conditions encountered globally, and recognizes that effective SOC enhancement strategies will vary substantially by region and context. Using common terminology for SOC-enhancing practices, examples of such eligible activities include, but are not limited to:

• Cover cropping: Growing crops during fallow periods to increase organic matter inputs to the soil 
• Reduced or no-till: Minimizing soil disturbance to preserve existing SOC stocks and promote accumulation 
• Compost and organic matter application: Adding external organic inputs to increase soil carbon content 
• Biostimulant and biological inoculant application: Use of microbially-based or biochemical products that stimulate plant growth and root exudation, promoting SOC formation pathways

In addition to climate mitigation, improvements to SOC in cropland systems can provide environmental and social co-benefits, including enhanced soil water retention and drought resilience, improved crop yields and long-term agricultural productivity, reduced dependency on synthetic fertilizers, mitigation of erosion and nutrient runoff into waterways, and support for soil biodiversity (The diversity of life across taxonomic and spatial scales. Biodiversity can be measured within species (i.e. genetic diversity and variations in allele frequencies across populations), between species (i.e. the total number and abundance of species within and across defined regions), within ecosystems (i.e. the variation in functional diversity, such as guilds, life-history traits, and food-webs), and between ecosystems (variation in the services of abiotic and biotic communities across large, landscape-level scales) that support ecoregions and biomes.)^4,5,6,7. Carbon finance (Resources provided to projects that are generating, or are expected to generate, greenhouse gas (GHG) Emission Reductions or Removals.) presents a meaningful opportunity to overcome the adoption barriers, including implementation costs, agronomic support, and the absence of financial incentives, that have historically constrained farmer uptake of SOC-enhancing practices.

Project Proponents (The organization that develops and/or has overall legal ownership or control of a Removal or Reduction Project.) must meet all the requirements set out in the Improved Soil Management Protocol and relevant Modules, as well as the requirements set out in this Module. Where this Module contains requirements that duplicate or conflict with those in other Modules, this Module takes precedence.

Throughout this Module, the use of “must” indicates a requirement, “should” indicates a recommendation, and “may” indicates an option or permission.

2.0 Sources and Reference Standards & Methodologies

This Module relies on and is intended to be compliant with the following Standards and Protocols:

• The Isometric Standard
• Improved Soil Management Protocol v1.0, Isometric
• ISO (A worldwide federation (NGO) of national standards bodies from more than 160 countries, one from each member country.)14064-2: 2019 Greenhouse Gases Part 2: Specification with guidance at the project level for quantification, monitoring, and reporting of greenhouse gas emission reductions (Lowering future GHG releases from a specific entity.) or removal enhancements

Additional reference standards that inform the requirements of this Module include:

ISO 14064-3: 2019 Greenhouse Gases Part 3: Specification with guidance for the verification (A process for evaluating and confirming the net Removals and Reductions for a Project, using data and information collected from the Project and assessing conformity with the criteria set forth in the Isometric Standard and the Protocol by which it is governed. Verification must be completed by an Isometric approved third-party (VVB).) and validation (A systematic and independent process for evaluating the reasonableness of the assumptions, limitations and methods that support a Project and assessing whether the Project conforms to the criteria set forth in the Isometric Standard and the Protocol by which the Project is governed. Validation must be completed by an Isometric approved third-party (VVB).) of greenhouse gas statements

• ISO 14040: 2006 Environmental Management, Lifecycle Assessment, Principles & Framework
• ISO 14044: 2006 Environmental Management, Lifecycle Assessment, Requirements & Guidelines

Additional principles that were considered in the development of this Module include:

• The Core Carbon Principles of The Integrity Council for the Voluntary Carbon Market, v1.1, ICVCM, 2024
• Criteria for High-Quality Carbon Dioxide Removal, Carbon Direct & Microsoft, 2025

3.0 Future Versions

This Module was developed based on the current state of the art and publicly available science regarding cropland soil organic carbon dynamics and land management interventions. This Module aims to be scientifically stringent and robust. We recognize that some requirements may exceed the status quo in the voluntary carbon market and that there are numerous opportunities to improve the rigor of this Module as new approaches and techniques emerge.

Additionally, this Module will be reviewed when there is an update to published scientific literature, government policies, or legal requirements which would affect net CO₂e removal quantification or the monitoring guidelines outlined in this Module, or at a minimum of every 2 years.

4.0 Applicability

In addition to the requirements outlined in the Improved Soil Management Protocol, Projects are subject to the following applicability requirements, which must be demonstrated in The Project Design Document (The document, written by a Project Proponent, which records key characteristics of a Project and which forms the basis for Project Validation and evaluation in accordance with the relevant Certified Protocol. (Also known as “PDD”).).

4.1 Eligible Land

In the context of this Module, cropland is defined as land used for the cultivation of annual or perennial crops, including arable land under crop rotation, fallow land within an active rotation cycle, and land under permanent crops such as orchards, vineyards, or plantations of non-timber commodity (A product that has been cultivated, raised or harvested primarily for food, shelter, or natural fiber.) crops. The unit of enrollment (see Section 4.2 of the Improved Soil Management Protocol) is a field.

4.1.1 Land Use Exclusions

4.1.1.1 Definitions for Native Ecosystems

• Native forest - A forest dominated by tree species indigenous to the region and that has been established and regenerated through natural processes. Generally with high biodiversity and intact ecological processes (nutrient cycling, natural disturbance regimes).
• Native grassland - A grassland dominated by grass and herbaceous species indigenous to the region and that has formed and persisted through natural processes, distinct from sown, "improved," or cultivated pasture made up of introduced forage species, and distinct from grassland created by clearing forest.
• Native wetland - A wetland whose hydrology, soils, and biological communities have formed and persisted through natural processes, and whose vegetation is dominated by water-adapted species indigenous to the region.

4.2 Support for Biodiversity and Community Livelihoods

Following the requirements under the Protocol, Project Proponents should implement practices that provide ecological benefits in addition to increasing carbon sequestration. SOC-enhancing practices including cover cropping, reduced tillage, and organic amendment application are encouraged where they deliver co-benefits for soil health, water quality, and habitat connectivity alongside their primary carbon benefit.

4.3 Other Requirements

Where cultivation is required as part of project establishment, soil inversion should be limited to 25 cm depth. This recommendation applies to areas where new management practices are being established as part of The Project and does not apply to any continuing agricultural activity that was occurring at the same depth prior to project initiation.

The Project's primary commitment is to maintain cumulative SOC stocks at or above the level corresponding to credits (A publicly visible uniquely identifiable Credit Certificate Issued by a Registry that gives the owner of the Credit the right to account for one net metric tonne of Verified CO₂e Removal or Reduction. In the case of this Standard, the net tonne of CO₂e Removal or Reduction comes from a Project Validated against a Certified Protocol.) or certificates previously issued, quantified through the monitoring framework set out in Section 9. Because the Module is practice-agnostic, Project Proponents may adapt the specific practices used over time, provided the outcome-based commitment continues to hold.

The Project must provide a minimum of 40 years of SOC storage (Describes the addition of carbon dioxide removed from the atmosphere to a reservoir, which serves as its ultimate destination. This is also referred to as “sequestration”.) in The project area, as defined by the length of The Project Commitment Period (see Section 5.1)

• The primary commitment is to the maintenance of cumulative SOC stocks at or above the level corresponding to Credits previously issued, as quantified through the monitoring framework set out in Section 9.
• Project Proponents may adapt the specific land management practices implemented during The Project Commitment Period, provided that any such adaptation: (i) does not breach any other applicability requirement of this Module or the parent Protocol; (ii) is documented in advance in the Monitoring Report and approved at the next verification; and (iii) is accompanied by evidence, through the Section 9 framework, that the 30th-percentile estimate of cumulative SOC stocks (calculated in accordance with Section 9.1.1) remains at or above the level corresponding to Credits previously issued. Reversion to baseline land management practices, or any practice change for which the 30th-percentile cumulative SOC stock estimate falls below the level corresponding to Credits previously issued, will be treated as a reversal (The escape of CO₂ to the atmosphere after it has been stored, and after a Certificate has been Issued. A Reversal is classified as avoidable if a Project Proponent has influence or control over it and it likely could have been averted through application of reasonable risk mitigation measures. Any other Reversals will be classified as unavoidable.) in accordance with the Isometric Standard.
• Independently of the outcome-based test set out above, Project Proponents must notify Isometric of any material change to the land management practices implemented within The Project Boundary as soon as practicable, and in any event no later than the next Monitoring Report.
  • A change is material where it alters the practices identified in The Project Design Document as responsible for generating the SOC increase, or where it could reasonably be expected to affect SOC stocks within The Project Area. Failure to notify a material practice change will be treated as a non-conformance and may result in the suspension of crediting until the change has been assessed and verified.

• Where The Project involves smallholder farmers or customary land users, land tenure arrangements must be consistent with the farmer and land user rights requirements of Section 6.6 of the Improved Soil Management Protocol.

• Planting trees and comparable land-use transformations are scoped under the Isometric Agroforestry and Reforestation Protocols.

5.0 Project Timelines

5.1 Project Commitment Period

The Project Commitment Period encompasses the Crediting Period and any Ongoing Monitoring Period commitments following the end of Crediting.

For grouped projects where new areas are added to The Project over time, The Project Timeline may be staggered across The Project areas to reflect different initiation times of Project activities.

5.2 Crediting Period

• Duration. The total Crediting Period (including renewals) can be no longer than the Project Commitment Period set at project initiation.
• Issuance (Certificates are issued to the Certificate Account of a Project Proponent with whom Isometric has a Validated Protocol after an Order for Verification and Certificate Issuance services from a Buyer and once a Verified Removal or Reduction has taken place.). Credit issuances occur throughout the Crediting Period. Credits are issued upon Verification of a Reporting Period. Credits issued at each verification event represent the cumulative net CO₂e removal from project initiation to the end of the current Reporting Period, less all credits previously issued under The Project.
• Renewal. The Crediting Period may be renewed up to the duration of The Project Commitment Period.
• Considerations for Grouped Projects. For grouped projects composed of multiple discrete areas, the individual Crediting Periods may be staggered across individual sites to reflect the different timings of project activity initiation within the sites.

5.3 Reporting Period

• Duration. The minimum duration of a Reporting Period is one year. The maximum duration of a Reporting Period is five years. Project Proponents may request an alternative length of the Reporting Period provided they submit suitable justification for the deviation (e.g., evidence of carbon stock impacts of intervention anticipated to take longer during establishment; matching relevant cultivation cycles).
• Verification. Verification of project activities by a third-party Verification and Validation Body (VVB (Third-party auditing organizations that are experts in their sector and used to determine if a project conforms to the rules, regulations, and standards set out by a governing body. A VVB must be approved by Isometric prior to conducting validation and verification.)) is conducted for each Reporting Period (see Section 7.2 of the Improved Soil Management Protocol).
• Last Reporting Period. Project Proponents must indicate the last Reporting Period to be submitted for Verification. Failure to initiate a Verification within 5 years of the previous Reporting Period or request an extension will conclude the Crediting Period.

5.4 Ongoing Monitoring Period

• Duration. If the Crediting Period is less than The Project Commitment Period, the Ongoing Monitoring Period must encompass the length of time from the end of the Crediting Period to the end of the Project Commitment Period.
• Monitoring. Monitoring for Reversals is the responsibility of The Project Proponent and must follow the same requirements for the measure re-measure approach used for quantification of [math: CO_2e_{stored}] (Section 9.1.2). If monitoring is not possible (e.g., do not have access to land), the area is considered to experience a full Reversal.

5.5 Durability

The Project Commitment Period must be set at the time of PDD submission. If the initial Crediting Period is less than The Project Commitment Period, The Crediting Period may be extended up to the duration of The Project Commitment Period. If the initial Crediting Period is not extended, the remainder of The Project Commitment Period must consist of an Ongoing Monitoring Period. For grouped projects, The Project Commitment Period must be the same duration for all areas, but Crediting Period and Ongoing Monitoring Period duration may vary across individual areas to reflect differences in land tenure.

5.6 Post-Project Commitment Period

• Definition. The indefinite period of time after The Project Commitment Period has ended.

5.7 Example Project Timelines

Project A is a smallholder cropland management project. It sets a Project Commitment Period of 40 years, with an initial Crediting Period of 20 years for all enrolled properties. At the end of the initial Crediting Period, most of the properties renew their enrollment for another 20 years and continue to accumulate Credits for the remainder of The Project Commitment Period. The remaining properties which did not renew their enrollment enter an Ongoing Monitoring Period to monitor for Reversals for the remaining 20 years of The Project Commitment Period. The Project Proponent is responsible for quantification and monitoring through all of The Project Area for the full duration of The Project Commitment Period, and the reported activities are verified by a Validation and Verification Body (VVB). All Credits from Project A have a durability of 20 years, equivalent to half of the 40 year Project Commitment Period.

Project B is a cropland project which sets a Project Commitment Period of 60 years, which fully consists of the Crediting Period. All Credits from Project B have a durability of 30 years, equivalent to half the 60 year Project Commitment Period.

Project C is a grouped cropland management project which sets a Project Commitment Period of 40 years, entirely composed of crediting periods. It is composed of two cohorts of enrolled properties - Group C1 which started in 2025 and Group C2 which started in 2026. The Project Commitment Period for Group C1 would run from 2025 to 2065, while The Project Commitment Period for Group C2 would run from 2026 to 2066. All credits would have a 20 year durability.

6.0 Environmental and Social Safeguarding

6.1 Cropland Management Activities

Project interventions involve deliberate changes to cropland management practices that may affect local ecosystems, communities, and land users beyond the direct project benefit.

7.0 Relation to Isometric Standard

7.1 Financial Additionality

SOC-enhancing land management practices frequently require upfront investment in new inputs, equipment, and agronomic expertise, while delivering climate benefits that are not directly captured by commodity markets.

Project Proponents must demonstrate that the adoption of project activities is contingent on Carbon Finance, i.e., that the practices would not be economically viable without the revenue generated from credit issuance.

Where project activities generate revenue from commodity production or other non-carbon sources within The Project area, Project Proponents must demonstrate that this revenue alone is insufficient to make The Project financially viable, in accordance with the financial additionality (An evaluation of the likelihood that an intervention that causes a climate benefit above and beyond what would have happened in a no-intervention Baseline scenario was the result of revenues from carbon finance.) requirements of the Improved Soil Management Protocol and the Isometric Standard.

Projects must not occur in regions where implemented SOC-enhancing practices are already being driven to adoption by market demand, agricultural policy, or regulatory requirements that would lead to equivalent practice changes without Carbon Finance.

8.0 System Boundary, Project Baseline and Leakage

8.1 System Boundary

A cradle-to-grave (​​Considering impacts at each stage of a product's life cycle, from the time natural resources are extracted from the ground and processed through each subsequent stage of manufacturing, transportation, product use, and ultimately, disposal.)GHG Statement (A document submitted alongside Claimed Removals and/or Reductions that details the calculations associated with a Removal or Reduction, including the Project's emissions, Removals, Reductions and Leakages, presented together in net metric tonnes of CO₂e per Removal or Reduction.) must be prepared encompassing the GHG emissions and removals relating to all activities within the system boundary. The Project Proponent is responsible for identifying all sources of emissions directly or indirectly related to project activities.

Any emissions from sub-processes or process changes that would not have taken place without the CDR (Activities that remove carbon dioxide (CO₂) from the atmosphere and store it in products or geological, terrestrial, and oceanic Reservoirs. CDR includes the enhancement of biological or geochemical sinks and direct air capture (DAC) and storage, but excludes natural CO₂ uptake not directly caused by human intervention.) Project must be fully considered in the system boundary. Any additional activity that ultimately leads to the issuance of Credits must be included in the system boundary.

The system boundary must include all relevant GHG SSRs controlled and related to The Project, as set out in Table 4 of the Improved Soil Management Protocol. Any emissions from sub-processes or process changes that would not have taken place without the CDR Project must be fully considered in the system boundary.

8.1.1 Activities Integrated into Existing Practices

Cropland management projects are implemented on land under active agricultural production, meaning that certain operational activities (e.g., planting, harvesting, fertilizer application, tillage passes) were occurring prior to, and may continue alongside, project activities. For the purpose of this provision, an "activity" may refer to an operational sub-unit, such as an individual equipment pass, a single fertilizer application event, or a discrete field management step, where such sub-units can be cleanly delineated by equipment, timing, and physical scope. Where a pre-existing activity is partially modified by The Project (for example, a field operation whose frequency is altered, or an application event whose inputs are changed), the activity must be partitioned into:

• Portions that are materially unchanged by The Project, which may be excluded from the system boundary under this provision; and
• Portions that are new, extended, or altered as a result of The Project, which must remain within the system boundary and be quantified in accordance with Section 9.5 of the Improved Soil Management Protocol.

Evidence supporting these conditions must be provided in the the LCA (An analysis of the balance of positive and negative emissions associated with a certain process, which includes all of the flows of CO₂ and other GHGs, along with other environmental or social impacts of concern.). This must include either:

• Historic farm records documenting the activity prior to project start. Acceptable records include management logs, operational records, equipment usage records, invoices, or equivalent documentation; or
• A signed affidavit from the relevant operator (e.g., farmer, land manager, or equivalent party) confirming the activity was part of routine operations prior to The Project.

And the following:

• A signed affidavit from the relevant operator confirming:
  • The equipment, timing, cadence, and intensity of the activity are not materially changed as a result of The Project; and
  • The activity would have continued at a comparable level absent The Project.

Where these conditions are met, only the emissions associated with the activity as it would have occurred in the baseline may be excluded. Any incremental emissions attributable to The Project must remain within the system boundary and be accounted for in the relevant emissions section.

8.1.2 Module-Specific SSR Considerations

8.1.3 Excluded Pools and Sources

In accordance with the Improved Soil Management Protocol, the following are excluded from the system boundary for cropland management projects:

• Aboveground woody biomass and belowground woody biomass carbon pools (these terms must be set to zero in Equation 7 of the Improved Soil Management Protocol);
• Deadwood, litter, and non-woody herbaceous biomass carbon pools, as these are considered transient

8.2 Project Baseline

8.2.1 Counterfactual Carbon Storage

8.2.2 Reductions from Project Baseline

In accordance with Section 8.2.1 of the Improved Soil Management Protocol, improved cropland management interventions may involve reductions in activities that result in emissions, such as reduced CO₂ from diesel use in tractors corresponding to fewer passes in a no-till intervention, or avoided N₂O emissions from reduced fertilizer inputs to soils.

Emissions reductions do not inflate removals by counting against CO₂e_Emissions,RP and are credited separately in the Agricultural Practices Reductions Module. Emissions outlined in Table 1 of the Improved Soil Management Protocol and reported in accordance with Section 9.5 of the Improved Soil Management Protocol must be strictly positive.

This is in keeping with principles of conservativeness and maintaining consistency with Isometric's approach to GHG Accounting for removals as outlined in the GHG Accounting Module v1.1.

8.3 Leakage

8.3.1 Overview of Leakage Assessment

Leakage emissions, [math: CO_2e_{Leakage}], occur when project activities lead to emissions that occur outside the system boundary of cropland management projects. For cropland management projects, the primary leakage risk is market-mediated leakage. This risk stems from project-induced reductions in agricultural production that may lead to land conversion and associated GHG emissions elsewhere to meet the supply shortfall. As a principle, Projects should seek to maintain or enhance yields from baseline levels.

Three key types of leakage can theoretically occur for cropland management projects, although, as noted above, the primary focus of the leakage assessment is market-mediated leakage:

• Activity-shifting leakage. Cropland management projects may displace activities in their project areas, leading to an increase in those activities outside of The Project area, which may result in potential land conversion. Examples are where the local community can no longer use the whole project site for subsistence, or where part of a farmer’s productivity is displaced as a result of project activities. This type of leakage is known as “Direct” leakage as the relevant stakeholders (Any person or entity who can potentially affect or be affected by Isometric or an individual Project activity.) can be identified and the activity-shifting is traceable.
• Market leakage. Cropland management projects may displace activities, which results in a reduction in supply of a commodity. Changes to the supply and demand equilibrium causes other market actors to shift their activities, leading to potential land conversion. This type of leakage is known as “Indirect” leakage because its effects cannot be isolated and measured directly. Quantifying the likelihood and potential magnitude of market leakage is complex and relies heavily on modeling and available literature.
• Ecological leakage. Project activities may lead to emissions in areas outside of The Project site as a result of ecological interactions, for example unintended hydrological impacts, introduction of disease, or secondary impacts of faunal influx.

