Ocean Accounting for Nature-based Solutions

TG-2.4 Environmental (including Ecosystem) Goods and Services Section 3.10.5 prohibits valuing an ecosystem asset at restoration cost, on the grounds that doing so “would make it impossible to separately assess the benefits of ecosystem restoration: they would by definition equal the costs.” The Technical Guidance therefore protects the separation between what an intervention cost and what it achieved, and nowhere sets out how to measure the second. This Circular does.

Its unit is the intervention. TG-6.11 Coastal Infrastructure Accounting records the structures built on a coast and values them as produced assets. The question here is different: an agency has decided to spend public or blended money on an ecosystem, in the expectation of a return. What must the accounting system record, over what horizon, against what counterfactual, and what must it publish so the decision can be defended, compared against an engineered alternative, and audited later?

Upon completing this guidance, practitioners will be able to:

1.1 Decision Use Cases

Coastal protection portfolio choice; national adaptation planning and Nationally Determined Contribution reporting; Kunming-Montreal Global Biodiversity Framework Targets 2 and 3; blue bond and blue carbon due diligence; public investment appraisal and budget classification; disaster risk reduction programming; and the design of offsets and compensation under environmental impact assessment.

TG-1.1 National Ocean Budgets, TG-1.4 Sustainable Ocean Planning, TG-1.6 Environmental Impact Assessment, TG-1.8 Project-Level Finance, TG-2.6 Ocean-related Investment and TG-2.9 Disaster Risk Indicators.

This Circular assumes the foundational concepts in TG-0.1 General Introduction, the asset framework in TG-3.1 Asset Accounts, and the ecosystem services framework in TG-2.4 Environmental (including Ecosystem) Goods and Services.

An intervention register recording site, area, date, technique, cost, implementing entity and land or sea tenure; the pre-intervention state of the ecosystem; a counterfactual or control; a time series of extent and condition; survival and failure monitoring; expenditure records split between capital and recurrent; service-flow indicators; and a mapping of beneficiaries.

The IUCN categories of NbS approach (Section 3.1.3) and the IUCN Global Standard criteria (Section 3.1.4); the Classification of Environmental Protection Activities (CEPA) and the Classification of Resource Management Activities (CReMA); the Classification of the Functions of Government (COFOG), for which the primary-purpose rule distinguishing coastal protection under Group 05.6 from transport infrastructure under Division 04 is set out at TG-6.11 footnote 67 and is not restated here; the IUCN Global Ecosystem Typology; and the restoration, rehabilitation and reclamation distinctions at TG-6.11 Section 3.4.2.

Readiness and sequencing. Whether a country is positioned to compile this account at all — mandate, data architecture, capacity and the order in which to build — is assessed under TG-0.8 Implementation Readiness Assessment, and is not restated here.

Concepts stated in TG-6.11 are cross-referenced, not summarised, so that each concept has one statement across the two Circulars.

3.1 What Counts as a Nature-based Solution

The first multilaterally agreed definition is United Nations Environment Assembly resolution 5/5 (2022), which defines nature-based solutions as:

“actions to protect, conserve, restore, sustainably use and manage natural or modified terrestrial, freshwater, coastal and marine ecosystems which address social, economic and environmental challenges effectively and adaptively, while simultaneously providing human well-being, ecosystem services, resilience and biodiversity benefits.”

The IUCN Global Standard (2020), operationalising IUCN Resolution WCC-2016-Res-069, defines them as “actions to protect, sustainably manage and restore natural or modified ecosystems that address societal challenges effectively and adaptively, simultaneously providing human well-being and biodiversity benefits.”

3.1.2 Distinguishing NbS from adjacent categories

TG-6.11 Section 1 defines grey, green and blue infrastructure, and that vocabulary is inherited here without restatement. TG-6.11 Section 3.4.2 sets out the restoration, rehabilitation and reclamation distinctions, which turn on whether the objective is pre-existing structure and function, functionality for service supply, or a useful state. Table 3.1.2 records only the distinctions those sources do not carry.

Table 3.1.2: Categories adjacent to NbS, and their accounting consequence

3.1.3 A typology for accounting

Two axes are used, and an ownership mapping. The typology below is adopted, not devised here.

The primary axis is the five categories of NbS approach set out by Cohen-Shacham et al. (2016), which group interventions by their primary ecological approach. Each category carries a distinct accounting consequence, which is what earns the axis its place here.