All market-leakage discounts under this Module are applied to removal credits. Where a Project concurrently generates emission reduction credits under the Agricultural Practices Reductions Module, no leakage discount is applied to those reduction credits. This allocation reflects the principle that market-mediated leakage from yield displacement is conservative (Purposefully erring on the side of caution under conditions of Uncertainty by choosing input parameter values that will result in a lower net CO₂ Removal or GHG Reduction than if using the median input values. This is done to increase the likelihood that a given Removal or Reduction calculation is an underestimation rather than an overestimation.) to avoid the risk that emissions associated with removals are undercounted leading to over-crediting of removals.

8.3.2 Productivity Assessment

8.3.2.1 Regionally-Indexed Pre-Project Productivity

Pre-Project Productivity ([math: PPP]) is defined as the annual productivity of a commodity type [math: c] on Project field [math: f] in relevant units (e.g., tonnes/ ha yr). [math: PPP] must be calculated on a per-commodity basis using field-level yields indexed against regional benchmarks. This approach normalizes field-level performance against county-level (or equivalent sub-national jurisdiction) yields, controlling for weather, pest, and commodity-price variation that affects all fields in the region equally.

[math: PPP] for commodity [math: c] is calculated as:

[math: PPP_{f,c} = \bar{Y}_{\text{regional,c}} \times \frac{\sum_{t \in T_{f,c}} Y_{f,c,t}}{\sum_{t \in T_{f,c}} Y_{\text{regional,c,t}}}]

(Equation 1)

Where:

• [math: PPP_{f,c}] is the Pre-Project Productivity for commodity [math: c] on field [math: f], in production units per hectare per year
• [math: \bar{Y}_{regional,c}] is the mean regional yield for commodity [math: c] in production units per hectare over the baseline period (minimum 5 years, may be longer)
• [math: Y_{f,c,t}] is the field-level yield for commodity [math: c] on field [math: f] in year [math: t] of the baseline period
• [math: Y_{regional,c,t}] is the regional yield for commodity [math: c] in year [math: t] over the baseline period
• [math: T_{f,c}] is the set of years in which field [math: f] was planted with commodity [math: c] during the baseline period (minimum 5 years, may be longer)

[math: PPP] must be calculated separately for each commodity within the field's crop rotation. For example, in a corn-soy rotation with a 5-year baseline, corn [math: PPP] is calculated from the corn rotations and soy [math: PPP] from the soy rotations within the 5-year baseline period.

8.3.2.2 Project-Scenario Productivity

Project-Scenario Productivity ([math: PSP]) for commodity [math: c] on field [math: f] in Reporting Period [math: RP] is calculated as:

[math: PSP_{f,c,RP} = \bar{Y}_{\text{regional,c}} \times \frac{\sum_{t \in W_{f,c,RP}} Y_{f,c,t}}{\sum_{t \in W_{f,c,RP}} Y_{\text{regional,c,t}}}, \qquad RP \geq t_0 + 4]

(Equation 2)

Where:

• [math: PSP_{f,c,RP}] is The Project-Scenario Productivity for commodity [math: c] on field [math: f] in Reporting Period [math: RP], in production units per hectare per year
• [math: \bar{Y}_{regional,c}] is the mean regional yield for commodity [math: c] in production units per hectare over the baseline period (minimum 5 years, may be longer) as defined in Equation 1
• [math: W_{f,c,RP}] is the set of years t satisfying [math: RP] - [math: 4 ≤ t ≤ RP] in which field f grew crop c
• [math: t_{0}] is the first project year
• [math: Y_{f,c,t}] is the Yield of field f for crop c in year t in production units per hectare
• [math: Y_{regional,c,t}] is the regional average yield for commodity [math: c] in production units per hectare, drawn from the same regional yield source used for [math: \bar{Y}_{regional,c}]

Project-scenario productivity is assessed on a rolling basis beginning in the fifth project year ([math: RP = t_{0} + 4]) and in each Reporting Period thereafter. The first assessment applies a retroactive lookback covering project years 1 through 5; subsequent assessments use the trailing five calendar years ending in Reporting Period [math: RP]. Within each window, only years in which field [math: f] grew crop [math: c] contribute to the calculation. No productivity assessment is performed in project years 1 through 4, though yield data for these years must be collected and reported from project year 1 for use in the year-5 assessment and subsequent rolling windows. Field yields are indexed to regional average yields over the same years, so that regional weather and market-driven variation does not distort the productivity estimate. The multiplier [math: \bar{Y}_{regional,c}] is fixed at project start and is identical to that used in [math: PPP_{f,c}] ensuring both metrics are directly comparable. Where crop c was not grown on field f in any year of the window, [math: PSP_{f,c,RP}] is undefined and no assessment is made for that field-crop pair.

8.3.3 Commodity-Specific Productivity Shortfall

The productivity shortfall for each commodity [math: c] in Reporting Period [math: RP] is:

[math: \Delta P_{c,f,RP} = \max(PPP_{c,f} - PSP_{c,f,RP}, 0)]

(Equation 3)

No leakage assessment is required for commodities where [math: \Delta P_{c,f,RP} = 0] (i.e., project-scenario productivity meets or exceeds the baseline).

8.3.4 De Minimis Threshold

The de minimis test is applied to the net productivity shortfall for each commodity, after within-project leakage mitigation has been applied per Equation 6. The Project-level expected production (EP) for commodity c is:

[math: EP_{c,RP} = \sum_f PPP_{f,c} \times A_{f,c,RP}]

(Equation 4)

Where the sum is taken over all fields [math: f] planted with commodity [math: c] in the Reporting Period with a defined [math: PSP_{f,c,RP}], and [math: A_{f,c,RP}] is the area planted with commodity [math: c] on field [math: f] in the Reporting Period, in hectares.

Where [math: NPP_{c,RP} / EP_{c,RP} \leq 0.03] (i.e., the net productivity decline is 3% or less relative to expected baseline production for commodity [math: c] across The Project), the decline is within the de minimis materiality (An acceptable difference between reported Removals/emissions or Reductions/emissions and what an auditor determines is the actual Removal/emissions or Reduction/emissions.) threshold and no leakage discount applies for that commodity in that Reporting Period. The de minimis threshold is a Materiality gate, not a deductible: where the net productivity decline exceeds 3%, leakage must be assessed on the full value of [math: NPP_{c,RP}], not only the portion exceeding the threshold.

The de minimis test operates on the net shortfall because the market-mediated leakage risk arises from The Project's net impact on commodity supply; production surpluses on fields within The Project relieve the same market pressure that shortfalls create.

8.3.5 Crediting Suspension Threshold

The crediting suspension threshold is applied to the gross productivity shortfall for each commodity, before within-project leakage mitigation. The gross shortfall (GS) ratio for commodity [math: c] is:

[math: GS_{c,RP} = \frac{\sum_f \Delta P_{f,c,RP} \times A_{f,c,RP}}{EP_{c,RP}}]

(Equation 5)

Where [math: GS_{c,RP} > 0.15] (i.e., the gross productivity decline exceeds 15% for any single commodity across The Project), The Project is ineligible for crediting for that Reporting Period unless a temporary exemption is granted under the exceptional circumstances provision detailed in Section 8.6. Credits attributable to that Reporting Period are forfeited and may not be earned back at any point during the remainder of the Crediting Period.

The crediting suspension threshold tests the gross shortfall, without offset from within-project surpluses, because the threshold serves a different function from the leakage quantification: it is a detector of systematic, practice-driven productivity failure. Permitting surpluses on some fields to mask severe declines on others would defeat this function, even where the net market impact is modest.

For fields that unenroll from The Project, whether due to productivity declines coinciding with The Project intervention or for any other reason, the field's productivity in its final enrolled year must be included in the calculation of [math: GS_{c,RP}] and [math: EP_{c,RP}] for the Reporting Period in which unenrollment occurs, in addition to the unenrollment true-up requirements of Section 8.3.7. Unenrollment does not remove a field's contribution to The Project-wide crediting suspension threshold assessment for that Reporting Period.

8.3.6 Independent Assessment per Reporting Period

The productivity assessment is performed independently for each Reporting Period against the original pre-project baseline. Each Reporting Period's leakage is calculated without reference to the leakage assessed in any prior Reporting Period. There are two exceptions to this rule: (i) leakage in years 1 through 4, as detailed in Section 8.3.6.1; and (ii) the unenrollment true-up under Section 8.3.7, which may apply an exit assessment against credits issued in prior Reporting Periods.

8.3.6.1 Leakage Assessment in Years 1 through 4

Leakage is first assessed in the fifth project year ([math: RP = t_{0} + 4]), coinciding with the first assessment of Project-Scenario Productivity under Equation 2. No leakage assessment is performed in project years 1 through 4; however, credits issued for Reporting Periods 1 through 4 are provisional with respect to leakage and remain subject to the retroactive assessment described here.

At the first assessment, [math: PSP_{f,c,RP}] is calculated using the full retroactive lookback covering project years 1 through 5, and any resulting leakage discount is applied to the cumulative credits issued or issuable for Reporting Periods 1 through 5. The de minimis threshold and within-project netting provisions apply to the full retroactive period. The crediting suspension threshold, however, does not apply retroactively. Where the first assessment identifies a productivity decline exceeding the crediting suspension threshold for any commodity, ineligibility for crediting applies to Reporting Period 5 only; credits issued for Reporting Periods 1 through 4 are not voided, though they remain subject to the standard leakage discount calculated at the first assessment. From the sixth project year onward, leakage is assessed independently in each Reporting Period using the trailing five-year window per Equation 2, without reference to leakage assessed in any prior Reporting Period, and applies only to credits for that Reporting Period.

8.3.7 Unenrollment True-Up

Fields that unenroll from The Project, or are removed from The Project for any reason, remain subject to a final productivity assessment ("true-up") in the Reporting Period in which unenrollment takes effect. The true-up ensures that a field's productivity shortfall history is settled before the field exits the leakage assessment, and that unenrollment cannot be used to remove underperforming fields from the dataset before their leakage impact is quantified.

8.3.7.1 Exit Assessment

For each unenrolling field [math: f] and each commodity [math: c] grown on that field, a final [math: PSP_{f,c,RP}] must be calculated per Equation 2 using the trailing five-year window ending in the field's final enrolled year. Where a field unenrolls before the first productivity assessment (project years 1 through 4), the exit assessment must be calculated using all available project years, and the field's yield data must additionally be included in the retroactive first assessment at year 5 as if the field remained enrolled through its final enrolled year.

The field's resulting shortfall (Equation 3) and, where the eligibility conditions of the Within-Project Leakage Mitigation provisions are met, surplus (Equation 6) are included in The Project-level aggregations (Equations 4 through 8) for the Reporting Period in which unenrollment takes effect, using the field's area in its final enrolled year. The field's contribution to the crediting suspension threshold assessment is as set out in the Crediting Suspension Threshold section (Section 8.3.5).

8.3.7.2 Settlement

Leakage attributable to an unenrolled field is quantified and applied within the exit Reporting Period's leakage calculation. Where the resulting leakage discount exceeds the credits available to The Project in the exit Reporting Period, the remainder must be deducted from subsequent issuances until fully settled; it may not be waived or earned back. The true-up does not reopen or retroactively void credits issued in prior Reporting Periods, except as provided for Reporting Periods 1 through 5 under the Leakage Assessment in Years 1 through 4 provisions.

8.3.7.3 Data Requirements

The Project Proponent must retain field-level yield records for each enrolled field through its final enrolled year, including for fields that signal intent to unenroll. Where yield data for any year in the exit assessment window are unavailable, the field's yield for those years must be deemed zero for purposes of the exit assessment, unless a temporary exemption is granted under the exceptional circumstances provision in Section 8.6. Documented crop failures on unenrolling fields are treated per the Crop Failure provisions.

8.3.8 Within-Project Leakage Mitigation

8.3.8.1 Eligibility for Leakage Mitigation

A field may generate leakage mitigation (a productivity surplus that can offset negative leakage elsewhere in The Project) only where all of the following conditions are met:

• The field is enrolled in The Project and implementing SOC-enhancing interventions under the Improved Soil Management Protocol and this Module. This requirement is to provide confidence that the yield increases are a result of The Project intervention(s), not incidental.
• The yield increase exceeds 3% above the Pre-Project Productivity for the commodity. This gate is deliberatively stricter than the treatment of shortfalls:  only the surplus portion exceeding 103% of [math: PPP_{f,c}]​ is counted (Equation 4), whereas shortfalls are counted in full from the first unit. This asymmetry is conservative and ensures that only meaningful, above-trend increases generate mitigation, not normal year-to-year variability or background yield growth.
• The commodity is not a switched commodity within its first five years on the field, per the Section 8.5.5 Legitimate Reasons for Cropping Change provisions.

8.3.8.2 Calculation of Positive Leakage

The productivity surplus for commodity [math: c] on field [math: f] in Reporting Period [math: RP] is:

[math: \Delta P^{+}_{c,f,RP} = \max(PSP_{c,f,RP} - PPP_{c,f} \times 1.03, 0)]

(Equation 6)

Where:

• [math: \Delta P^{+}_{c,f,RP}] is the productivity surplus for commodity [math: c] on field [math: f], in production units per hectare.
• [math: PSP_{c,f,RP}] is The Project-Scenario Productivity for commodity [math: c] on field [math: f] (Equation 2).
• [math: PPP_{c,f}] is the Pre-Project Productivity for commodity [math: c] on field [math: f] (Equation 1).
• The factor of 1.03 applies the 3% positive-leakage gate.

The total project-level productivity surplus for commodity [math: c] is:

[math: Total Surplus_{c,RP} = \sum_f \Delta P^{+}_{c,f,RP} \times A_{c,f,RP}]

(Equation 7)

Where the sum is taken by over all fields [math: f] within The Project that are eligible for positive leakage and are planted with commodity [math: c] in the Reporting Period, and [math: A_{c,f,RP}] is the area planted with commodity [math: c] on field [math: f] in the Reporting Period, in hectares.

8.3.8.3 Within-Project Leakage Mitigation Conditions

8.3.8.3.1 Commodity Matching

Positive leakage may only offset negative leakage within the same commodity. A corn yield surplus cannot offset a soy yield shortfall as these are different markets with different supply chains and different land conversion intensities. Netting across commodities would conflate distinct market responses and risk underestimating actual leakage. A consequence of this rule is that shortfalls on a commodity grown nowhere else in The Project, including switched commodities, cannot be offset at all; this outcome is intentional and conservative.

8.3.8.3.2 Leakage Mitigation Banking Prohibition

Positive leakage may only offset negative leakage within the same Reporting Period. Positive leakage cannot be carried over to offset negative leakage in a subsequent (or prior) Reporting Period. For the purposes of this rule, the first assessment's retroactive lookback covering project years 1 through 5 constitutes a single assessment event, and netting operates across the full lookback.

8.3.8.3.3 Order of Operations for Positive Leakage Estimation

The threshold tests and netting are applied in the following sequence for each commodity [math: c] in each Reporting Period:

1. Calculate field-level shortfalls (Equation 3) and eligible field-level surpluses (Equation 6).
2. Aggregate to The Project level: gross shortfall (numerator of Equation 5), Total Surplus (Equation 7), and expected production (Equation 4).
3. Apply the Crediting Suspension Threshold to the gross shortfall ratio (Equation 5). Within-project surpluses do not offset the threshold test.
4. Apply within-project netting to obtain Net Project Productivity, [math: NPP_{c,RP}] (Equation 8).
5. Apply the de minimis test to [math: NPP_{c,RP} / EP_{c,RP}].
6. Where the de minimis threshold is exceeded, carry [math: NPP_{c,RP}] forward to the induced land conversion calculation (Equation 9).

This sequence ensures the market-response model operates on the net supply impact after within-project mitigation, while the Crediting Suspension Threshold retains its function as a detector of systematic productivity failure.

8.4 Leakage Quantification

8.4.1 Net Project Productivity

Net Project Productivity for each commodity [math: c] is calculated as:

[math: NPP_{c,RP} = \max\left( \sum_f \left( \Delta P_{f,c,RP} \times A_{f,c,RP} \right) - Total\ Surplus_{c,RP},\ 0 \right)]

(Equation 8)

Where:

• [math: \Delta P_{c,f,RP}] is the productivity shortfall for commodity [math: c] on field [math: f] (Equation 3), in production units per hectare per year.
• [math: A_{c,f,RP}] is the area planted with commodity c on field f in the Reporting Period, in hectares.
• [math: TotalSurplus_{c,RP}] is the within-project positive-leakage surplus for commodity [math: c] (Equation 7), in production units per year. The sum is taken over all fields [math: f] within The Project planted with commodity [math: c] in the Reporting Period.

The max ensures [math: NPP_{c,RP}] is non-negative even when within-project positive leakage exceeds the gross shortfall.

8.4.2 Induced Land Conversion

Project Proponents are required to estimate the amount of new land brought into production, [math: ha_{LC}]. This estimate must be informed by:

• Pre-Project Productivity of a commodity type at The Project site (after accounting for system productivity);
• The estimated proportion of this productivity that would be replaced with new production via an increase in supply of the commodity type;
• The increase in supply that would result in new land being brought into production; and
• The yield of new land being brought into production.

The new land brought into production must be calculated separately for each commodity type being displaced as a result of The Project.

For each commodity [math: c] with [math: NPP_{c,RP} > 0], the hectares of induced land conversion are calculated as:

[math: ha_{LC,c,RP} = \frac{NPP_{c,RP} \times IS_c \times NL_c}{Y_{NL,c}}]

(Equation 9)

Where:

• [math: ha_{LC,c,RP}] is the induced land conversion for commodity [math: c] in the Reporting Period, in hectares.
• [math: NPP_{c,RP}] is the Net Project Productivity, in appropriate units (e.g., tonnes per year).
• [math: IS_c] is the Increased Supply fraction for commodity [math: c], the proportion of the foregone deficit that will be replaced by increased supply elsewhere, calculated from supply and demand elasticities in accordance with Section 8.4.2.1 of this Module and Appendix D.
• [math: NL_c] is the proportion of increased supply for commodity [math: c] that will result in new land being brought into production, sourced in accordance with Section 8.4.2.2 of this Module and Appendix D.
• [math: Y_{NL,c}] is the yield on new land brought into production for commodity [math: c], in production units per hectare per year, determined in accordance with Section 8.4.2.3 of this Module.
• [math: c] is commodity type.

Where The Project falls into regions for which Isometric has provided default [math: IS] and [math: NL] values in Appendix D, those default values must be used. For all other regions, values must be sourced from literature following the procedures set out in Appendix D.

8.4.2.1 Estimating Increase Supply, IS

Increased Supply ([math: IS]) is the proportion of foregone productivity that will be replaced by increased supply elsewhere. This is underpinned by the premise that foregone production will not necessarily be replaced in totality by increased supply elsewhere as a result of elasticities of supply and demand. Global markets for commodities have been assumed for the purposes of the leakage assessment.

Estimates for IS are determined using the following equation:

[math: IS = \frac{\varepsilon_{s,c}}{\varepsilon_{s,c} + \lvert \varepsilon_{d,c} \rvert}]

(Equation 10)

Where:

• [math: IS] is increased supply.
• [math: \epsilon_{s}] is elasticity in supply, as a ratio.
• [math: \epsilon_{d}] is elasticity in demand, as a ratio.
• [math: c] is commodity type.

Isometric has carried out a literature review of [math: \epsilon_{s}] and [math: \epsilon_{d}] values for certain regions. Values for [math: \epsilon_{s}] and [math: \epsilon_{d}] for these regions are provided in Appendix D. Where The Project falls into these regions, the default values provided must be used. This is because understanding which values to use from literature is challenging as academic papers are typically not written with this purpose or audience in mind. Isometric has completed this work for certain regions to lessen this complexity and provide consistency across projects.

The default values also serve as an example of appropriate values to select from the literature for other regions; however, it should be noted that the quality of research differs across regions. For all other regions, values for [math: \epsilon_{s}] and [math: \epsilon_{d}] must be sourced from literature. The procedure and requirements for sourcing default values for [math: \epsilon_{s}] and [math: \epsilon_{d}] are set out in Appendix D.

8.4.2.2 Estimating Increased Supply That Will Result in New Land, NL, Being Brought into Production

[math: NL] considers the percentage of increased supply that will result in new land brought into production for the commodity type. This is underpinned by the premise that not all increased supply will result in new lands being brought into production. Some increased supply may be made up of intensification of activities and increased yields on existing production lands.

Isometric have carried out a literature review of [math: NL] values for certain regions. Values for [math: NL] for these regions are provided in Appendix D; their use is required where The Project falls into those regions, per Leakage Parameter Sourcing. The procedure and requirements for sourcing default values for NL are set out in Appendix D

The default values also serve as an example of appropriate values to select, however it should be noted that the quality of research differs across regions.