Table 3.1.3: Categories of NbS approach, and their accounting consequence

The second axis is intensity of ecosystem management, which predicts which entry points at Section 3.2 an intervention will touch: Type 1 conservation and monitoring, Type 2 restoration and rehabilitation, and Type 3 creation and engineering. The two axes are not redundant, since a single IUCN category can contain interventions of differing intensity.

The ownership mapping is a lookup, not a typology. It records which thematic Circular owns the ecosystem an intervention acts on: shoreline protection to TG-6.11 and TG-6.2 Mangrove and Coastal Wetland Accounting; habitat and fisheries to TG-6.2, TG-6.3 Seagrass and TG-6.7 Fisheries Accounting; carbon to TG-2.8; land-sea interaction to TG-6.13; productive systems to TG-3.9 Aquaculture Accounts and TG-6.8 Aquaculture Thematic Methods; and nature-positive built development to TG-6.9 Offshore Energy, TG-6.10 Maritime Transport and Ports and TG-6.11.

3.1.4 The eligibility screen

Eligibility turns on one question of mechanism.

Does an ecosystem process provide the primary functional mechanism addressing the challenge?

Wetland restoration for flood regulation passes, because water absorption and flow attenuation are ecosystem processes. Ecotourism in a wetland does not, because the primary mechanism is a business model and the ecosystem is the attraction. The distinction has a direct accounting consequence: an activity that fails the test is not routed through the ledger at Section 3.2. Its expenditure is recorded, whilst no attributable ecosystem enhancement arises from it, and treating it as an NbS would overstate the intervention’s ecosystem effect.

The SEEA EA defines ecosystem enhancement as “the increase in the value of an ecosystem asset over an accounting period that is associated with an improvement in the condition of the asset during that accounting period.” Enhancement is therefore measured through the change in the net present value of expected service flows, and not as the expenditure incurred. The SEEA EA is explicit that there should be “no prior expectations regarding the results of such a comparison” between the two.

The rule for attributing change is established, and is applied here without restatement. Extent accounts attribute every addition and reduction by cause, distinguishing managed change arising from direct human activity from unmanaged change arising from natural processes. Condition change is attributed on the same basis, and where causality cannot be reliably attributed compilers should not attempt to disaggregate the change: the total should be recorded without attribution and the limitation documented in account metadata.

For an intervention, the managed share is the quantity of interest, and the counterfactual machinery at Section 3.3 is what separates it from ambient change.

Type 1 conservation interventions and ecosystem protection approaches produce avoided loss rather than empirical gain. This is the hardest case in the ledger, because a successful intervention may leave the observed state unchanged: the benefit is the difference between the observed state and a decline that did not occur.

Avoided degradation is a recordable outcome, and it requires a modelled counterfactual, not an observed one. The method is at Section 3.3.4, since it depends entirely on the counterfactual construction and cannot be separated from it.

3.3 Baselines, Counterfactuals and Additionality

Counterfactual design is established in the Technical Guidance and is not re-derived here. TG-1.3 Section 3.4.3 specifies control-site selection, before-after-control-impact (BACI) designs, matching-based approaches and regression discontinuity designs, together with the instruction that where no design is feasible compilers should report descriptive condition trends with an explicit caveat that causal attribution is not supported. This section covers only what is specific to an intervention.

3.3.1 The restoration counterfactual is harder than the protection counterfactual

An MPA counterfactual can draw on comparable unprotected sites, which usually exist. A restoration counterfactual requires a comparable degraded and unrestored site, and such sites are systematically scarce for a structural reason: sites are selected for restoration precisely because they are degraded, and often because they are the most degraded available. The pool of candidate controls is therefore biased by the same variable the design is trying to control for.

Three practical responses are available. A staged rollout across sites creates a natural experiment, where sites scheduled for later intervention serve as controls for earlier ones. Matched degraded sites can be drawn from a wider region, at the cost of weaker environmental comparability. A restoration-gradient design uses variation in intervention intensity across a single site, in place of a treated-untreated contrast.

Tiers follow the graduated convention used throughout the Technical Guidance, under which Tier 1 represents the minimum viable approach and higher tiers require greater data or capacity.

Table 3.3.2: Counterfactual tiers for intervention accounting

The tier is not a presentational choice. It determines the width of the uncertainty band on the attributable change, and a Tier 1 account should not report an attributable change as a point estimate.