8.4.2.3 Estimating Yield Productivity on New Land, NL, Brought into Production

[math: Y_{NL,c}] considers the yield on new land brought into production for commodity [math: c]. This is assessed based on the observed productivity in the region in the pre-project period. Here, the value of the regional mean yield ([math: \bar{Y}_{regional,c}]) used in Section 8.3.2.1 for assessing pre-project productivity must be assumed as the the value of [math: Y_{NL,c}].

8.4.2.4 Determining the Carbon Stock Emission Factor, EFCarbon Stock

[math: EF_{Carbon Stock}] must be derived from the IPCC average national aboveground biomass content of the land cover for the relevant ecosystem. Mean carbon stocks should be derived from aboveground biomass estimates in Table 3A.1.4 of the IPCC Good Practice Guidance for Land Use, Land Use Change and Forestry⁸.

Carbon stocks should be determined using the ratio of mass of CO₂ to mass of C, and carbon fraction, [math: CF], specified for the relevant ecosystem/vegetation type by the IPCC.

8.4.3 Leakage Emission Calculation

Total leakage emissions for the Reporting Period are the sum of leakage across all commodities with a productivity shortfall:

[math: CO_2e_{Leakage,RP} = \sum_c \left( ha_{LC,c,RP} \times EF_{Carbon Stock} \right)]

(Equation 11)

Where:

• [math: CO_2e_{Leakage,RP}] is the total leakage emission for the Reporting Period, in tonnes CO₂e.
• The sum is taken over all commodities [math: c] for which [math: NPP_{c,RP} > 0].

[math: CO_2e_{Leakage,RP}] is included as part of [math: CO_2e_{Emissions,RP}] as set out in the Improved Soil Management Protocol. The leakage discount is applied exclusively to removal credits; no leakage discount is applied to emission reduction credits issued under the Agricultural Practices Reductions Module.

[math: CO_2e_{Leakage}] is quantified for every Reporting Period.

8.5 Crop Rotation Requirements

8.5.1 Baseline Rotation

8.5.2 Rotation Maintenance

The Project Proponent must maintain the same set of commodities as the baseline rotation within a ±1 year tolerance over each 5-year window. For example, a baseline rotation of corn-corn-soy-corn-soy permits any combination that includes 2–4 years of corn and 1–3 years of soy within the next 5 years.

For fallow rotations, The Project can maintain the use of fallow rotations at the same historical rate. If the number of fallow rotations within The Project exceeds the historical rate (e.g., a third fallow rotation within a 5-year period when the baseline featured two), the forgone production for the additional fallow rotation is assumed to equal the average production of the most productive commodity grown on the field during the baseline period.

8.5.3 Pre-Registered Crop Changes

Crop changes beyond the ±1 year tolerance are permitted if pre-registered at least 1 month before planting with supporting evidence that the change is driven by regional market trends, agronomic factors, or farm-level economic conditions unrelated to The Project. Pre-registered crop changes that meet these criteria do not incur a leakage penalty.

8.5.4 Unregistered Crop Changes

Any unregistered crop change that results in a shift to a lower-value commodity or an increase in fallow rotations beyond the baseline rate must be treated as a productivity shortfall for leakage purposes. The forgone production is assumed to equal the average production of the most productive commodity grown on the field during the baseline period.

8.5.5 Legitimate Reasons for Cropping Change

Crop rotation changes beyond the ±1 year tolerance may be accommodated without leakage penalty where The Project Proponent demonstrates that the change is driven by market, agronomic, or environmental factors unrelated to project enrollment. All changes under this provision must be pre-registered per the Pre-Registered Crop Changes requirements and must satisfy one of the following objective tests.

8.5.5.1 Relative Price Test (Commodity-to-Commodity Switches)

A switch from commodity [math: B] to commodity [math: A] is justified where the relative price of [math: A] has materially risen against its baseline relationship:

[math: PR = \frac{F_{A,t+1} / F_{B,t+1}}{\bar{P}_{A,bsl} / \bar{P}_{B,bsl}} > 1.2]

(Equation 12)

where [math: F_{A,t+1}]and [math: F_{B,t+1}] are harvest-time futures prices for the upcoming cultivation cycle, observed on the date of pre-registration, and [math: \bar{P}_{A,bsl}] ​, [math: \bar{P}_{B,bsl}] are average marketing-year prices over the baseline period (per USDA NASS or equivalent). This threshold is grounded in the empirical acreage-price response literature (Hendricks et al. 2014; Miao et al. 2016), which finds corn acreage responses of 7–9% to a 20% relative price increase.

8.5.5.2 Input Price Test (Nitrogen)

A switch to a lower-nitrogen-requiring commodity is justified where regional nitrogen fertilizer prices exceed the baseline-period average by more than 40%, as measured by a published reference series such as the USDA AMS Illinois Production Cost Report (anhydrous ammonia) or an equivalent regional index, observed on the date of pre-registration. Switches to higher-nitrogen-requiring commodities are not eligible under this test regardless of nitrogen price movements, given the risk that improved soil condition attributable to The Project is enabling intensification.

8.5.5.3 Pest, Disease, and Water Context Test

A switch is justified where the Proponent documents (i) the pest, disease, or water constraint via USDA or state extension reporting, and (ii) a corresponding regional acreage response, demonstrated by a shift of at least 5% in USDA intended or observed planted acreage for the affected commodity in the relevant region.

8.5.5.4 High-Value Crop Transition

A switch into a higher-value commodity (e.g., vegetables, cotton) is justified where the Proponent provides evidence of a secured offtake contract, a documented major shift in local demand, or comparative crop budgets demonstrating the economic case. Such transitions are presumed to increase the quality-adjusted quantity of agricultural output and therefore do not incur a leakage penalty, subject to the productivity assessment below.

8.5.5.5 Exclusions

No flexibility is available under this subsection for switches from higher-value commodities (grains, oilseeds, cotton, vegetables) to lower-value or lower-intensity uses (hay, alfalfa, grazing, fallow). Such changes are treated under the Unregistered Crop Changes and fallow provisions regardless of registration or supporting evidence.

8.5.5.6 Prior-Penalty Carry-Over

Where a field has been assessed a leakage penalty for commodity [math: c] in the most recent Reporting Period in which [math: c] was grown, and the field subsequently switches away from [math: c] under any provision of this subsection, the most recent leakage penalty assessed for [math: c] must be carried over and applied to that field for the first year following the switch, calculated per Section 8.4. This prevents a productivity-shortfall history from being extinguished through crop switching.

8.5.5.7 Productivity Assessment for Switched Crops

Where a field transitions to a commodity [math: c'] not present in the baseline rotation, no direct [math: PPP_{f,c'}]​ exists. The Pre-Project Productivity for the new commodity shall be constructed by transferring the field's demonstrated relative productivity to the new crop:

[math: PPP_{f,c'} = \bar{Y}_{\text{regional,c'}} \times \overline{YR}_{f}]

(Equation 13)

where [math: \overline{YR}_{f}] is the acre-year-weighted average of [math: YR_{f,c}] across all commodities [math: c] in the field's baseline rotation, and [math: \bar{Y}_{\text{regional,c'}}] is the long-run average regional yield for [math: c'], fixed per the existing multiplier convention. [math: PSP_{f,c',RP}]​ is then calculated per Equation 2 once yield observations for [math: c'] accrue, and the productivity shortfall per Equation 3.

Fields may not generate positive leakage (productivity surplus) for a switched commodity during its first five years on the field. Additionally, because within-project netting requires commodity matching, shortfalls on a switched commodity that is not grown elsewhere in The Project cannot be offset by surpluses on other commodities.

8.6 Exceptional Circumstances Affecting PSP

Where [math: PSP] is significantly lower than projected due to exogenous natural causes beyond The Project's reasonable control (extreme weather, drought, flooding, region-wide pest or disease outbreaks), The Project Proponent may request that [math: PSP] be evaluated in a broader regional context.

The regional indexing approach (Equations 1–2) already controls for most region-wide shocks; if the field and the region both experience a drought, the yield ratio ([math: Y_{field}/Y_{regional}]) is largely unaffected. The exceptional circumstances provision therefore applies to localised events that affect The Project field but not the broader region.

Evidence must demonstrate that:

1. The cause of the [math: PSP] shortfall is unrelated to project design, management decisions, or land-use change; and
2. Comparable agricultural systems in the surrounding region did not experience similar yield impacts during the same period.

Evidence may include regional or county-level yield statistics, meteorological or disaster records, government or insurance reports, and peer-reviewed or authoritative third-party data.

Where Isometric determines that the exogenous cause contributed to [math: PSP] shortfalls, Isometric may adjust the [math: PSP] by applying a counterfactual estimate that reflects the localized nature of the impact for the purposes of leakage assessment.

8.6.1 Crop Failure

If a crop failure occurs in a Reporting Period, The Project Proponent must provide evidence that the failure was caused by factors outside The Project's control (e.g., extreme weather, pest outbreak). If accepted as a genuine crop failure:

• For commodities present in the baseline rotation, the field's yield for that Reporting Period is imputed as [math: \bar{Y}_{regional,c,t}×YR_{f,c}] the regional yield for that year scaled by the field's pre-project yield ratio. This prevents a crop failure from inflating the productivity shortfall while remaining consistent with the regional indexing applied throughout the productivity assessment.
• For switched commodities with no baseline observations (i.e., those assessed under the Section 8.5.5 Legitimate Reasons for Cropping Change provisions), the field's yield is imputed as [math: \bar{Y}_{regional,c,t}×\overline{YR}_{f}] consistent with the transferred-ratio approach used to construct [math: PPP_{f,c′}]​.

8.7 Worked Leakage Example

Setup: A project with 100 hectares enrolled in a corn-soy rotation. Baseline [math: PPP] (regionally indexed): corn = 180 bu/ha, soy = 50 bu/ha. [math: IS] = 0.70 (global calories, from Appendix D), [math: NL] = 0.28 (US cropland, from Appendix D). [math: Y_{NL}] = 150 bu/ha (corn) and 45 bu/ha (soy). [math: EF_{Carbon Stock}] = 200 t CO₂e/ha.

Reporting Period 1 (Corn Year):

• PSP (regionally indexed = 170 bu/ha. Decline = (180 − 170) / 180 = 5.6%.
• Exceeds 3% de minimis gate. Leakage assessed on the full shortfall: [math: \Delta P] = 180 − 170 = 10 bu/ha.
• [math: NPP] = 10 × 100 ha = 1,000 bu/yr.
• [math: ha_{LC}] = (1,000 × 0.70 × 0.28) / 150 = 1.307 ha.
• [math: CO_2e_{Leakage,1}] = 1.307 × 200 = 261.3 t CO₂e.

Reporting Period 2 (Soy Year):

• PSP (regionally indexed) = 46 bu/ha. Decline = (50 − 46) / 50 = 8%.
• Exceeds 3% de minimis gate. Leakage assessed on the full shortfall: [math: \Delta P] = 50 − 46 = 4 bu/ha.
• [math: NPP] = 4 × 100 ha = 400 bu/yr.
• [math: ha_{LC}] = (400 × 0.70 × 0.28) / 45 = 1.742 ha.
• [math: CO_2e_{Leakage,2}] = 1.742 × 200 = 348.4 t CO₂e.

Note: the soy-year leakage is larger than the corn-year leakage despite a smaller absolute shortfall in bushels (4 vs 10), because soy has a lower yield on new land ([math: Y_{NL}] = 45 vs 150), meaning more land must be converted per unit of displaced soy production. Each Reporting Period is assessed independently against the baseline.

Reporting Period 3 (Corn Year):

• PSP (regionally indexed) = 176 bu/ha. Decline = (180 − 176) / 180 = 2.2%.
• Below 3% de minimis gate. No leakage assessed.
• [math: CO_2e_{Leakage,3}] = 0.

Note: the 2.2% corn yield decline in RP3 is within the 3% de minimis threshold. The decline is treated as natural year-to-year variability rather than a project-induced productivity loss. Because the de minimis threshold is a gate (not a deduction), no portion of the 2.2% decline is subject to leakage; the entire Reporting Period is leakage-free for corn.

9.0 Net Carbon Quantification

The net removals are calculated following the requirements within the Improved Soil Management Protocol. Under this Module, the carbon storage ([math: CO_2e_{stored}]) is considered to consist of soil organic carbon. Terms for aboveground and belowground woody biomass must be set to 0.

9.1 Calculation of CO₂e_stored

The following calculations apply equally to both Quantification Approach 1 (direct measurement and remeasurement, Section 9.1.2) and Quantification Approach 2 (biogeochemical modeling with remeasurement, Section 9.1.3). They consume the stratum level SOC stock outputs from either approach and produce The Project-level total stock and cumulative stock change used for credit issuance.

The Project-level total carbon stock at any time [math: t] is expressed as total carbon mass of soil organic carbon across all strata within the primary quantification unit:

[math: CO_2e_{stored,t} = \sum_{j=1}^{J} \left( SOC_{j,t} \times A_j \right)\times \frac{44}{12}]

(Equation 14)

Where:

• [math: CO_2e_{stored,t}] is The Project-level carbon stock at time [math: t] in tonnes CO₂e
• [math: SOC_{j,t}] is the SOC stock in stratum [math: j] at time [math: t] in tonnes C ha^-1
• [math: A_j] is the area of stratum [math: j] in ha
• [math: J] is the total number of strata
• [math: \frac{44}{12}] is the mass ratio for converting carbon to CO₂e

[math: SOC_{j,t_n}] is substituted as follows depending on the quantification approach and crediting event type. All substitution targets are calculated on an equivalent soil mineral mass (ESM) basis using the reduced cumulative-mass formulation across the depth increments required under Section 9.1.2.4, with a single stratum-level reference mineral mass and linear interpolation between increments as defined in Equations 18–19, and with parameters estimated from the sampled depth increments set out in that Section.

• Approach 1, all events: [math: SOC_{j,t_n}] from Equation 19 (ESM-corrected direct measurement on a mineral-mass basis).
• Approach 2, interim crediting events: [math: SOC^{model}_{j,t_n}] from Equation 21 (mineral-mass model output using baseline reference mineral mass).
• Approach 2, true-up events: [math: SOC^{model,ESM}_{j,t_n}] from Equation 22 (ESM-corrected mineral-mass model output using true-up-updated reference mineral mass; ).

The values of [math: CO_2e_{stored,t}] are then used in Equation 6 of the Improved Soil Management Protocol to derive the Reporting-Period change in stored carbon ([math: CO_2e_{stored,RP}]), which feeds Equations 3 and 4 of the Protocol to calculate net removals ([math: CO_2e_{Removal,RP}]).

9.1.1 Uncertainty Propagation

• It does not require the assumption that measurement or model errors are uncorrelated beyond the field level, making it more appropriate for the structured spatial and temporal error patterns present in both direct measurement campaigns and biogeochemical SOC models;
• It can directly estimate the correlation of errors across years, avoiding underestimation of uncertainty on shorter time horizons where annual errors may be positively correlated;
• Where implemented as a hierarchical bootstrap (see below), it requires minimal assumptions about the shape of the underlying data distributions, which is particularly important for paired SOC stock differences that are typically non-normal and exhibit short-range spatial autocorrelation.

Monte Carlo simulation must be implemented as follows:

• [G-5V3Y-0, Project Proponents must demonstrate convergence by showing that the uncertainty estimate does not materially change (> 1%) with additional iterations. ]An appropriate number of iterations must be performed. Project Proponents must demonstrate convergence by showing that the uncertainty estimate does not materially change (> 1%) with additional iterations. [/G-5V3Y-0]
• [G-0SPX-0, Each iteration of the Monte Carlo simulation must be independent of other iterations.]Each iteration of the Monte Carlo simulation must be independent of other iterations. Within a given iteration, the sources of uncertainty contributing to the credit estimate must be sampled jointly, with correlations between sources preserved in accordance with the correlation requirements set out below; the requirement of inter-iteration independence does not imply that sources sampled within an iteration are independent of one another. The materially uncertain sources sampled at each iteration may include [/G-0SPX-0]
  • (i) input feature uncertainty, sampled from the probability distributions of materially uncertain inputs;
  • (ii) model parameter or weight uncertainty, sampled from the posterior distribution for Bayesian models or via bootstrap resampling for frequentist methods;
  • (iii) ensemble-member sampling, where the ensemble represents diversity in training data, hyperparameters, or architecture;
  • or (iv) any combination of these, depending on which sources are material to the model in use. The relevant sources differ by approach as specified below.
• [G-D4V8-0, Input probability distributions must be justified with reference to empirical data or published literature and must not be assumed normal without justification. Where distributions are non-normal, appropriate alternative distributions must be used. ]Input probability distributions must be justified with reference to empirical data or published literature and must not be assumed normal without justification. Where distributions are non-normal, appropriate alternative distributions must be used. [/G-D4V8-0]
• [G-MEJJ-0, Correlations between uncertain inputs must be explicitly assessed and, where material, accounted for through appropriate joint sampling procedures rather than assuming independence. ]Correlations between uncertain inputs must be explicitly assessed and, where material, accounted for through appropriate joint sampling procedures rather than assuming independence. [/G-MEJJ-0]
• [G-83E7-0, The correlation structure of errors across time periods must be characterized and incorporated into the simulation, to avoid underestimation of cumulative uncertainty over multi-year Reporting Periods.]The correlation structure of errors across time periods must be characterized and incorporated into the simulation, to avoid underestimation of cumulative uncertainty over multi-year Reporting Periods.[/G-83E7-0]

The following uncertain inputs must be characterized and included in the Monte Carlo simulation:

• In the event that any sampling locations are substituted during a given Reporting Period (see Section 9.1.2.2.4), the unpaired component of variance for these locations must be propagated separately from the paired component, with the two components using sample-count weighting. The procedure for this must be clearly documented.
• Survey uncertainty arising from estimating stratum-level mean SOC stocks from a finite sample of locations or modeled units within each stratum must also be characterized and propagated, distinct from per-location measurement and analytical uncertainty.
• Model-remeasure only: interim crediting events: model prediction error (characterized by RMSE from the validation dataset); where residuals are approximately symmetric (|skewness| ≤ 0.5), RMSE may be used as the basis for a symmetric error distribution; where significant skewness is detected under Section 9.1.3.2.3, model prediction error must instead be sampled from a fitted asymmetric distribution (e.g., empirical bootstrap of validation residuals or a parametric skew-normal distribution) that preserves the observed tail behavior parameterization uncertainty; and temporal error correlation across modeled time steps between resampling campaigns. Note that at interim events there is no measured ESM, so the reference ESM carries uncertainty from the original baseline measurement.
• Model-remeasure only: true-up events: model prediction error, characterized as set out above for interim crediting events and updated through the true-up procedure in Section 9.1.3.1.6, and parameterization uncertainty as similarly updated through the true-up procedure in Section 9.1.3.1.6. Measured ESM from the resampling campaign replaces the reference value, reducing uncertainty relative to interim events.

The conservative estimate used for credit issuance must correspond to the 30th percentile of the distribution of net CO₂e removal estimates across all simulation iterations, in line with the Isometric standard. Where the 30th percentile estimate is negative, no credits may be issued for that Reporting Period.

Where model-based error propagation is used under Measure-Model, the coverage of model predictions must be evaluated at each true-up event. Where coverage is materially below the nominal confidence level, the input uncertainty distributions must be revised before the simulation is rerun. This requirement does not apply under Measure-Remeasure.

9.1.1.1 Recommended Approach: Hierarchical Bootstrap

For propagating sampling and analytical uncertainty under both Approach 1 and Approach 2, Project Proponents are recommended to implement a hierarchical bootstrap Monte Carlo procedure, since it makes minimal assumptions about the shape of the underlying data distributions and naturally captures within-stratum and between-stratum variability. The procedure has three steps:

Step 1: Location resampling within strata. Within each stratum [math: j], resample sampling locations with replacement, drawing [math: n_j] locations where [math: n_j] is the number of sampling locations in stratum [math: j]. Resampling is performed on locations (not on individual cores or depth increments), so that the paired structure ([math: t_0] and [math: t_n] measurements at the same location) is preserved within each bootstrap iteration. Location resampling at stratum [math: j] may only be applied where rarefaction analysis (see Step 4 below) demonstrates that [math: n_j] supports stable estimates; otherwise The Project must fall back to a parametric Monte Carlo at the stratum level with explicit assumptions documented and reported at each Verification.

Step 2: Observation perturbation. For each resampled location, perturb the measured SOC concentration and soil mineral mass per depth increment by drawing from the analytical uncertainty distribution. Where a location is drawn more than once in a single bootstrap iteration, the same analytical perturbation must be applied to all repeated draws of that location, so that analytical noise is not double-counted as sampling variance. Where significant spatial autocorrelation is detected (see Step 3), the effective sample size [math: n_{eff}] replaces [math: n_j] in the resampling count for that stratum.

Step 3: Spatial autocorrelation test. Within each stratum, test for spatial autocorrelation in the location-level paired SOC stock differences using Moran's [math: I] or the empirical SOC residual variogram. Where significant autocorrelation is detected at the typical inter-location distance, compute an effective sample size [math: n_{eff}] for that stratum using a documented adjustment (e.g., Cressie's variance-inflation formula based on the fitted variogram), and substitute [math: n_{eff}] for [math: n_j] in Step 1.

Step 4: Stratum and project aggregation. Compute the stratum-level mean SOC stock change from the resampled locations. Compute The Project-level mean as the area-weighted average of stratum-level means. Repeat steps 1–4 across iterations until the 30th-percentile estimate (see below) has converged within the 1% threshold of Section 9.1.1.

Parametric fallback. Where the number of sampling locations in a stratum is too small to support stable bootstrap estimates, or where the rarefaction analysis fails (see below), Project Proponents may fall back to a parametric Monte Carlo simulation under Section 9.1.1, drawing from a stratum-level distribution justified against empirical data or literature. The fallback must be flagged in the Monitoring Report and a conservative additional uncertainty penalty applied at Isometric's direction.

9.1.2 Approach 1 — Direct Measurement and Remeasurement

Direct measurement is used to quantify changes in SOC stocks under this approach. This approach is applicable where predictive models are unavailable, have not been validated for the relevant soil type or land management context, or have not been sufficiently parameterized for The Project area. Project Proponents may also elect to use direct measurement where they prefer not to rely on modeled outputs for SOC stock change quantification.

Under this approach, baseline SOC stocks within The Project Area are established through direct field measurement at project initiation ([math: t_0]) and re-measured at each subsequent Reporting Period. The counterfactual SOC trajectory is established through paired measurement of business-as-usual control plots, as set out in Section 9.2.1. This approach is applicable to the quantification of SOC stock changes and is not used for other GHG sources or sinks within the system boundary, which are quantified separately under Section 9.5 of the Improved Soil Management Protocol.

9.1.2.1 Soil Sampling Requirements

Project Proponents are encouraged to directly adopt or adapt QA/QC procedures from established published frameworks, including those produced by the Food and Agriculture Organization of the United Nations (FAO) and available via the FAO Soils Portal, the ISO soil sampling standards (including ISO 18400-104: Soil Quality - Sampling - Part 104: Strategies), or the IPCC Good Practice Guidance for Land Use, Land-Use Change and Forestry (2003).

The following requirements apply to all sampling and re-sampling campaigns:

• Sample locations must be georeferenced to a horizontal accuracy of 3 m (or better) to enable re-sampling at the sampling locations set at project initiation during subsequent Reporting Periods;
  • At each re-sampling, samples must be collected within a 5 m radius of the initial location set at project initiation, making sure to avoid the exact location of any prior sampling disturbance, with the coordinates of the new sample collected and recorded in the Monitoring Report;
• Intra-annual variability must be considered in the sampling design, and all sampling and re-sampling campaigns must be conducted during the same crop phase (e.g., pre-planting, post-harvest) and within +/- 30 days across Reporting Periods to ensure comparability;
• Where organic amendments have been applied, Project Proponents must delay sampling or re-sampling to the latest practicable point after the previous application and the earliest practicable point before the next application, in order to minimize confounding effects on measured SOC stocks.
• Sample locations should avoid areas subject to localized edge effects that would render them unrepresentative of the stratum as a whole, such as field margins, tramlines, or gateways. Any such exclusions must be clearly disclosed, justified, and spatially documented.

9.1.2.2 Sampling Design and Stratification

9.1.2.2.1 Quantification Units

The Project Proponent must define a hierarchy of quantification units in The Project Monitoring Plan. At minimum, this must specify:

• The primary quantification unit: the unit at which credits are issued and uncertainty is reported. This is typically The Project Area or, for grouped projects, an enrolled property or farm.
• The sampling unit: the unit at which sampling locations are selected and at which a single SOC stock estimate is generated. Sampling units must nest within primary quantification units.
• Where applicable, intermediate units (e.g., field, management zone) used to organize the design.

The choice of units must reflect the management, soil, and climatic heterogeneity of The Project area and the practical constraints of sampling, including for projects involving smallholders. Sub-field stratification is not required where a coarser unit can be shown to deliver equivalent or better project-level precision.

9.1.2.2.2 Sampling Design

The default sampling design is stratified random sampling, in which each sampling unit is divided into strata more homogeneous in expected SOC than the unit as a whole. Project Proponents may instead propose:

• Conditioned Latin hypercube sampling (cLHS) or other covariate-informed probability sampling designs, where ancillary variables (e.g., elevation, slope, remote-sensed indices, digital soil maps) are available and shown to correlate with SOC stocks; or
• Model-based sampling designs (e.g., spatial coverage sampling, geostatistical designs) where The Project intends to use a model-based estimator at The Project level.

Any design must be documented in The Project Design Document and justified with reference to peer-reviewed literature. Grid sampling and unstratified simple random sampling are not permitted. Project Proponents should consider, as part of the design choice, the robustness of the design to anticipated point-level data loss arising from withdrawal, sampling-access loss, or laboratory error. A design with a large number of fine-grained strata can be more sensitive to such losses than a coarser design

9.1.2.2.3 Stratification Factors

Where stratified random sampling is used in a paired-difference design (i.e., the same locations are resampled across Reporting Periods), stratification factors should be selected for their expected influence on the rate of SOC change under project management, not on absolute baseline SOC levels. Within-location baseline absolute SOC largely cancels out in the paired difference; residual within-stratum variance is dominated by heterogeneity in management practice intensity, climate, soil texture, and initial SOC saturation. Where The Project area is homogeneous in these factors, a coarser sampling unit may deliver equivalent project-level precision to a finely stratified design.

9.1.2.2.4 Missed and Substituted Samples during Re-Sampling

Project Proponents must use the sample design set at project initiation for sampling during subsequent Reporting Periods. Sampling locations may be substituted over the course of The Project Commitment Period only when there are documented barriers, including:

• Loss of contractual access (e.g., disenrollment of a property)
• Active hazards that prevent access to sampling site (e.g., natural disaster)

In such a scenario, the sampling location must be replaced with another location in the same stratum selected using the same probability sampling rule that was applied at project initiation. The substitution must be clearly documented in the Monitoring Report for review as part of the subsequent Verification. Uncertainty associated with the substitution also must be accounted for following the requirements in Section 9.1.1. This substitute sampling point must then be used for the remainder of the Reporting Period.

Project Proponents must track all substitutions over the entire Project Commitment Period. If the number of substitutions within a single stratum exceeds 15% of sampling points within that stratum within a single Reporting Period, or cumulatively exceeds 25% of the sampling points within that stratum over The Project Commitment Period, the stratum must be re-evaluated in accordance with Section 9.1.2.2.5.

If the sampling location is only temporarily inaccessible (e.g., temporary flooding) or data collection was missed because of a documented operational issue (e.g., unforeseen capacity issues that did not allow sampling in the temporal window), the sampling location will be considered missing for the given Reporting Period. These locations may be excluded from the stratum-level paired-difference calculation for that Reporting Period and uncertainty appropriately handled. These missing sampling points must be clearly documented at the relevant Verification. If data are missing for a sampling location during a second consecutive Reporting Period, the sampling point must be substituted.

9.1.2.2.5 Redefining Strata

Strata must be defined to support direct comparison of SOC stocks across Reporting Periods. Project Proponents may re-aggregate, split, or otherwise revise strata at a Reporting Period where this improves project-level precision, where field divergence under management has rendered the original stratification non-homogeneous, or where additional data has become available. Any change must:

• Preserve a documented mapping between the previous and revised stratification, so that prior measurements can be reconciled to the new design;
• Demonstrate that the revised design does not selectively exclude areas of expected SOC loss; and
• Be reported in the Monitoring Report and approved at verification.

Project-specific strata, their areas, the sampling locations within each, and any revisions across Reporting Periods must be reported as an annex to project documentation at every verification.

9.1.2.3 Sample Size Determination

The requirements of this Section govern the design of the sampling campaign and, specifically, whether the campaign supports reliable estimation of the change in SOC stocks within each stratum. They do not determine the size of the uncertainty discount applied to issued credits, which is set separately under Section 9.1.1. Project Proponents are recommended to conduct a formal power analysis to set the number of samples at design stage; a recommended minimum-detectable-difference (MDD) power-analysis procedure is set out in Appendix E. This procedure can help ensure a well-powered design that can decrease the associated uncertainty.

A minimum of three composite samples per stratum is required as an absolute floor to support variance estimation, since a within-stratum standard deviation cannot be computed from fewer than three observations. This minimum enables estimation only; it is not a representation of the sample size needed for a well-powered design, which is determined by the rarefaction analysis below (and, as a design aid, by the recommended power analysis in Appendix E) and is typically substantially larger. Strata containing fewer than three samples must be pooled with the most similar neighboring stratum for estimation, following a documented and reproducible pooling rule.

Beginning at the first re-sampling event ([math: t_1]), Project Proponents must conduct a rarefaction analysis to demonstrate empirically that each stratum's sample count supports stable estimates of the mean SOC stock change. The rarefaction procedure is:

1. For each stratum, compute the bootstrap mean of the location-level paired SOC stock difference [math: \Delta SOC_{loc} = SOC_{loc,t_n} - SOC_{loc,t_0}] using progressively larger random subsets of the available sampling locations, starting from a minimum subset of 5 locations and increasing in increments of 1 (or 5, where the stratum contains > 50 locations) up to the full sample set.
2. Repeat each subsample size at least 200 times to generate a distribution of bootstrap means at each subsample size.
3. Identify the smallest sample size [math: n_j] at which the width of the 70% confidence interval of the bootstrap mean changes by less than 5% of the overall observed range of [math: \Delta SOC_{loc}] in the stratum, for all subsequent increments. The stratum's design sample size [math: n_j] must equal or exceed [math: n_j].

Where any stratum fails the rarefaction criterion (no [math: n^*_j \leq n_j] exists), location resampling under Section 9.1.1 Step 1 is not permitted for that stratum at that event, and the parametric fallback under Section 9.1.1 applies. The Project Proponent must increase [math: n_j] for the next Reporting Period to meet the rarefaction criterion.

9.1.2.4 Sampling Depth and Equivalent Soil Mass

Soils must be sampled using a minimum of two depth increments across the sampled profile at all sampling locations. The number of increments and the depths at which the profile is subdivided may be selected by The Project Proponent, provided at least two increments are used and the choice is documented and justified at validation (for example, with reference to expected vertical distribution of SOC change, soil horizons, rooting depth, or tillage depth). A default subdivision of 0–15 cm and 15–30 cm is recommended where The Project Proponent has no project-specific basis for an alternative. Where soils at a location are shallower than the sampled depth, the deepest increment must be reported to the sampled depth and documented. SOC content analysis may be performed on a single composited sample per increment provided soil mass data are recorded separately for each increment.

Projects must apply an ESM correction. The ESM correction must be applied across all sampled depth increments using soil mass data from all sampling locations.

For sampling events post-project initiation, this may require extension (up to 10 cm) to a further depth in order to capture the mass established at project initiation. The actual depth of collection must be recorded and reported, and the additional mineral mass below the nominal increment boundary must be included in the ESM correction calculation for the deepest increment. Where a 10 cm additional buffer does not capture an equivalent mass of soil compared to project initiation, the deepest increment must be reported on a fixed-depth basis and documented as such. Note that baseline measurements at project initiation do not require multiple depth increments for ESM purposes within a given depth band.

The credited cumulative SOC stock change must be calculated on this basis.

9.1.2.5 Sample Collection and Processing

The following requirements apply:

• Both the intended and actual sampling point locations must be recorded and georeferenced.
• SOC content, oven-dry fine soil mass, and sample volume must be obtained from the same sample or from adjacent samples taken during the same sampling event, if the sample size is insufficient. Where multiple cores are combined into a single sample, all cores must be taken from the same depth and fully homogenized prior to subsampling for the different measurements and this must be disclosed in the documentation of the sampling procedures.
• Soil cores must be collected using equipment of documented cross-sectional area. The auger or core diameter and cross-sectional area used at each sampling event must be recorded and reported at each Verification event. Core dimensions must be held constant across all sampling events within a project, including between baseline and all subsequent remeasurement campaigns. Where a change in core dimensions is unavoidable, The Project Proponent must document the change, demonstrate that sampled volumes remain comparable, and propagate any resulting volume uncertainty through the ESM correction and the Section 9.1.1 Monte Carlo simulation.

All organic material (e.g., living plants, crop residue (A product that is not an economic driver of the process it is produced in.)) must be cleared from the soil surface prior to sampling. This must be documented via a photograph taken after sample removal.

• SOC stocks within each stratum must be calculated on an equivalent soil mineral mass (ESM) basis. Use of soil mineral mass, rather than fine soil mass, ensures that the reference quantity against which SOC stocks are normalized is itself insensitive to changes in soil organic matter content driven by project activities. Sampling events are denoted [math: t_n], where [math: t_0] is the baseline measurement at project initiation and [math: t_1, t_2,]... are subsequent resampling events.

Fine soil mass per unit area. For each depth increment [math: d] in stratum [math: j] at sampling event [math: t_n]​, the oven-dry fine soil mass per unit area is:

[math: M_{fine,j,d,t_n} = \frac{m_{fine,j,d,t_n}}{\pi \left(\dfrac{\phi}{2}\right)^{2} \times N} \times 1000]

(Equation 15)

Where:

• [math: M_{fine,j,d,t_n}]​​ is the oven-dry fine soil mass per unit area in depth increment [math: D] of stratum [math: j] at sampling event [math: t_n]​ (kg m⁻²);
• [math: m_{fine,j,d,t_n}] is the oven-dry mass of the fine (< 2 mm) soil recovered for that increment ([math: g]), after exclusion of coarse material retained on the 2 mm sieve;
• [math: \phi] is the internal diameter of the auger or corer (mm);
• [math: N] is the number of cores composited into the sample, such that [math: π(ϕ/2)2×N] is the total cross-sectional area sampled (mm²);
• the factor 1000 converts g m⁻² to kg m⁻².

Coarse material must be prevented from passing through the 2 mm sieve. Deriving the fine soil mass per unit area directly from sample mass and corer cross-sectional area in this way eliminates the need for independent bulk-density sampling and the imprecision associated with it (Wendt & Hauser 2013). This is the reason core dimensions ([math: \phi]) and the compositing design (N) must be documented and held constant across sampling events.

Soil mineral mass per unit area. The soil mineral mass per unit area for each depth increment is derived from the fine soil mass per unit area, adjusted to exclude the mass of soil organic matter:

[math: M_{min,j,d,t_n} = M_{fine,j,d,t_n} \times \left(1 - 1.724 \times C_{j,d,t_n} \times 10^{-3}\right)]

(Equation 16)

Where:

• [math: M_{fine,j,d,tn}] is the oven-dry fine soil mass per unit area as derived in Equation 15 (kg m⁻²);
• [math: C_{j,d,t_n}] is the measured SOC concentration in that increment (g C kg⁻¹ fine soil); and
• 1.724 is the van Bemmelen conversion factor from organic carbon mass to organic matter mass. Where Project Proponents have project-area-specific data on the carbon fraction of soil organic matter, that value may be substituted for 1.724 with justification at validation. This protocol is only scoped to mineral soils.

Sampled volume, where required, is the product of the cross-sectional area sampled and the increment depth. ESM, where reported, is a derived quantity calculated as the oven-dry fine soil mass divided by the sampled volume; it is used only for reporting and diagnostic purposes. Neither bulk density nor sampled volume is used as a direct input for the quantification calculations.

Drying and sieving procedures must follow laboratory-specific standard operating procedures (SOPs) and must be applied consistently across all samples throughout The Project lifetime, including where there is a change in analytical laboratory. Sample processing procedures must be reported in detail, explicitly describing sieving and grinding procedures.

"Composite" in this Section refers to the physical homogenization of multiple cores collected at a single sampling location, prior to laboratory analysis. The composite is the analytical sample; the sampling location is the statistical unit. Pairwise SOC stock differences computed at the sampling-location level (the difference between the composite measurement at [math: t_n] and at [math: t_0]) are the basis for stratum-level variance estimation in Section 9.1.2.3. Project Proponents are not required to physically pool samples across sampling locations or across strata at any point.

Composite sample (composite). A single analytical sample formed by physically combining and fully homogenizing two or more soil cores collected at the same sampling location and from the same depth increment, prior to laboratory analysis or subsampling. The composite sample is the analytical unit, the specimen on which SOC concentration and soil mass are measured, whereas the sampling location is the statistical unit at which SOC stocks are estimated and at which paired ([math: t_0] vs [math: t_n]​) stock differences are computed for variance estimation under Section 9.1.2.3.

Compositing is a physical operation and is distinct from statistical pooling. Cores must not be combined across different sampling locations, across different depth increments, or across strata. Where cores are composited, all constituent cores must derive from the same sampling location and the same depth increment and must be fully homogenized before any subsampling for separate measurements; soil mass data must be recorded separately for each depth increment even where SOC content is determined on a single composite per increment (Section 9.1.2.4).

9.1.2.6 Sample Handling and Storage

Following sampling, samples may be temporarily stored on-site in a location protected from sunlight, humidity, and precipitation, with different soil materials kept separate. Soil samples must be shipped within five days, or stably stored (e.g., dried or refrigerated, but not frozen) in a way that will be maintained until analysis. Once shipped, samples must be stored under environmentally controlled conditions that minimize biological activity (e.g., dried or refrigerated, but not frozen) until analysis. The duration of refrigerated storage prior to analysis must not exceed three months.

9.1.2.7 Analytical Laboratory Requirements

All samples collected throughout The Project lifetime should be analyzed by the same laboratory wherever practicable. A transition to a different laboratory is permitted, including in cases where the incumbent laboratory is no longer able or willing to process project samples, provided that:

• The change is justified in writing and documented for review at the subsequent verification;
• The replacement laboratory satisfies the accreditation requirements set out above; and
• A documented cross-calibration procedure is completed prior to the new laboratory's results being used for quantification. The cross-calibration must involve parallel analysis of a representative set of samples (including, where available, archived samples and certified reference materials) by both laboratories, with results compared against pre-defined acceptance criteria for bias and precision. Any systematic offset identified must be addressed through method alignment or a documented correction, and the cross-calibration report must be submitted as part of project documentation.

The selected laboratory must quantify and report analytical error statistics to The Project Proponent on a regular basis, derived from repeated analyses of the same sample and from analyses of certified reference materials. The laboratory must provide documentation of its internal quality control program, including:

• Use of certified soil reference materials with known SOC content (e.g., NIST SRM 2710, BCR-129, or an equivalent traceable certified reference material appropriate for the SOC content range of The Project samples);
• Monitoring of variation in analysis against defined error thresholds; and
• Participation in external proficiency testing schemes (e.g., round-robin testing) or registration as a member of the Global Soil Laboratory Network (GLOSOLAN), or an equivalent nationally recognized program.

Quantification of analytical uncertainty for Monte Carlo propagation. For each batch of project samples, the analytical uncertainty distribution used to perturb SOC concentration measurements in the Monte Carlo simulation (Section 9.1.1) must be characterized as follows:

1. Systematic bias correction. Identify any systematic bias in the analytical run against the certified reference material(s), expressed as the mean signed difference between measured and certified SOC values. Where systematic bias exceeds the CRM's certification uncertainty, project sample measurements must be corrected for the bias and the residual uncertainty (after correction) documented in the Monitoring Report.
2. Random error characterization. The random analytical error must be characterized as the greater of:

• (i) the relative standard deviation (RSD) of laboratory duplicate analyses on project samples within the same analytical batch, expressed as a fraction of the measured value; or
• (ii) the relative error derived from CRM recovery, i.e., the standard deviation of CRM measurements within the batch divided by the certified value.

3. Monte Carlo input distribution. The random error from Step 2 must be used to parameterize a mean-zero perturbation around each measured SOC concentration value within the Monte Carlo simulation under Section 9.1.1. The distribution may be assumed Gaussian provided the laboratory's quality-control data support this assumption; where the duplicate/CRM distribution is materially non-Gaussian, the empirical distribution must be used directly via inverse-transform sampling or a documented alternative.

A minimum of 5% of project samples within each analytical batch must be analyzed as laboratory duplicates. Where the batch size is small (< 20 samples), a minimum of three duplicates must be analyzed regardless of percentage.

The Project Proponent must retain run-level analytical data for all project samples, including individual measurement results, replicate analyses, calibration records, reference material results, and associated QC flags, and ensure data is available for audit by Isometric or a VVB. Quality control documentation, together with any cross-calibration reports, must be submitted as an annex to project documentation at each verification.

Where a Project Proponent utilizes an accredited laboratory for all analyses, The Project Proponent must provide raw data files to Isometric and the VVB, if requested.

9.1.2.7.1 Third Party Validation

Where a Project Proponent utilizes laboratory facilities within an academic institution or a non-accredited commercial laboratory, this Protocol requires that 10% of samples are sent to an accredited laboratory for validation. External laboratory validation is required for both quantification and validation samples in this scenario. The analyses conducted by the third party lab must be sufficient to calculate gross CDR and confirm the analytical results of the academic or non-accredited institution.

Prior to data submission, The Project Proponent is required to identify the sampling locations that will be sent to a third part laboratory to Isometric and the VVB for approval. The third party facility must be approved by Isometric. Data from the third party validation must be sent directly to Isometric from the accredited laboratory.

If the results of the third party validation show significant discrepancies with the overall dataset, an audit will be conducted by Isometric and the VVB. As part of this audit, Isometric or the VVB may request that additional samples are sent for third party validation. In this instance, Isometric or the VVB will select the samples for validation. Other materials that may be requested in an audit include:

• Run logs of analytical instruments within the academic or non-accredited facility;
• Information on sample preparation, such as laboratory notebooks and SOPs;
• Chain of custody of the samples analyzed;
• Raw, uncorrected data files for each sample analyzed.

9.1.2.8 Sample Measurement

SOC content must be measured using dry combustion (Dumas method) with known and reported measurement uncertainty from the specific analytical run (i.e., default values not accepted).

Walkley-Black (wet) oxidation and loss on ignition (LOI) are not permitted except where no other analytical method is available, in which case their use must be justified in The Project Monitoring Plan and approved at validation. Approval may require a subset of samples to be sent for analysis using dry combustion for calibration and validation of these methods. Project Proponents must document the known limitations of these methods and apply appropriate uncertainty adjustments to the resulting SOC stock estimates.

9.1.2.9 Proximal Sensing Techniques

9.1.2.10 SOC Stock Calculation (Approach 1)

The ESM-corrected SOC stock density is obtained by expressing cumulative SOC mass as a function of cumulative soil mineral mass and evaluating it, by linear interpolation between measured increments, at a fixed reference mineral mass common to all sampling events. This is the reduced cumulative-mass formulation common to the methods of Gifford & Roderick⁹ and Ellert & Bettany¹⁰ as set out in Wendt & Hauser¹¹. A single method is prescribed and conservatism is applied at the crediting step (the 30th-percentile estimate under Section 9.1.1).

Increment quantities. For each depth increment dd d in stratum [math: j] at sampling event [math: t_n]​, and at each sampling location, compute the soil mineral mass per unit area [math: M_{min,j,d,t_n}] (Equation 16) and the SOC mass per unit area:

[math: M_{C,j,d,t_n} = C_{j,d,t_n} \times M_{fine,j,d,t_n}]

(Equation 17)

where [math: C_{j,d,t_n}] is the measured SOC concentration (g C kg⁻¹ fine soil), [math: M_{fine,j,d,t_n}]​​ is the fine soil mass per unit area (Equation 15, kg m⁻²), and [math: M_{C,j,d,t_n}] is the SOC mass per unit area for the increment (g C m⁻²).

Cumulative profiles. At each sampling location, accumulate mineral mass and SOC mass from the soil surface to the bottom of each increment [math: d]:

[math: \text{cumulative mineral mass to } d = \sum_{k=1}^{d} M_{min,j,k,t_n} \qquad \text{cumulative SOC mass to } d = \sum_{k=1}^{d} M_{C,j,k,t_n}]

Reference mineral mass. Define the reference soil mineral mass for stratum [math: j] as the minimum, across all sampling events, of the total cumulative soil mineral mass of the stratum-mean profile:

[math: M_{\min,\mathrm{ref},j} = \min_{t_n} \left( \sum_{k=1}^{D} \bar{M}_{\min,j,k,t_n} \right)]

(Equation 18)

where [math: D] is the total number of depth increments sampled (common to all sampling locations and all sampling events; see Section 9.1.2.4) and [math: \bar{M}_{min,j,k,t_n}]​​ is the mean, across all sampling locations in stratum [math: j], of the increment-[math: k] mineral mass at event [math: t_n]. Taking the across-event minimum guarantees that the reference mass lies within the sampled stratum-mean column at every event, so no extrapolation beyond the measured profile is required¹¹. At the sampling event that yields the minimum, the reference mass equals the stratum-mean profile total and the ESM correction ratio for the stratum-mean profile is unity.

ESM-corrected stock density. Let aa a be the deepest increment whose cumulative mineral mass does not exceed the reference mass, and let [math: b = a + 1] be the increment within which the reference mass falls:

[math: a = \max \left\{ d : \sum_{k=1}^{d} M_{\min,j,k,t_n} \leq M_{\min,\mathrm{ref},j} \right\}, \quad b = a + 1]

The ESM-corrected SOC stock density at a sampling location in stratum [math: j] at sampling event [math: t_n]​ is then:

[math: \mathrm{SOC}_{j,t_n} = \left[ \sum_{k=1}^{a} M_{C,j,k,t_n} + \left( M_{\min,\mathrm{ref},j} - \sum_{k=1}^{a} M_{\min,j,k,t_n} \right) \frac{M_{C,j,b,t_n}}{M_{\min,j,b,t_n}} \right] \times \frac{1}{100}]

(Equation 19)
Where:

• [math: SOC_{j,tn}]​​ is the ESM-corrected SOC stock density in stratum [math: j] at sampling event [math: t_n]​ (t C ha⁻¹);
• [math: \sum_{k=1}^{a} M_{C,j,k,t_n}] is the cumulative SOC mass to the bottom of increment [math: a] (g C m⁻²);
• [math: \sum_{k=1}^{a} M_{\min,j,k,t_n}]​​ is the cumulative soil mineral mass to the bottom of increment [math: a] (kg m⁻²);
• [math: M_{C,j,b,t_n} / M_{\min,j,b,t_n}] is the ratio of SOC mass to soil mineral mass within increment [math: b], used as the interpolation slope across that increment (g C kg⁻¹ mineral soil). This is distinct from the measured SOC concentration [math: C_{j,d,t_n}]​​, which is expressed per fine soil;
• [math: M_{\min,\mathrm{ref},j}] is the reference soil mineral mass (Equation 18, kg m⁻²);
• the factor [math: 1/100] converts g C m⁻² to t C ha⁻¹.

When [math: a = D] the reference mass equals or exceeds the location's full profile mineral mass, increment bb b does not exist, and the interpolation term is set to zero; Equation 19 reduces to the location's full-profile SOC mass divided by 100.

Where the profile is sampled as a single depth increment ([math: D=1]), set [math: a=0] with both cumulative sums equal to zero; Equation 19 then reduces to the single-layer ESM correction:

[math: \mathrm{SOC}_{j,t_n} = M_{\min,\mathrm{ref},j} \times \frac{M_{C,j,1,t_n}}{M_{\min,j,1,t_n}} \times \frac{1}{100}]

(Equation 20)

The same equation therefore serves both single-layer and multiple-layer assessments. (Under Section 9.1.2.4 a minimum of two depth increments is required at project sampling events; the [math: D=1] reduction is retained for completeness and for the baseline case described below.)

Estimation and aggregation across locations. The reference soil mineral mass [math: M_{min,ref,j}] (Equation 18) is a stratum-level constant defined on the stratum-mean mineral profile. Equation 19 must be evaluated at each sampling location, using that location's own cumulative SOC and mineral masses against the common stratum reference [math: M_{min,ref,j}]​. The stratum-level stock [math: SOC_{j,t_n}]​​ carried into Equation 14 is the mean of the location-level values across all sampling locations in the stratum. Evaluating Equation 19 at the location level is required so that the location-level paired stock differences [math: \Delta SOC_{loc} = SOC_{loc,t_n} - SOC_{loc,t_0}]​​ used in the rarefaction analysis (Section 9.1.2.3) and the hierarchical bootstrap (Section 9.1.1) are defined consistently with the crediting point estimate.

Where an individual location's total profile mineral mass at an event is less than [math: M_{min,ref,j}] (which can arise because the reference is defined on the stratum mean) that location's cumulative SOC mass must be evaluated at its own profile total (that is, a is set to the location's deepest increment and the interpolation term is zero) rather than extrapolated beyond the measured column. This truncation omits any SOC that would lie between the location's profile total and the reference mass and is therefore conservative for crediting. Each occurrence must be documented at Verification, and where truncation recurs at the same location across consecutive sampling events that location must be sampled to greater depth at the next campaign so that the measured column reaches the reference mass.

Linear interpolation yields an approximately unbiased point estimate of the ESM-corrected stock provided increments are thin enough that the SOC–mineral-mass relationship is close to linear within each increment; the point estimate must be incorporated into the Monte Carlo simulation (Section 9.1.1) to account for uncertainty.

Linear interpolation assumes SOC concentration is uniform within each increment. Where increments are thick and the SOC–mineral-mass relationship is strongly curved (for example across buried or spodic horizons), this introduces interpolation error. That error propagates as increased uncertainty in the Section 9.1.1 Monte Carlo simulation and, through the conservative lower bound, into the credited quantity. Project Proponents expecting a non-linear SOC distribution with depth should sample thinner increments across the affected depth intervals to reduce it.

Baseline event. At the baseline event, before any resampling event exists, the minimization in Equation 18 is taken over a single event and therefore returns that event's own stratum-mean profile total; the ESM correction ratio is unity at baseline by construction. The correction takes full effect from the first resampling event [math: t_1] onwards, once the cross-event minimum is defined over more than one event.

The variance terms used in Section 9.1.1 (Uncertainty Propagation) must reflect the sample size at each depth increment across all sampling locations.

Aggregation to total carbon stock. Equation 19 produces an SOC stock density for a single stratum at a single event. These densities are aggregated to The Project level, area-weighted across strata and converted to CO₂e, by Equation 14; no separate aggregation is defined here. Under this Module the stored-carbon pool is SOC only (the aboveground and belowground woody biomass terms are set to zero), so The Project total carbon stock is the SOC stock delivered by Equation 14.

9.1.3 Approach 2 — Biogeochemical Modeling with Remeasurement

Under this approach, an approved biogeochemical model (see Section 9.1.3.2.1) is used to estimate SOC stock changes between resampling campaigns based on measured initial SOC stocks, implemented land management practice changes, soil characteristics, and climatic conditions within each quantification unit.

Direct measurement of SOC stocks is required at a minimum of every five years. Remeasurement data must be used to re-estimate model prediction error and recalibrate the model against observed conditions at each Reporting Period (true-up procedure, see Section 9.1.3.2.4).

9.1.3.1 Model Requirements

9.1.3.1.1 Model Eligibility

Any model used to contribute to the quantification of net CO₂e removal under this Module must be demonstrated to be well-validated and skillful for the purpose for which it is used, including the relevant soil types, land management practices, climatic conditions, and geographic context of The Project area. Recommended biogeochemical models include established process-based models such as DayCent, RothC, and CENTURY, as well as other models that meet the following eligibility criteria. However, even recommended models must be demonstrated to be fit for purpose in the context of The Project according to the below criteria.

Model eligibility is context-specific. A model that is eligible for one project area is not automatically eligible for a project in a different ecoregion, soil type, or management context. Where the same model is to be used both to predict The Project scenario and to estimate the counterfactual under Section 9.2.2, eligibility must be demonstrated separately for The Project management practices and the baseline management practices.

The selected model must be capable of simulating SOC dynamics across the sampled depth profile required under Section 9.1.2.4.

9.1.3.1.2 Model Calibration and Parameterization

All data sources used in parameterization must be available to Isometric and the VVB. Where proprietary data sources are used, Project Proponents must demonstrate that equivalent publicly available data were not available and must provide sufficient metadata to allow independent assessment of data quality.

Digital soil maps may be used for model initialization with approval from Isometric.

9.1.3.1.3 Model Validation

Prior to use for credit-relevant quantification, the selected biogeochemical model must be validated for the specific conditions of The Project area. Validation results and supporting data must be submitted in the PDD and are subject to review and approval by Isometric at project validation.

Data points used for model parameterization or calibration must not be used for validation. This prohibition applies to all data sources used in model development, including literature-derived values used at initial parameterization and any project-collected data subsequently used for recalibration under the true-up procedure. A minimum of 20% of all available representative data points must be withheld from parameterization and reserved exclusively for validation. Representative means that the withheld data points must span the full range of SOC stock values, soil types, climatic conditions, and land management practices present in The Project area. A validation subset that satisfies the 20% threshold but clusters at one end of the observed distribution does not meet this requirement.

The validation dataset must:

• Include representative coverage of SOC stock values across the sampled depth profile required under Section 9.1.2.4;
• Be drawn from the same ecoregion and encompass soil types, climatic conditions, and land management practices comparable to those of The Project area; and
• Include the range of management practice changes implemented under The Project, so that model performance under project-relevant conditions can be directly assessed.
• Where the model is to be used to estimate the counterfactual under Section 9.2.2, additionally include observations representative of the baseline management practices that will be input to the counterfactual simulation, so that model performance under baseline-relevant conditions can be directly assessed.

Where the validation dataset is used to support a modeled counterfactual under Section 9.2.2, model performance must be reported separately for project-like and baseline-like management contexts and for project-area-relevant and project-area-non-relevant climatic contexts. The [math: R^2] performance threshold and bias requirements set out in this section apply independently to each project-like × climatic-coverage subset and to each baseline-like × climatic-coverage subset; where they are not met for the baseline-like × project-area-relevant climatic-coverage subset specifically, the model is not eligible to be used for counterfactual estimation under Section 9.2.2, regardless of its performance under other subsets.

Climatic coverage of the validation dataset must be reassessed at each Reporting Period against the climatic conditions actually experienced by The Project area in that period. Where the validation dataset does not include observations spanning the climatic conditions actually experienced (defined as fewer than three validation points within the climatic envelope of the Reporting Period under each of project-like and baseline-like management) the climatic-coverage gap must be documented and triggers the additional consequences set out in Section 9.1.3.2.3.

Based on assessment against the validation dataset, the model must demonstrate the following:

• [math: R^2], calculated as the coefficient of determination [math: (1 - \frac{RSS}{TSS})] against validation dataset SOC stock values, must be greater than 0. A 90% confidence interval for [math: R^2] must be calculated and must exclude 0, demonstrating statistically significant predictive skill. [math: R^2] must not be calculated as the coefficient of determination of a linear regression of actual against predicted values, as this can produce misleadingly high values for biased models.
• Where model-based error propagation is used, the coverage of model predictions must be evaluated. The proportion of observed values falling within the model's predicted uncertainty intervals must be assessed and reported. This requirement does not apply where model-assisted or analytic error propagation is used.
• No systematic bias in predictions across the range of SOC stock values observed in The Project area. Where bias is present, it must be characterized, documented, and corrected prior to use.

The following statistics must be calculated and reported as documentation requirements at validation, but do not constitute enforceable performance thresholds: root mean square error (RMSE), reported in t C ha[math: ^{-1}]; and mean bias error (MBE), reported in t C ha[math: ^{-1}], as the primary diagnostic for systematic overprediction or underprediction; and the skewness of the residual distribution, reported alongside a quantile-quantile plot of residuals against a normal distribution. Significant skewness in the marginal residual distribution does not by itself indicate model misspecification and does not constitute grounds for rejection under the residual structure requirements below; however, where significant skewness is detected (|skewness| > 0.5, or where the Q-Q plot indicates material asymmetry in the tails), the asymmetric error distribution must be characterised and used in Monte Carlo propagation as set out in Section 9.1.1 rather than a symmetric RMSE-based distribution.

Residual prediction errors must show no statistically significant structure with respect to soil type, management practice, or time, tested using an appropriate statistical test (e.g., ANOVA/Kruskal-Wallis for soil type and management practice; a trend or autocorrelation test for time) at a significance threshold of p < 0.10. Where stratum-level sample sizes are insufficient to support a determinative test, residuals may instead be assessed over the generalized parameter space described later in this Section, subject to the same consultation and approval requirements set out there. Where significant structure is detected, the model must be investigated and recalibrated, and may not be approved for use until the structure is resolved.

Where a model does not meet the [math: R^2] performance threshold or exhibits systematic bias that cannot be corrected, it must be reparameterized or recalibrated before resubmission for Isometric approval. Data points previously used for validation may not be reused for parameterization or recalibration.

Where The Project Proponent can demonstrate that representative in-situ data are not reasonably available for the specific combination of soil, climatic, and management conditions of The Project area, model performance over a generalized parameter space encapsulating The Project conditions may be considered for validation, subject to consultation with and approval by Isometric. The Project Proponent must document the data scarcity, characterize the parameter space against which generalized validation is conducted, and demonstrate that this space is sufficiently broad and representative to provide reasonable confidence in model performance under project conditions. The performance thresholds set out in this Section apply unchanged to the generalized validation.

9.1.3.1.4 Completeness of of Validation Data

Studies satisfying the mandatory inclusion criteria may be excluded from the corpus only on the basis of one or more of the following grounds, with each exclusion documented individually at the study level and subject to scrutiny at validation:

1. Methodological incompatibility: the study uses an SOC measurement method (e.g., wet oxidation with no dry-combustion comparator, or a non-comparable depth-aggregation procedure) that cannot be reconciled with The Project's measurement framework;
2. Documented data quality issues: the study has been subject to a published correction, retraction, or methodological criticism that materially undermines the reliability of its SOC stock measurements;
3. Soil-type incompatibility: the study covers a soil type that is not present in The Project area, where soil type is the primary control on the model's predictions;
4. Replacement by superseding study: the study has been superseded by a later peer-reviewed remeasurement of the same locations, in which case the later study replaces (rather than is added to) the earlier one in the corpus; or
5. Inaccessibility: the underlying data are not available to The Project Proponent through reasonable efforts, including direct request to the corresponding author. Where this exclusion is invoked, The Project Proponent must document the specific data-request actions taken and the response (or non-response) received.

Exclusion on grounds outside this list is not permitted. In particular, a study may not be excluded on the basis that its inclusion would widen the Monte Carlo input distributions or increase the uncertainty discount.

Changes to the corpus or to the validation subset after pre-registration require explicit Isometric approval and constitute a fresh validation event for which model performance metrics calculated prior to the change may not be relied upon.

9.1.3.1.5 Spatial Blocking of Calibration and Validation Data

Project Proponents must report at validation, for each bias-test subset (project-like, baseline-like, and the climate-coverage portions of each):

• The empirical SOC residual variogram or equivalent autocorrelation diagnostic, including the fitted autocorrelation range and 90% confidence interval on the range estimate;
• The minimum, median, and 5th-percentile distances between validation locations and the nearest calibration location;
• Where the distance-based separation design is used, evidence that the 5th-percentile distance exceeds the autocorrelation range; and
• Where the block-level hold-out design is used, the spatial definition of each held-out block, the rationale for the choice of block size, and a demonstration that no calibration-point neighborhood (defined by the autocorrelation range) crosses a block boundary into a held-out block.

Where The Project area is too compact to support distance-based separation beyond the autocorrelation range, or where the available validation dataset is too small to populate block-level hold-outs at the required statistical-power thresholds, within-project spatial blocking is not feasible and The Project Proponent must instead use regional validation data drawn from beyond The Project boundary, with the regional dataset itself satisfying the management-coverage, climate-coverage, distributional-distance, and spatial-blocking requirements of this Section. The regional dataset must be drawn from the same ecoregion, must satisfy the comparability criteria set out elsewhere in this Section, and must include sufficient spatial extent for distance-based separation to be operationalized within it. Use of the fallback must be documented at validation, including a quantitative demonstration that within-project blocking is infeasible and a justification of the regional dataset's representativeness of The Project area.

Where post-recalibration validation is conducted under Section 9.1.3.2.3 following a triggered bias test, the spatial-blocking requirements of this sub-section apply equally to the post-recalibration validation. New calibration points used in recalibration must be spatially separated from validation points under the same design as was used at initial validation. Where new calibration data are spatially correlated with existing validation data such that this requirement cannot be satisfied without restructuring the validation set, The Project Proponent must construct a fresh validation set under the spatial-blocking requirements before the recalibrated model may be used for credit-relevant quantification.

9.1.3.1.6 Model True-Up

At a maximum interval of 5 years, the biogeochemical model must be updated using new SOC stock measurements collected during the resampling campaign in accordance with Section 9.1.2. A true-up event is additionally required, before the scheduled maximum interval, for any Reporting Period in which the climatic-coverage reassessment under Section 9.1.3.2.3 shows that a climatic driver used by the model falls below (validation minimum − 0.2 × validation range) or above (validation maximum + 0.2 × validation range) of that driver across the validation dataset. In such a Reporting Period the crediting event is a true-up event rather than an interim crediting event, and the measured result is subject to divergence testing and reconciliation under Section 9.1.3.2.3.

The true-up procedure serves three purposes: integrating new observational evidence into the model's cumulative data record; reassessing model error and uncertainty; and detecting and correcting systematic bias before it propagates into subsequent credit estimates.

New field measurement data must be added to the full cumulative dataset comprising all prior literature-derived values and project-collected data used in parameterization and validation. New data points must first be used for model error reassessment before any portion is considered for recalibration, the same data points may not be used simultaneously for both purposes within the same Reporting Period.

At each true-up event, the sampling collected under Section 9.1.2.3 must be used to update model error statistics and test for systematic bias in the modeled profile. The MBE-based bias test (10% of mean observed SOC stock change) applies to the modeled profile. Where systematic overprediction is detected, the consequences set out in this section apply.

Updated error statistics must be calculated against the new observations and reported in the Monitoring Report at verification. At minimum the following must be reported, calculated separately for The Project-like and baseline-like subsets of the cumulative validation dataset (as defined in Section 9.1.3.2.3) and for each subset's project-area-relevant climatic-coverage portion: [math: R^2] calculated as [math: 1 - \frac{RSS}{TSS}], with a 90% confidence interval that must exclude 0; RMSE in t C ha[math: ^{-1}]; residual skewness with a Q-Q plot; and MBE in t C ha[math: ^{-1}], calculated as the mean of (predicted − observed) across all validation points in each subset, reported alongside its standard error and a one-sided 90% upper confidence bound on [math: |MBE|]. The pooled statistics for the cumulative validation dataset as a whole must additionally be reported. Updated error statistics must be used to revise the Monte Carlo simulation input probability distributions in accordance with Section 9.1.1, including, where residual skewness has changed materially, revisions of the error distributions. Where the true-up reveals greater model uncertainty than was previously characterized, the higher uncertainty estimate must be applied.

Three parallel bias tests apply at each true-up. Each test combines a statistical-significance criterion (to control false positives and false negatives at small or noisy validation sample sizes) with a residual minimum effect-size criterion (to prevent triggering on statistically detectable but practically immaterial bias). Any single test, if triggered, requires the mandatory consequences set out below.

Pooled bias test. The pooled-MBE test fires where both of the following are satisfied:

• The lower bound of a one-sided 90% confidence interval on MBE on the cumulative validation dataset as a whole exceeds zero, i.e., the test rejects the null hypothesis of no overprediction at the 10% one-sided significance level; and
• The point estimate of MBE on the cumulative validation dataset exceeds 5% of the mean observed SOC stock in The Project area across the cumulative dataset.

Where the test fires with a positive MBE point estimate (the model over-predicts), this constitutes systematic pooled overprediction. A negative MBE point estimate, or failure of either criterion, does not trigger the mandatory consequences but must be documented and investigated where statistically significant.

Differential bias test (management-conditional). The differential test fires where both of the following are satisfied:

• The lower bound of a one-sided 90% confidence interval on [math: (MBE_{proj} - MBE_{base})] exceeds zero, where [math: MBE_{proj}] and [math: MBE_{base}] are the point estimates on The Project-like and baseline-like subsets of the cumulative validation dataset respectively, and the confidence interval is calculated using a Welch-style standard-error formulation that accommodates unequal sample sizes and unequal within-subset variances between the two subsets; and
• The point estimate [math: (MBE_{proj} - MBE_{base})] exceeds 5% of the mean observed SOC stock in The Project area across the cumulative dataset.

Where the test fires, the model is biased toward over-estimating The Project effect (over-crediting). A negative point estimate, or failure of either criterion, does not trigger the mandatory consequences but must be documented and investigated where statistically significant.

Differential bias test (climate-conditional). The climate-conditional differential test fires where both of the following are satisfied:

• The lower bound of a one-sided 90% confidence interval on [math: (MBE_{proj,C} - MBE_{base,C})] exceeds zero, where [math: MBE_{proj,C}] and [math: MBE_{base,C}] are the point estimates on The Project-like and baseline-like subsets of the cumulative validation dataset restricted to the climatic-coverage portion overlapping with the climatic conditions actually experienced by The Project area during the Reporting Period, calculated using the same Welch-style standard-error formulation as the management-conditional test; and
• The point estimate [math: (MBE_{proj,C} - MBE_{base,C})] exceeds 5% of the mean observed SOC stock in The Project area across the cumulative dataset.

Where either subset under the climate-conditional test contains fewer than three validation points within the relevant climatic-coverage portion, the test is not statistically defined, and the climatic-coverage gap consequences set out below apply in lieu of the test result.

Counterfactual bias assessment and correction. The management-conditional and climate-conditional differential tests assess whether the model systematically over-estimates The Project effect by mis-estimating the counterfactual, evaluated against the baseline-like subset of the validation dataset. This assessment must be performed at initial model approval and re-performed upon any recalibration or other change to the model parameterization, and upon any update to the validation dataset. It is not triggered by true-up events in themselves, as a true-up provides no new observation of the counterfactual.

Where a differential test fires, the modeled counterfactual SOC stock change must be adjusted by subtracting [math: MBE_{base}] (or [math: MBE_{base,C}] where the climate-conditional test fires) from the modeled counterfactual for each Reporting Period to which the affected parameterization applies. The adjustment must be capped at the lesser of (i) 25% of the unadjusted modeled counterfactual stock change for the period, or (ii) the absolute value of the bias point estimate from the relevant subset, with the cap and any application of it documented in the Monitoring Report. Where the climate-conditional test cannot be statistically defined, the climatic-coverage gap consequences below apply. Where a counterfactual bias correction is applied, the combined project-effect estimate for each affected Reporting Period is calculated as the difference between The Project SOC stock change (measured or modeled, according to the Reporting Period type) and the bias-corrected modeled counterfactual.

Consequences where a bias test fires. The bias tests diagnose the presence, source, and location of model bias. The consequences of the pooled and management-conditional bias tests are as follows.

Where a bias test fires, the source of the bias must be investigated, documented, and reported to Isometric. The Project Proponent must respond in accordance with Section 9.1.3.2.3, by either recalibrating the model or applying an increased model prediction uncertainty in the Section 9.1.1 propagation. Where the source cannot be identified or corrected, continued use of the model must be approved by Isometric before the next Reporting Period commences. Reconciliation of project-side crediting against field measurement is governed by Section 9.1.3.2.3; the counterfactual bias correction is governed by the counterfactual bias assessment above.

Climatic-coverage gap consequences. Where the climate-conditional differential test cannot be statistically defined for a Reporting Period because the validation dataset contains fewer than three validation points within one or both subsets restricted to the climatic conditions actually experienced, the consequences are as follows:

1. For the current Reporting Period, the modeled counterfactual estimate must be subjected to a default conservative adjustment: the modeled counterfactual SOC stock change must be increased by an amount equal to the larger of (i) the absolute value of [math: MBE_{base}] on the broader baseline-like subset (i.e., the management-conditional, not climate-conditional, point estimate), or (ii) 5% of the mean observed SOC stock n The Project area across the cumulative dataset. The combined project-effect estimate is then calculated as the difference between the modeled (or measured, depending on Reporting Period type) project SOC stock change and the adjusted modeled counterfactual. The adjustment must be applied without the bias-correction cap that applies under the bias-test consequences above.
2. The Project Proponent must, prior to the next Reporting Period, expand the validation dataset to remedy the climatic-coverage gap. Where the climatic conditions experienced are unprecedented in the available validation evidence base for The Project's ecoregion, this requirement may be satisfied by acceptance from Isometric of a documented gap, with the conservative adjustment under (1) continuing to apply at each subsequent Reporting Period until the gap is closed.

9.1.3.1.7 Optional Recalibration

Project Proponents may elect to recalibrate model parameters at any true-up event using the updated cumulative dataset, subject to the following requirements:

• Recalibration must use the full cumulative dataset of literature-derived and project-collected data, excluding all data points reserved for validation;
• A minimum of 20% of all available representative data points in the updated cumulative dataset must be withheld from recalibration and used for post-recalibration validation, consistent with the data separation requirements of Section 9.1.3.2.3. A validation subset that satisfies the 20% threshold but clusters at one end of the observed distribution does not meet this requirement;
• Post-recalibration model performance must satisfy the [math: R^2] performance threshold of Section 9.1.3.2.3, that is, [math: R^2] greater than 0 with a 90% confidence interval excluding 0, and must demonstrate no systematic bias before the recalibrated model is used for credit-relevant quantification. Where recalibration was undertaken in response to a bias test firing, it must specifically target the source(s) of the identified bias and must restore each fired test to non-firing status (both criteria failing on the post-recalibration validation dataset), evidenced through subset-by-subset performance reporting under Section 9.1.3.2.3;
• All recalibration inputs, revised parameter values, updated validation results, evidence that the 20% representative validation split requirement has been satisfied, updated performance statistics, and revised Monte Carlo input probability distributions must be submitted to Isometric and approved prior to use of the recalibrated model in the next Reporting Period.

Data points previously used for validation may not be reused for parameterization or recalibration at any point in The Project lifetime.

9.1.3.1.8 Remeasurement (True-up) Soil Sampling Requirements

Remeasurement serves model calibration and true-up purposes rather than primary quantification. The following requirements apply to all resampling campaigns conducted under this approach:

• The QA/QC, georeferencing, seasonal consistency, and organic amendment timing requirements set out in Section 9.1.2.1 apply equally to remeasurement campaigns.
• The sampling design and stratification requirements of Section 9.1.2.2 apply equally to remeasurement.
• Stratification for remeasurement must be consistent with the stratification used for initial baseline measurement and for model parameterization, to ensure that remeasurement data are directly comparable to modeled outputs at the stratum level.
• The power analysis requirements of Section 9.1.2.3 apply to remeasurement campaigns. The [math: MDD] must be set at a level sufficient to detect the model prediction error identified at initial validation, in addition to the expected project-induced SOC stock change.
• At true-up events, field measurements of ESM must be collected during the resampling campaign and used to apply an ESM correction to the model output, consistent with the ESM approach selected under Section 9.1.2.4. The treatment of ESM differs by event type:
  • Interim crediting events; the model output is used on a fixed-depth basis without an ESM correction, using the reference ESM from baseline measurements.
  • True-up events: measured ESM data from the resampling campaign must be used to apply a full ESM correction to the model output, consistent with the approach described in Section 9.1.2.4.

The requirements of Section 9.1.2 and all subsections apply equally to remeasurement under Approach 2. In-situ proximal sensing (Section 9.1.2.9) may also be used for remeasurement, subject to the same requirements.

9.1.3.2 SOC Stock Calculation (Approach 2)

9.1.3.2.1 Interim Crediting Events

At interim events no field measurement is available, so the mineral-mass profile and the reference mass are held at their baseline ([math: t_0]​) values. The cumulative-mass estimator (Equation 19) evaluated at the baseline reference mass therefore has correction ratio unity ([math: a = D], interpolation term zero) and reduces to the modelled SOC mass over the baseline profile:

[math: \mathrm{SOC}_{j,t_n}^{\mathrm{model}} = \left( \sum_{d=1}^{D} \hat{C}_{j,d,t_n} \times M_{\mathrm{fine},j,d,t_0} \right) \times \frac{1}{100}]

(Equation 21)

Where:

• [math: SOC^{model}_{j,t_n}] is the modelled SOC stock density in stratum [math: j] at modelled time step [math: t_n]​, on a mineral-mass basis (t C ha⁻¹);
• [math: \hat{C}_{j,d,t_n}]​​ is the modelled SOC content in depth increment dd d of stratum [math: j] at time [math: t_n]​ (g C kg⁻¹ fine soil);
• [math: M_{fine,j,d,t_0}] is the baseline fine soil mass per unit area for increment [math: d] in stratum [math: j] (Equation 15, kg m⁻²), estimated from all sampling locations within stratum [math: j];
• [math: D] is the total number of depth increments sampled under Section 9.1.2.4. Where soils within stratum [math: j] are shallower than the sampled depth, [math: D] is reduced accordingly and the deepest increment is reported to the sampled depth;
• the factor [math: 1/100] converts g C m⁻² to t C ha⁻¹.

This equation evaluates the cumulative-mass estimator at the baseline reference mineral mass [math: M_{\min,\mathrm{ref},j}(t_0) = \sum_{k=1}^{D} \bar{M}_{\min,j,k,t_0}]​​; because no post-baseline mineral-mass measurement is available at an interim event, the cross-event correction ratio is unity and the interpolation term vanishes. Modelled SOC mass is formed on the fine soil mass basis, consistent with Equation 17.

Estimation of [math: M_{min,ref,j,d}^{(t_0)}]: the baseline reference soil mineral mass for each sampled increment is estimated using baseline ([math: t_0]​) measurements from all sampling locations within stratum [math: j].

9.1.3.2.2 True-Up Events

At true-up events, field measurements from the resampling campaign are available, so the full cumulative-mass ESM correction (Equation 19) is applied to the model output, with measured SOC mass replaced by modeled SOC mass:

[math: \mathrm{SOC}_{j,t_n}^{\mathrm{model,ESM}} = \left[ \sum_{k=1}^{a} \hat{M}_{C,j,k,t_n} + \left( M_{\min,\mathrm{ref},j} - \sum_{k=1}^{a} M_{\min,j,k,t_n} \right) \frac{\hat{M}_{C,j,b,t_n}}{M_{\min,j,b,t_n}} \right] \times \frac{1}{100}]

(Equation 22)

Where:

• [math: \mathrm{SOC}_{j,t_n}^{\mathrm{model,ESM}}]​ is the ESM-corrected modelled SOC stock density in stratum [math: j] at true-up event [math: t_n]​, on a mineral-mass basis (t C ha⁻¹);
• [math: \hat{M}_{C,j,d,t_n} = \hat{C}_{j,d,t_n} \times M_{\mathrm{fine},j,d,t_n}]​​ is the modelled SOC mass per unit area for increment [math: d], formed from the modeled SOC content (g C kg⁻¹ fine soil, updated through the true-up procedure) and the measured fine soil mass at the true-up event (Equation 15, kg m⁻²);
• [math: M_{\min,j,d,t_n}] is the measured soil mineral mass per unit area for increment [math: d] at the true-up event (Equation 16, kg m⁻²);
• [math: M_{\min,\mathrm{ref},j}] is the reference soil mineral mass (Equation 18), the across-event minimum of the stratum-mean profile total over [math: t_0]​ and all true-up events including the current one;
• [math: a] and [math: b = a+1] are defined as in Equation 19;
• the factor [math: 1/100] converts g C m⁻² to t C ha⁻¹.

As in Approach 1, Equation 22 is evaluated at each sampling location against the stratum reference [math: M_{min,ref,j}]​, and the stratum-level value is the mean across locations. The cross-event minimization is applied consistently across all sampled increments using all sampling locations.

The cross-event minimization under the ESM correction is applied consistently across all sampled increments using all sampling locations.

[math: SOC^{model,ESM}*{j,t_n}] is substituted for [math: SOC*{j,t_n}] in Equation 14 for the purposes of calculating The Project-level total SOC stock at true-up events.

9.1.3.2.3 Model Divergence Testing and Reconciliation

With the true-up events under the model and remeasurement approach, the values across the two methods must be reconciled, and any model overprediction must be detected and corrected. These requirements have implications for all credit issuances under this approach, at both interim crediting events and true-up events.

Governing principle. Crediting under this Module and the Improved Soil Management Protocol is cumulative. At all times, the total credits issued to The Project must correspond to the cumulative increase in SOC stock established at the most recent true-up event, less credits issued in respect of periods before that event. Each true-up event re-establishes the cumulative SOC stock from field measurement and reconciles total credits issued to it, consistent with the cumulative crediting basis and the derivation of net removals under Equation 10 of the Improved Soil Management Protocol.

Divergence test. At each true-up event, the measured cumulative change in SOC stock since the most recent true-up event (or since t₀ for the first true-up event) must be compared against the modeled cumulative change over the same interval. The test statistic is the difference between the measured and modeled cumulative change, standardized by the combined standard error of the two quantities. The comparison must be conducted as two one-sided tests at the 90% confidence level, one for overprediction and one for underprediction, comparing the standardized difference against the corresponding one-sided critical value. The distributional basis for the test must be documented, and a small-sample correction (use of the t-distribution in place of the normal approximation) must be applied where sample sizes are low.

The combined standard error must be formed from the sampling uncertainty of the measurement and the model prediction uncertainty, treated as independent (combined in quadrature) unless a correlation between them is demonstrated:

• The measurement sampling uncertainty must be the same within-stratum sampling uncertainty characterized for propagation through the Section 9.1.1 Monte Carlo simulation, so that the divergence test and credit issuance use a consistent uncertainty basis.
• The model prediction uncertainty must be the uncertainty in the model's cumulative prediction over the interval since the most recent true-up event, propagated consistently with Section 9.1.1, including any temporal correlation of model errors across periods. The per-period model prediction error must not be used in place of the cumulative interval uncertainty, as this would understate the uncertainty against which the divergence is tested.

The outcome of the divergence test must be reported at every true-up event, whether or not a significant difference is found.

Establishing the corrected SOC stock. On the basis of the divergence test, the corrected cumulative SOC stock at the true-up event ([math: SOC^{corrected}_{j,t_n}]​) is established as follows, and used in place of the modeled stock in Equation 14 for the derivation of net removals:

• Overprediction (measurement significantly below model). The corrected stock is the lower of the modeled cumulative stock or the measured cumulative stock.
• Underprediction (measurement significantly above model). The corrected stock is the conservative lower bound of the measured cumulative stock, evaluated at the confidence level used for issuance under Section 9.1.1.
• No significant difference (conservative reconciliation applies). The corrected stock is the lower of the modeled cumulative stock or the conservative lower bound of the measured cumulative stock, evaluated at the confidence level used for issuance under Section 9.1.1.

In each case the corrected stock is established at the module level; the derivation of the corresponding cumulative change and net removals from [math: t=t0] is performed under Equations 3-5 of the Improved Soil Management Protocol.

Reconciliation of credits. Total credits issued to The Project must be reconciled to the corrected cumulative SOC stock (as given effect through the derivation of cumulative net removals under Equations 3 and 4 of the Improved Soil Management Protocol):

• Where total credits issued exceed the corrected position, the excess constitutes overcrediting. No credits are issued for the current reporting period, and the excess is compensated from the Buffer Pool, subordinate to any reversal compensation claims on the Buffer Pool in the same period. Any amount so compensated constitutes a replenishment obligation on The Project Proponent, which is senior to any further issuance: no credits may be issued to The Project until the Buffer Pool has been fully replenished by the compensated amount.
• Where the corrected position exceeds total credits issued, the shortfall is issued as credits attributed to the current reporting period.

Where credits conservatively forgone at a true-up event (through reconciliation to a conservative lower bound) are confirmed by a subsequent true-up event that establishes a higher corrected cumulative SOC stock, the difference is issued at that subsequent event through the reconciliation above. No separate tracking of forgone amounts is required, as the cumulative reconciliation gives effect to this automatically.

Detection roles. The divergence test is the sole trigger for reconciliation under this Section. The bias tests under Section 9.1.3.2.3 serve to diagnose the source and location of a divergence and to guide the forward response; they are recommended at each true-up event and become required where the divergence test identifies a significant difference. Bias in the counterfactual is addressed through model validation against baseline-like observations under Section 9.1.3.2.3 rather than through reconciliation under this Section, as The Project counterfactual is not directly observable.

Forward response. Where the divergence test identifies significant overprediction, reconciliation of the elapsed period under this Section is mandatory. For subsequent periods, The ProjectProponent must either recalibrate the model in accordance with Section 9.1.3.2.3 or apply update the model prediction uncertainty to correspond to the latest true up data and use this increased uncertainty value for all issuances until the next true-up event. In either case, modeled increments for subsequent periods must be applied to the corrected cumulative SOC stock established at the true-up event, and not to the model's uncorrected cumulative trajectory.

Escalation. The model loses eligibility for interim model-based crediting, and crediting reverts to the measure-and-remeasure procedure under Section 9.1.2 for the remainder of The Project, where either: (i) significant overprediction is identified at two consecutive true-up events; or (ii) the overcrediting excess at a single true-up event exceeds 20% of the credits issued in respect of the periods since the most recent true-up event. Eligibility for interim model-based crediting may be restored where the model is re-validated against the full requirements of Section 9.1.3.2.1 and 9.1.3.2.3 and re-approved by Isometric, including demonstration of performance under the conditions associated with the divergence. Following restoration, a single subsequent identification of significant overprediction re-suspends eligibility.

9.2 Assessment of Counterfactual Carbon Storage

9.2.1 Counterfactual Assessment via Measurement of Control Plots

Projects using a measure-remeasure quantification approach for carbon storage within The Project Area will determine the counterfactual carbon storage by applying the same measurement techniques within control plots. The control plots must be areas where project interventions are not taking place, but are otherwise representative of The Project Area. As such, these areas represent “business as usual” management practices that would have continued in absence of The Project intervention.

1. Candidate-set identification. At enrollment, participating landowners must identify the contiguous areas within their holdings that they are willing and able to maintain under business-as-usual management for the duration of The Project Commitment Period and to make available for sampling under the protocol. The aggregate candidate set across all participating landowners (the "Control Plot Candidate Set") must, in aggregate, span the soil, climate, and management strata represented in The Project Area. Where the Control Plot Candidate Set materially under-represents one or more strata of The Project Area, defined as fewer than three contiguous candidate areas within a stratum, or aggregate candidate area within a stratum less than 5% of the corresponding stratum in The Project Area, The Project Proponent must either (a) extend the candidate set to remedy the gap before validation, or (b) document the gap, justify why representativeness can nonetheless be reasonably assumed, and accept the consequences for uncertainty quantification under Section 9.1.1.
2. Stratified random selection. Control plots must then be drawn from the Control Plot Candidate Set using a stratified random sampling approach, with stratification reflecting the soil, climate, and management variability of The Project Area. Project Proponents must provide details of how the stratification and selection process was done, including the random seed and procedure used, and must document the relationship between the Control Plot Candidate Set and The Project Area (i.e., how the candidate set was constructed and any areas that were considered for inclusion but excluded, with reasons).

At minimum, the total area of the control plots must be equivalent to 2.5% of The Project Area, with a minimum of three control plots for each stratum within The Project Area. Management practices within the control plots must continue to follow "business as usual" practices, and Project Proponents must provide evidence and details on the implementation of these practices at each Reporting Period. At project initiation, the management practices must be evidenced with records from the enrolled properties over a minimum of three years or full crop rotation (whichever is longer) prior to project implementation. In order of preference, this evidence can include:

1. Farm management logs
2. Farm management plans
3. Affidavits from enrolled land owners

At least once every five years, The Project Proponent must demonstrate that the implemented practices within the control plots continue to represent regional management trends. This must be evidenced by regional data records from government bodies, academic or research institutions, international organizations, and/or peer-reviewed literature. In the event that there is a shift in these business-as-usual management practices, the management of the control plots must be updated to reflect these practices for the subsequent Reporting Period.

Where a control plot becomes unavailable during The Project Commitment Period for reasons outside The Project Proponent's reasonable control (e.g., land sale, change of operator, withdrawal of the participating landowner), a replacement control plot must be drawn from the Control Plot Candidate Set under the same stratified random sampling procedure, within the same stratum, and re-baselined at the next sampling event. Discontinuity in the control-plot time series introduced by such a replacement must be documented and propagated through the uncertainty framework under Section 9.1.1. Replacement of control plots for reasons within The Project Proponent's control (e.g., agronomic preference) is not permitted.

Project Proponents must follow the same requirements for sampling design, sample collection, sample analysis, and reporting as for quantification of carbon storage within The project area as described in Section 9.1.2 and all subsections therein. If control plots are directly adjacent to project intervention plots, sampling should occur at least 10 meters from the boundary. Project Proponents must follow the same sampling and analysis procedures for both Project Area and control samples. The total counterfactual storage at any time ([math: t]) is then calculated as:

[math: CO_2e_{counterfactual,t} = \frac{44}{12} \sum_{i=1}^{n} SOC_{i,t} A_i]

(Equation 23)

Where:

[math: CO_2e_{counterfactual,t}] is the total storage of carbon in the counterfactual scenario at time [math: t], in tonnes CO2e

[math: SOC_{i,t}] is the mean soil carbon density at time [math: t] in in the control plots corresponding to strata [math: i] of The Project Area of [math: n] total strata in the control plots calculated following the procedures in Section 9.1.2, in tonnes C ha^-1

[math: A_i] is the area of strata [math: i] in The Project Area, in ha

9.2.1.1 Assessing counterfactual uncertainty

Assessment of uncertainty in the value of [math: CO_2e_{counterfactual}] estimated via measure remeasure techniques must follow the same requirements and procedure for assessing uncertainty in the measure remeasure approach for Project Area carbon quantification.

9.2.2 Counterfactual Assessment via Modeling

For each parameter within the model related to management practices, Project Proponents must provide evidence of the historical parameter values based on records from the enrolled properties over a minimum of three years or full crop rotation (whichever is longer) prior to project implementation. In order of preference, this evidence can include:

1. Farm management logs
2. Farm management plans
3. Affidavits from enrolled land owners

In addition to property-specific data, The Project Proponent must also indicate for each variable where regional data records from government bodies, academic/research institutions, international organizations, and/or peer-reviewed literature exist.

For each Reporting Period, The Project Proponent must report any newly available data based on the regional data sources, as well as report availability of any new regional data sources. Where there is a difference between the recent regional parameter values and the historical farm parameter values, the more conservative (i.e. the one yielding a higher counterfactual discount) must be used for the prediction of [math: CO_2e_{counterfactual}]. If there is a lag between the collection and reporting of regional data, the historical data may be used for years with as of yet unreported data, but the values must be updated in subsequent Reporting Periods to reflect the most recent data availability.

In lieu of the above data sources, Project Proponents may elect to take a probabilistic modeled approach for assessing counterfactual management practices following the procedure and requirements in Section 9.2.2.1. The model used to estimate [math: CO_2e_{counterfactual}] must satisfy the eligibility and validation requirements of Sections 9.1.3.1 with respect to the baseline management practices being modeled, in addition to The Project management practices.

9.2.2.1 Development of Probabilistic Management Practice Model

Projects may elect to develop a statistical model for describing management practices in absence of The Project interventions within The Project area. This model must be developed and validated based on property- and region-specific data. The model must predict the probabilistic distribution of management practices within The Project area based on variables which are dynamically updated over the course of The Project. The Project Proponent should consider all potentially relevant factors to farm activity and management, including but not limited to:

• Climate variables
• Fertilizer prices
• Commodity prices
• Recent practices/management decisions

The model must be approved by Isometric prior to use, with this section outlining criteria which will form the basis of the assessment. If approved, the distributions of management practices generated by the model can then be used as input parameters for the biogeochemical model for the prediction of a distribution of [math: CO_2e_{counterfactual}] values. Following initial approval, the continued performance of the model at each Reporting Period based on recent data must be demonstrated for continued use. If the model still falls below the performance thresholds, The Project must revert to the default procedure for modeled counterfactual determination.

9.2.2.1.1 Model Development and Reporting

9.2.2.1.2 Model Validation and Updates Over Time

To be eligible for use, the performance of the model must be demonstrated at The Project and regional level using data that is withheld from the training process and both spatially and temporally blocked from the training data. The validation data must cover at least 80% of the range of each management parameter which is fed into the biogeochemical model. To be approved for use, the model skill for predicting the validation data must be demonstrated via:

• Root mean square error of less than 30% of the mean value for all continuous variables
• Classification accuracy of 75% for categorical variables
• Misclassifications of parameter values that would deflate the baseline (and thus decrease the counterfactual discount) must not exceed 15%
• Mean prediction error not significantly different from 0 at 90% confidence interval for The Project area overall and for each strata; If a bias is present it must be demonstrated that it results in a more conservative baseline assessment

The input parameters for the model are also subject to the sensitivity analysis (An analysis of how much different components in a Model contribute to the overall Uncertainty.) requirements under the Protocol as part of the broader uncertainty assessment.

For each Reporting Period, the continued model performance must be demonstrated against all newly available regional data since the prior Reporting Period. The model may be recalibrated, but all details must be reported and the recalibration must apply across the full modeled project time period, including retroactively

9.2.2.2 Assessing Modeled Counterfactual Uncertainty

The relative uncertainty of the modeled counterfactual carbon storage should use the same uncertainty assessed for the model using the validation against collected measurements.. Where Project Proponents elect to use the method described in Section 9.2.2.1 to generate probabilistic descriptions of management practice parameter values, the distributions of [math: CO_2e_{counterfactual}] generated as part of the procedure should be propagated through the broader assessment of uncertainty for The Project.

9.2.2.3 Measure-Remeasure Project with Modeled Counterfactual

Where this approach is used, The Project-scenario SOC stock change must be quantified by direct measurement under Section 9.1.2 and all its subsections, and the counterfactual SOC stock change must be modeled under Section 9.2.2, subject to the additional requirements of this Section. The model used for the counterfactual must satisfy the eligibility, validation, spatial-blocking, true-up, and (where applicable) multi-model-ensemble requirements of Section 9.2.2.3.4 with respect to the baseline management practices being modeled.

9.2.2.3.1 Loss of Error Cancellation

The quantification pairings derive the net project effect as the difference between two estimates established on a common basis. Measurement, seasonal, laboratory, and model-structural biases common to both estimates therefore substantially offset one another in the difference. This pairing does not achieve such offsetting: any systematic bias in the modeled counterfactual propagates without offset into the credited net effect. The requirements of this Section must be applied to ensure that residual bias in the modeled counterfactual acts only in the direction that reduces credit issuance.

The counterfactual term must be calculated from the same measured baseline SOC stock at project initiation (t₀), the same stratification, and the same reference mineral mass, so that the difference reflects only the divergence in SOC trajectory between The Project and counterfactual scenarios. This section should be applied in addition to Section 9.2.

9.2.2.3.2 One-Sided Counterfactual Bias Correction

The modeled counterfactual must be corrected for systematic bias in the conservative direction only. Where the bias tests in Section 9.1.3.1.6 detect that the model systematically predicts the counterfactual to lose more, or gain less, SOC than the baseline-like validation evidence supports, the modeled counterfactual stock change must be adjusted upward to remove the detected bias. Where the model predicts the counterfactual to gain more SOC than the validation evidence supports, no adjustment may be made; bias in this direction is conservative for crediting and must be retained.

9.2.2.3.3 One-Sided Uncertainty Treatment of the Counterfactual

The uncertainty discount framework of Section 9.1.3.1 applies to the measured project-scenario component as for Approach 1. The modeled counterfactual component must receive the 1.0 SD true-up discount. Counterfactual model uncertainty must be propagated through the Section 9.1.1 Monte Carlo simulation.

The conservative estimate used for credit issuance must correspond to the 30th percentile of the distribution of [math: \Delta \mathrm{SOC}_{\mathrm{net,RP}}] across all simulation iterations, in line with Section 9.1.1. Where the 30th-percentile estimate is negative, no credits may be issued for that Reporting Period.

9.2.2.3.4 Multi-Model Ensembles

10.0 Storage and Durability of CO2e Removals

10.1 Project Risk Assessment and Management

The Cropland Soil Carbon Risk Assessment is independently evaluated by a third-party VVB. The Cropland Soil Carbon Risk Assessment is used to determine the risk profile of The Project, including risks to Credit delivery and storage. Aspects of The Project which have higher risk exposure must be accompanied by an appropriate risk mitigation plan. To safeguard against high risk projects, The Project must score below the indicated thresholds to be eligible for crediting under this Module. The Cropland Soil Carbon Risk Assessment must be updated each Reporting Period by The Project Proponent and increased risk scores will result in additional mitigation activities.

10.2 Buffer Pool

10.3 Ongoing Monitoring for Reversals

10.3.1 Reversal Detection

If monitoring reveals a loss event representing a reduction of carbon stored in soil carbon stocks greater than 1% of the cumulative tonnes of CO₂e removed by The Project (based on total number of Credits issued), The Project Proponent must follow the requirements for reporting, investigating, and compensating for the Reversal set out in Section 10.4.3 of the Improved Soil Management Protocol (Buffer Pool Compensation Process).

If The Project Proponent is unable to conduct the required sampling within a property (e.g., no longer has requisite access to property), the property is considered to have experienced a full Reversal.

10.3.2 Reversal Quantification

Quantification of Reversals must be calculated via the same methods and procedures as are used for the quantification of carbon storage via the measure and re-measure approach (Section 9.1.2). Declines in the measured soil carbon stocks within The Project Area are conservatively assumed to represent carbon that was immediately released to the atmosphere. In the scenario that measurements are not able to be carried out in the relevant portion of The Project Area, the area will be considered to have experienced a full reversal of the cumulative carbon removed.

10.3.3 Remote Monitoring of Reversals (Alternative to Default Full-Reversal Treatment)

Where a portion of The Project Area is within an Ongoing Monitoring Period and The Project Proponent has lost contractual access to that portion such that direct field sampling under Section 9.1.2 cannot be conducted, the default treatment is full Reversal of the cumulative tonnes credited to the affected portion, drawn from the Buffer Pool.

As an alternative to that default, a Project Proponent may elect to apply the Remote Monitoring of Reversals procedure set out in this sub-section, subject to the eligibility conditions, monitoring requirements, and conservatism rules below. Election must be made in the Monitoring Report covering the Reporting Period in which the loss of access first occurred and is irrevocable for that portion of The Project Area until the procedure either succeeds through to the end of the Ongoing Monitoring Period or fails into the default treatment.

Each Remote Monitoring Report must be independently reviewed by The Project's VVB at the same frequency as The Project's primary verification cycle.

10.3.3.1 Eligibility

Remote Monitoring of Reversals is available only where all of the following conditions are met for the affected portion of The Project Area:

1. The Project intervention(s) implemented at the property are limited to practices whose continuity or reversion can be reliably detected from publicly or commercially available remote sensing data. Eligible interventions are: tillage regime change, cover-crop adoption, residue management, and crop-rotation change. Where project interventions are not directly surface detectable, remote sensing monitoring may be implemented if it can be demonstrated that i) carbon from prior project interventions remains stored in absence of identified disturbances, and ii) all identified disturbances can be monitored via remote sensing. For (i), it must be demonstrated with empirical data that storage is not expected to be impacted by non-detectable factors including irrigation changes, cessation of amendment application (if part of project interventions), and compaction management.
2. The Project Proponent has, at validation, set out a documented Remote Monitoring Plan for The Project that specifies the remote sensing data sources, indices, classification methods, spatial resolution, revisit frequency, accuracy thresholds, and ground-truth validation approach. The Remote Monitoring Plan must be in place at the time of loss of access; it cannot be retroactively constructed.
3. The portion of The Project Area to which the procedure is applied does not, in aggregate over The Project Commitment Period, exceed 10% of the original Project Area. Where loss of access exceeds this threshold, the default full-reversal treatment under Section 10.3.1 applies to the surplus.

10.3.3.2 Required Remote Monitoring

Remote Monitoring under this sub-section must satisfy each of the following requirements:

1. Spatial resolution. Sufficient to resolve sub-field management variability at the affected property. As a default, ≤ 10 m ground sample distance for optical / multispectral imagery; lower-resolution data may be permitted with documented justification.
2. Revisit frequency. Sufficient to reliably detect each eligible practice within its agronomic window. As a default, ≤ 14-day revisit during the active growing and post-harvest seasons.
3. Detection accuracy. The classifier(s) used for each practice must be validated against ground-truth data including, at minimum, The Project Proponent's own historical sampling campaign data, at an accuracy of ≥ 85% per practice, with documented confusion matrices and false-negative rates separately reported.
4. Independence and reproducibility. RS data sources, classification methods, and accuracy assessments must be documented in sufficient detail that an independent verifier can reproduce the classification outputs from the source data. Proprietary classifiers are permitted only where the verifier has independent access to the underlying data and to a documented specification of the classifier's logic.
5. Reporting frequency. A Remote Monitoring Report covering each affected portion of The Project Area must be submitted at every Reporting Period for the duration of the Ongoing Monitoring Period.

10.3.3.3 Triggers

The following events, detected through the Remote Monitoring procedure, trigger the immediate consequences set out below:

1. Practice reversion. Detection of reversion to baseline practice, including (but not limited to) tillage events on a property whose project intervention is no-till, removal of cover-crop establishment in two or more consecutive seasons on a property whose intervention is cover-cropping, or a return to the baseline crop rotation, triggers full Reversal of the cumulative tonnes previously credited to the affected portion of The Project Area, drawn from the Buffer Pool, and termination of the Remote Monitoring procedure for that portion.
2. Acute disturbance. Detection of fire, land-use conversion, or other acute disturbance events affecting the soil carbon pool triggers full Reversal of the cumulative tonnes previously credited to the affected portion of The Project Area, drawn from the Buffer Pool, and termination of the Remote Monitoring procedure for that portion.
3. Loss of remote sensing coverage. Where remote sensing data of sufficient quality (per Section 10.3.3.2) cannot be obtained for the affected portion for two consecutive Reporting Periods, the procedure terminates and full Reversal of the cumulative tonnes previously credited to the affected portion applies, drawn from the Buffer Pool.
4. End of Ongoing Monitoring Period. At the end of the Ongoing Monitoring Period, where Remote Monitoring has not previously triggered a full reversal, the procedure terminates without further accounting consequence.

11.0 Acknowledgements

Isometric would like to thank the following external contributors to this Module:

• Nuala Fitton PhD
• Benjamin Dube PhD
• Matthew Gammans PhD

12.0 Definitions and Acronyms

Independent components of Isometric Certified Protocols which are transferable between and applicable to different Protocols.

The amount of CO₂ emissions that would cause the same integrated radiative forcing or temperature change, over a given time horizon, as an emitted amount of GHG or a mixture of GHGs. One common metric of CO₂e is the 100-year Global Warming Potential.

The term used to represent the CO₂ taken out of the atmosphere as a result of a CDR process.

Soil Organic Carbon

A document that describes how to quantitatively assess the net amount of CO₂ removed by a process. To Isometric, a Protocol is specific to a Project Proponent's process and comprised of Modules representing the Carbon Fluxes involved in the CDR process. A Protocol measures the full carbon impact of a process against the Baseline of it not occurring.

An activity or process or group of activities or processes that alter the condition of a Baseline and leads to Removals or Reductions.

An assessment of what would have happened in the absence of a particular intervention – i.e., assuming the Baseline scenario.

A set of data describing pre-intervention or control conditions to be used as a reference scenario for comparison.

The diversity of life across taxonomic and spatial scales. Biodiversity can be measured within species (i.e. genetic diversity and variations in allele frequencies across populations), between species (i.e. the total number and abundance of species within and across defined regions), within ecosystems (i.e. the variation in functional diversity, such as guilds, life-history traits, and food-webs), and between ecosystems (variation in the services of abiotic and biotic communities across large, landscape-level scales) that support ecoregions and biomes.

Resources provided to projects that are generating, or are expected to generate, greenhouse gas (GHG) Emission Reductions or Removals.

The organization that develops and/or has overall legal ownership or control of a Removal or Reduction Project.

A worldwide federation (NGO) of national standards bodies from more than 160 countries, one from each member country.

Lowering future GHG releases from a specific entity.

A process for evaluating and confirming the net Removals and Reductions for a Project, using data and information collected from the Project and assessing conformity with the criteria set forth in the Isometric Standard and the Protocol by which it is governed. Verification must be completed by an Isometric approved third-party (VVB).

A systematic and independent process for evaluating the reasonableness of the assumptions, limitations and methods that support a Project and assessing whether the Project conforms to the criteria set forth in the Isometric Standard and the Protocol by which the Project is governed. Validation must be completed by an Isometric approved third-party (VVB).

The document, written by a Project Proponent, which records key characteristics of a Project and which forms the basis for Project Validation and evaluation in accordance with the relevant Certified Protocol. (Also known as “PDD”).

A product that has been cultivated, raised or harvested primarily for food, shelter, or natural fiber.

The defined temporal and geographical boundary of a Project.

The use of satellite, aircraft and terrestrial deployed sensors to detect and measure characteristics of the Earth's surface, as well as the spectral, spatial and temporal analysis of this data to estimate biomass and biomass change.

Any process or activity that releases a greenhouse gas, an aerosol, or a precursor of a greenhouse gas into the atmosphere.

A retirement pathway in which an existing EAC is retired to enable the issuance of a new EAC with different specified characteristics.

A publicly visible uniquely identifiable Credit Certificate Issued by a Registry that gives the owner of the Credit the right to account for one net metric tonne of Verified CO₂e Removal or Reduction. In the case of this Standard, the net tonne of CO₂e Removal or Reduction comes from a Project Validated against a Certified Protocol.

Describes the addition of carbon dioxide removed from the atmosphere to a reservoir, which serves as its ultimate destination. This is also referred to as “sequestration”.

The escape of CO₂ to the atmosphere after it has been stored, and after a Certificate has been Issued. A Reversal is classified as avoidable if a Project Proponent has influence or control over it and it likely could have been averted through application of reasonable risk mitigation measures. Any other Reversals will be classified as unavoidable.

Certificates are issued to the Certificate Account of a Project Proponent with whom Isometric has a Validated Protocol after an Order for Verification and Certificate Issuance services from a Buyer and once a Verified Removal or Reduction has taken place.

Third-party auditing organizations that are experts in their sector and used to determine if a project conforms to the rules, regulations, and standards set out by a governing body. A VVB must be approved by Isometric prior to conducting validation and verification.

The amount of time carbon removed from the atmosphere by an intervention – for example, a CDR project – is expected to reside in a given Reservoir, taking into account both physical risks and socioeconomic constructs (such as contracts) to protect the Reservoir in question.

An evaluation of the likelihood that an intervention that causes a climate benefit above and beyond what would have happened in a no-intervention Baseline scenario was the result of revenues from carbon finance.

GHG sources, sinks and reservoirs (SSRs) associated with the project boundary and included in the GHG Statement.

Those gaseous constituents of the atmosphere, both natural and anthropogenic (human-caused), that absorb and emit radiation at specific wavelengths within the spectrum of terrestrial radiation emitted by the Earth’s surface, by the atmosphere itself, and by clouds. This property causes the greenhouse effect, whereby heat is trapped in Earth’s atmosphere (CDR Primer, 2022).

Any process, activity, or mechanism that removes a greenhouse gas, a precursor to a greenhouse gas, or an aerosol from the atmosphere.

A location where carbon is stored. This can be via physical barriers (such as geological formations) or through partitioning based on chemical or biological processes (such as mineralization or photosynthesis).

Sources, Sinks and Reservoirs

The term used to describe greenhouse gas emissions to the atmosphere as a result of Project activities.

Emissions that are produced by a specific CDR process and are directly controllable.

​​Considering impacts at each stage of a product's life cycle, from the time natural resources are extracted from the ground and processed through each subsequent stage of manufacturing, transportation, product use, and ultimately, disposal.

A document submitted alongside Claimed Removals and/or Reductions that details the calculations associated with a Removal or Reduction, including the Project's emissions, Removals, Reductions and Leakages, presented together in net metric tonnes of CO₂e per Removal or Reduction.

Activities that remove carbon dioxide (CO₂) from the atmosphere and store it in products or geological, terrestrial, and oceanic Reservoirs. CDR includes the enhancement of biological or geochemical sinks and direct air capture (DAC) and storage, but excludes natural CO₂ uptake not directly caused by human intervention.

An analysis of the balance of positive and negative emissions associated with a certain process, which includes all of the flows of CO₂ and other GHGs, along with other environmental or social impacts of concern.

Life cycle GHG emissions associated with production of materials, transportation, and construction or other processes for goods or buildings.

A calculation, series of calculations or simulations that use input variables in order to generate values for variables of interest that are not directly measured.

Any person or entity who can potentially affect or be affected by Isometric or an individual Project activity.

Purposefully erring on the side of caution under conditions of Uncertainty by choosing input parameter values that will result in a lower net CO₂ Removal or GHG Reduction than if using the median input values. This is done to increase the likelihood that a given Removal or Reduction calculation is an underestimation rather than an overestimation.

United States Department of Agriculture

An acceptable difference between reported Removals/emissions or Reductions/emissions and what an auditor determines is the actual Removal/emissions or Reduction/emissions.

A lack of knowledge of the exact amount of CO₂ removed by a particular process, Uncertainty may be quantified using probability distributions, confidence intervals, or variance estimates.

A mathematical approach for estimating the possible outcomes of an uncertain event through repeated random sampling. It can also be referred to as a "multiple probability simulation".

A product that is not an economic driver of the process it is produced in.

An analysis of how much different components in a Model contribute to the overall Uncertainty.

13.0 Appendix A: Risk Assessment

The Cropland Soil Carbon Risk Assessment is used to assess the overall delivery and storage risk associated with the cropland management activities and may inform the Buffer Pool contribution during Credit delivery (see Section 10.2). The assessment must first be filled in by The Project Proponent with corresponding evidence supplied and must then be validated by a VVB. During project Validation, discrepancies between The Project Proponent’s self reported score and VVB may result in monitoring or risk mitigation activities, or project ineligibility. Eligible projects must have an initial risk score ≤ 20 and initial risk category scores at or below the following thresholds:

* Project Proponent Capacity Risk ≤ 7

* Financial Viability Risk ≤ 8

* Social Governance Risk ≤ 11

* Disturbance Risk ≤ 13

All risk categories have a minimum score of 0, regardless of the outcome of the Cropland Soil Carbon Risk Assessment.

For projects with discrete planting areas, the risk assessment must include all geographic areas relevant to The Project. For risk indicators that are geographically explicit (e.g., disturbance risks), the score may be calculated via an area-weighted average and rounded up to the nearest whole number.

If Project Proponents choose to forgo a flat 20% Buffer Pool contribution (see Section 10.2), the Cropland Soil Carbon Risk Assessment will inform Buffer Pool contributions for The Project according to the process outlined in Appendix A for each Reporting Period and in accordance with the requirements in Section 10.2.

After each new Cropland Soil Carbon Risk Assessment evaluation, Isometric will update the required percentage of newly issued Credits that must be contributed to the Buffer Pool by The Project. We encourage Project Proponents to continuously monitor, mitigate, and reduce risks.

Table A1. Cropland Soil Carbon Risk Assessment, with the score to be filled out for each question.

14.0 Appendix B: Calculating Extreme Weather Risk Scores

The following section outlines how Isometric calculates the indicator scores for climate-related extreme weather disturbance risks to carbon permanence within a project’s region (Figure B1). Project Proponents should use Table B1 to lookup the Isometric-calculated values for their project's region and include those scores in their Risk Assessment (see Appendix A).

Extreme weather risks are assessed via two indicators: a temperature indicator (extreme heat and/or cold) and a hydrologic indicator (flooding and/or drought). Each indicator includes both historical data (1961-2015) and projected future extreme events. Historical data indicate the likelihood of extreme events based on past climate patterns, e.g., projects in regions with extended dry periods are expected to experience increased water stress as part of their typical climate. Climate model projections describe how changing climate conditions, relative to historical patterns, might present an increased risk of disturbances. Areas where there is a larger shift towards extreme conditions under future climate relative to their historical baseline have a greater disturbance risk (e.g., drier conditions relative to historical averages increases risk for drought-driven mortality).

To calculate the scores in Table B1, Isometric uses values from the Intergovernmental Panel on Climate Change AR6 report¹⁴. The temperature indicator is calculated using data describing the annual number of frost days (FD, minimum temperature below 0°C) and annual number of days with a maximum temperature above 40°C ([math: TX_{40}]) to capture extreme cold and extreme heat risks, respectively. The hydrologic indicator is calculated using data describing the maximum 5-day precipitation ([math: RX_{5Day}]) and annual maximum number of consecutive dry days (CDD) to capture risks of flooding and drought, respectively. All values come from the CMIP6 climate models. Future projections use the SSP2-4.5 medium term (2041-2060) scenario and are assessed as the change in value relative to a historical baseline (1961-1990).

For each indicator subcomponent, the region’s terrestrial median value is compared with the global terrestrial distribution of the same variable (Table B2). To convert the regional value into a subscore, regional values below the global 50th percentile are considered Low Risk, regional values equal to or greater than the global 50th percentile but below the 75th percentile are Medium Risk, and any regional values equal to or greater than the global 75th percentile are High Risk. Low Risks are given a subscore of 0, Medium Risks are 0.25, and High Risks are 0.5. The overall score for each of the indicators is calculated by summing the corresponding subscores, as described below:

[math: Indicator_{Temperature} = Historical_{TX_{40}} + Future_{TX_{40}} + Historical_{FD}]

(Equation B1)

[math: Indicator_{Hydrological} = Historical_{CDD} + Future_{CDD} + Historical_{RX_{5Day}} + Future_{RX_{5Day}}]

(Equation B2)

Projected change in the number of frost days is not included as a subcomponent since it is projected that they will decline under future climate across the globe, representing a low risk of future extreme cold events.

Figure B1. Map and lookup table for IPCC regional codes

[Image: image]

Table B1. Regional Lookup Table of Disturbance Risk

Table B2. Global Benchmark Values for Extreme Weather Risks.

15.0 Appendix C: Buffer Pool Calculations

By default, Projects are subject to a flat 20% Buffer Pool contribution as outlined in Section 10.3. Project Proponents may opt to calculate a project-specific Buffer Pool contribution based on the outputs of their Cropland Soil Carbon Risk Assessment for each Reporting Period.

The following steps are used to convert the outputs of the Cropland Soil Carbon Risk Assessment into a Buffer Pool contribution:

1. Sum the total score for each risk category in Table A1.
2. Map the risk score for each risk category into a Buffer Pool contribution using Table C1.
3. Sum the Buffer Pool contribution for each risk category to obtain the total Buffer Pool contribution.

Table C1. Risk score to Buffer Pool contribution conversion for each risk category.

The Buffer Pool contribution for each risk category is determined using a sigmoid function described by Equation C1. The Buffer Pool contribution for each risk category ranges from 2.5% to 10%.

[math: BP_{risk} = \frac{L}{1 + e^{-k(x - x_0)}} + 2.5]

(Equation C1)

Where:

• [math: BP_{risk}] is the Buffer Pool contribution for a given risk category.
• [math: L] = 7.5 is the range of Buffer Pool contributions within each risk category (2.5-10%)
• [math: k = \frac{10}{cutoff-1}] is the steepness parameter of the sigmoid curve and determines how quickly the function transitions between its minimum and maximum values.
• [math: x_0 = \frac{cutoff}{2}] is the midpoint of the sigmoid curve.
• [math: x] is the risk score for a given risk category
• [math: cutoff] is the value above which The Project fails the Soil Carbon Risk Assessment, noted in Appendix A.

Regardless of whether The Project is taking the flat contribution or risk assessment-adjusted contribution, the additional contribution related to contract coverage (see Protocol Section 10.4.1) still applies in addition to the flat or risk-adjusted buffer pool contribution.

15.1 Project-Specific Buffer Pool Contribution Example

The Project has completed the Cropland Soil Carbon Risk Assessment and obtained the following risk scores in a Reporting Period:

• Project Proponent Capacity Risk = 2
• Financial Viability Risk = 4
• Social Governance Risk = 3
• Disturbance Risk = 3

Mapping these risk scores to Table C1, the total Buffer Pool contribution for The Project is:

3.1% + 6.3% + 3.1% + 2.9% = 15.4%

If The Project did not have contracts in place for all enrolled properties for the full duration of the Crediting Period, the total would be 20.4%.

16.0 Appendix D: Market Leakage Parameters

Isometric has carried out a literature review of [math: ε_s] and [math: ε_d] values to inform [math: IS], as well as values for [math: NL] for certain regions. Where The Project falls into these regions, the default values provided must be used. This is because understanding which values to use from literature is challenging as academic papers are typically not written with this purpose or audience in mind. Isometric has completed this work for certain regions to lessen this complexity and provide consistency across projects.

These default values also serve as an example of appropriate values to select, however it should be noted that the quality of research differs across regions.

The following sections set out the procedure to be followed to obtain [math: IS] and [math: NL] values and set out the default values to be used for the regions studied.

16.1 Regions Studied

The regions considered in the literature review were:

• Brazil
• Panama
• Mexico
• United States

These regions were selected following a review of projected project demand. Isometric will update this analysis with additional regions iteratively based on demand. Values for other regions will be reviewed by Isometric on a case by case basis.

16.2 [math: IS] Values

[math: IS] represents the amount of production that is diverted to other locations. The [math: IS] value does not provide any information on where or in what manner that production is produced.

Procedure for determining [math: IS] values:

1. Define a crop-region pair broadly enough to reasonably assume that the supply of all other crop and regions is zero. For example, for livestock, “beef in Mato Grasso” is too specific and “all meat globally” is too broad, “beef in South America” may come closer to a balance.
2. Examine the existing academic literature for papers that estimate the supply/demand of the crop-region pair or some other similar pair. For example, we might use “meat in South America” in lieu of “beef in South America” if the former estimates are available.
3. Ensure that the paper meets the criteria for a valid estimate:

• Analysis uses either a “dynamic panel” or “instrumental variables” technique.
• Based on time-series variation in prices, rather than cross-sectional (i.e., using prices that are varying over time, rather than spatial differences in transportation costs).
• Published in the last 15 years in a reputable economics or land use journal, or is published in a report for a reputable organization such as the European Union or California Air Resources Board.
• Analysis clearly notes whether the estimate is to be interpreted as a short-run or long-run estimate.
• Analysis uses planting season prices (rather than harvest season).

4. Use the appropriate formula to calculate [math: IS] from the supply and demand elasticity
5. In cases where no academic literature exists in the relevant context, select the most similar available default value (e.g., for cocoa, we might use the default value for coffee).

Where possible:

*

• Supply and demand elasticities used should be estimated within the same paper; and

  *

• Paper should be cited by a reputable organization compiling a meta-analysis or parameterizing a partial equilibrium model for policy analysis (CARB, EU, FAO, etc.).

Table D1.[math: IS] default values.

16.3 [math: NL] Values

Procedure for determining [math: NL] values:

In an ideal world, there would be estimates of the specific types of land use that were converted and their locations. However, this data is not available. Instead, The Project Proponent should focus on the most important elements of potential land use change from a carbon emissions perspective. [math: NL] values proposed aim to capture the net effect of a one unit removal of crop area on forestland conversion. These [math: NL] values will be smaller in magnitude than [math: NL] values that incorporate the possibility of conversion of grazing land or the conversion of lower-value crops to higher-value crops. Focusing on forests is more tractable and likely provides a large share of the relevant land use change emissions, since forest conversion is relatively permanent in a way that livestock to cropland conversion is not. In general, the [math: NL] values are more speculative than the [math: IS] values and often rely on assumptions about the yield-price elasticity that have not been empirically confirmed.

Two possible methodologies for obtaining [math: NL] values are set out here. Method B is in most cases the preferred approach. This is because the necessary conditions to implement Method A (limited trade/ disconnected markets and demand driven quantity increase) are rarely met in practice. Method A should only be used in special cases and justified appropriately. Both methods are set out below:

* Method A: In cases where a large increase in deforestation has accompanied a large increase in cropland, the ratio of land deforested for agriculture to total new agriculture is taken. Note, this procedure is only accurate for cases where (1) the deforestation followed a large demand-driven increase in production and (2) where the land is not well-connected to international markets. This approach is not reflected in the default values, as it is not an acceptable methodology for the majority of crop-regions.

* Method B: In most cases, such as the US, analyses of large changes in land use due to a policy shock is relied upon, and then the ratio of the percentage change in agricultural land to the percentage change in production is taken. This way of calculating [math: NL] is represented in the following definition:

[math: NL = \frac{{Gross\: new\: production\:}_x}{{Gross\: new\: production\:}_x +\ {\Delta \ Average\: yield \:}_x\: +\  Total\: land \:area}]

(Equation D1)

Where:

• [math: Gross\:new\:production] is Gross New Production From Extensification due to [math: x].
• [math: \Delta \ Average\: yield] is Change in regional average yield due to [math: x].
• [math: Total \:land \:area] is Total land area under study.
• [math: x] is a 1-tonne reduction in supply, or a 1-unit price increase.

[math: x] is variable under the assumption that changes to supply are predominantly channeled through price changes¹⁹.

By dividing the numerator and denominator, the above equation can be reformulated as:

[math: NL = \frac{{\Delta \ Area\:}_x}{{\Delta \ Area\:}_x +\ {\Delta \ Yield \:}_x\:}]

(Equation D2)

Where:

• [math: \Delta Area] is the change in area due to [math: x].
• [math: \Delta Yield] is the change in yield due to [math: x].
• [math: x] is a 1-tonne reduction in supply, or a 1-unit price increase.

The following default values have been gathered using Method B.

Table D2.[math: NL] default values.

17.0 Appendix E: MDD Power Analysis for Sample Design

The power analysis applies at the primary quantification unit level, not stratum-by-stratum. Where a stratified design is used, the total number of samples should be allocated across strata to meet The Project-level MDD using optimal (Neyman) allocation based on stratum area and within-stratum variance, or an equivalent allocation rule documented in the Project Monitoring Plan.

Where ancillary variables (e.g., remote-sensed indices, terrain attributes, digital soil maps) are used either to inform the design or to support a regression or model-assisted estimator at The Project level, the sample size calculation may incorporate the variance reduction expected from those covariates, provided the correlation has been quantified using project-area or comparable regional data and is reported transparently.

The MDD-based power analysis below uses the SD of location-level paired SOC stock changes (the differences between [math: t\_0]​ and the most recent sampling event at each location), because this directly reflects the noise in what The Project is trying to detect. Cross-sectional SD in absolute SOC stocks is not used and is generally a poor predictor of paired-difference variance.

The minimum number of samples needed to detect a given MDD at The Project level is calculated as:

[math: MDD = \frac{S}{\sqrt{n}} \times (t_{\alpha,\upsilon} + t_{\beta,\upsilon})]

(Equation E1)

[math: n \geq \left(\frac{S \times (t_{\alpha} + t_{\beta})}{MDD}\right)^2]

(Equation E2)

Where:

[math: MDD] is the minimum detectable difference in SOC stocks (t C ha[math: ^{-1}]);

[math: S] is the standard deviation of the SOC stock change at fixed sampling locations, i.e., the location-level differences [math: \Delta SOC_{loc} = SOC_{loc,t_n} - SOC_{loc,t_0}] computed at each composite sampling location, pooled across strata weighted by stratum area (t C ha[math: ^{-1}]). "Pooled" here refers to the statistical aggregation of within-stratum SDs into a project-level SD for the purposes of this power analysis, and does not imply any physical pooling of samples.

[math: n] is the minimum number of samples required;

[math: \upsilon = n - 1] is the degrees of freedom;

[math: t_{\alpha}] is the two-sided critical value of the t-distribution at significance level [math: \alpha]. [math: \alpha] should not exceed 0.05, meaning the sampling design should control the probability of falsely concluding a SOC change has occurred when none has, to no more than 5%.

[math: t_{\beta}] is the one-sided quantile of the t-distribution corresponding to the probability of a Type II error [math: \beta]. [math: \beta] should not exceed 0.10, meaning the sampling design should achieve at least 90% statistical power to detect a true SOC change of magnitude [math: MDD] when one is present.

The within-stratum standard deviation [math: S] used in the design-stage power analysis must be estimated using the most direct evidence reasonably available for The Project area, in the following order of preference:

1. Project-area pre-sampling. A pre-sampling campaign within The Project area, designed to characterize within-stratum SOC variability at the spatial scale at which the full sampling campaign will be conducted. Pre-sampling is the preferred source under all conditions and is the default where database- or literature-derived variance is otherwise the only available source (see paragraph below).
2. Empirical variance estimates from peer-reviewed literature. Within-stratum or within-field SOC variance values reported in peer-reviewed studies conducted in The Project's ecoregion, on comparable soil types, and at comparable spatial scales. Where a range of values is reported across studies, the upper end of the range applicable to The Project area should be used.
3. Gridded predictive products. Variance estimates derived from gridded SOC prediction products (e.g., SoilGrids, SSURGO, or comparable national or regional digital soil maps) may be used only subject to the following conditions:

• The Project Proponent should explicitly document that the variance reported by the product reflects between-pixel prediction uncertainty conditional on the product's covariate structure, and does not, on its own, capture within-pixel local SOC heterogeneity at the spatial scale at which field samples will be drawn.
• The variance estimate used in the power analysis should be inflated to account for the within-pixel heterogeneity component. The inflation factor should be documented and justified using either (i) regional empirical studies of within-field or within-pixel SOC variability for comparable soil and management contexts, or (ii) a default inflation of the standard deviation by a factor of [math: \sqrt{2}] (i.e., a doubling of the variance) where (i) is unavailable. Any inflation factor below [math: \sqrt{2}] on the standard deviation should be justified at validation against published evidence specific to The Project's ecoregion.
• Where the gridded product reports prediction uncertainty at a coarser spatial resolution than The Project's quantification units, the variance estimate should be additionally inflated, or rejected as a source, to reflect the resolution mismatch. Where the product does not report any estimate of prediction uncertainty, it may not be used as a variance source.

Where database- or literature-derived variance is the only design-stage source available, The Project Proponent should conduct a pilot pre-sampling round of at least 10 sampling locations per stratum, drawn under the same probabilistic design as will be used for the full sampling campaign, prior to finalizing the sample size and committing to the first sampling campaign. The variance observed in the pilot replaces the database- or literature-derived estimate in the power analysis. The pilot samples are not credit-relevant and may be drawn at reduced analytical cost (e.g., proximal sensing where validated under Section 9.1.2.9, or single-increment composited samples).

The initial power analysis remains provisional and should be updated at each subsequent re-sampling event using observed within-stratum variance from the preceding Reporting Period. Where the observed variance is materially higher than was assumed at the design stage, the consequences for crediting precision are absorbed through the uncertainty discount at credit issuance under Section 9.1.1; where the observed variance is materially higher than the design-stage power analysis can support at the proponent's chosen MDD, the sample size should be increased for the next Reporting Period.

18.0 Relevant Works