During the years in which a benefit is still maturing, the intervention effect and ambient recovery are confounded. A site that would have partially recovered without intervention will show gains that are not wholly attributable, and the maturation profile at Section 3.4 means the intervention’s own contribution is at its smallest in exactly the period when monitoring is most intensive.

Compilers should not report an attributable change over the establishment period alone. Where the accounting period falls inside that window, the change should be reported with the establishment status stated, and the counterfactual extended over the full maturation horizon before an attributable total is drawn.

3.3.4 Restoration against protection, and the two additionality tests

Whether a benefit is verifiable at all turns on which limb the intervention sits in.

Restoration produces an empirical net gain. The intervention is measured at a later period and differenced against the baseline, so both terms are observed.

Both differ from additionality as used in crediting. TG-1.8 Section 3.4 records that crediting standards differ in their requirements for baseline setting, additionality demonstration, permanence buffers and leakage assessment. Crediting additionality asks whether an activity would have happened without the revenue from credits, and its consequence is eligibility. Accounting additionality asks whether an observed change is attributable to the intervention, and its consequence is which account entry the change belongs in. An intervention can be attributable and not creditable, or creditable and only weakly attributable. The two tests should never be conflated, and a compiler asked for one should not supply the other.

3.4 Time, Lag and Survival

3.4.1 Survival, and what the evidence supports

The largest synthesis of marine coastal restoration cost and feasibility covers 235 studies and 955 observations across coral reef, seagrass, mangrove, saltmarsh and oyster reef restoration, and records average survival of restored organisms alongside cost and project duration. Median average survival by ecosystem, computed from that dataset, is set out in Table 3.4.1.

Table 3.4.1: Median reported survival of restored organisms, by ecosystem

Three cautions apply to any use of these figures.

Coverage is uneven. Survival is reported for between 8 and 47 per cent of observations depending on ecosystem, and the coral figure rests on 17 observations against 286 records. The coral literature separately reports transplant and post-transplant survival, better populated at 77 and 87 observations with medians of 45 and 70 per cent, but these measure different stages and must not be pooled with each other or with the headline figure. Compilers using a coral survival rate should state which stage it refers to.

The distributions are wide. Reported survival spans zero to 100 per cent in every ecosystem, so a median is a central tendency across heterogeneous techniques, sites and monitoring periods, and not a design expectation for any particular project.

Survival is not permanence. These figures describe the survival of restored organisms over project monitoring periods, whose reported median duration is one to three years. They do not describe persistence over an asset life, which is the subject of Section 3.5.

3.4.2 Recording expected rather than design benefit

The quantity the monetary ecosystem asset account should carry is the expected delivered benefit, not the design capacity. Where a site-specific survival estimate exists it should be used. Where none exists, an ecosystem-level rate from Table 3.4.1 may be applied as a Tier 1 estimate, with the source and the coverage caution documented.

$B_{expected}=B_{capacity}\times p_{survival}\times m\left(t\right)$Bexpected​=Bcapacity​×psurvival​×m(t)

where $p_{survival}$psurvival​ is the probability of persistence and $m\left(t\right)$m(t) is the maturation factor at time $t$t, taking values between zero and one. A compiler applying no survival adjustment is implicitly setting $p_{survival}$psurvival​ to one, which is not a neutral choice and overstates the asset.

Compilers should therefore treat the maturation profile as a documented assumption, not an evidenced parameter. Ecosystem-specific reviews give indicative establishment periods for coral, seagrass and mangrove systems and should be cited where used. Where no defensible profile is available, the account should report the benefit as not yet established and should not interpolate one. The shortfall against design capacity belongs in the account, not absorbed into it.

3.4.4 Discounting, and the accounting period

Discount rate selection is governed by TG-1.9 Section 3.1.2, including its treatment of social against market rates, its fallback where no published social discount rate exists, and its requirement for sensitivity analysis across a plausible range. No further guidance is offered here, and compilers should not adopt a different rate for NbS interventions than for other ecosystem assets in the same account.

A benefit arriving in year 12 does not enter the current period’s service flow account. What enters is the change in the asset’s net present value arising from the intervention, which reflects the expected future flow. The distinction matters where an intervention is completed and revalued in one period whilst delivering nothing measurable in it.

3.5 Permanence, Reversal and Risk

Restored and protected ecosystems can be lost. Tropical cyclones, marine heatwaves, disease outbreaks, dieback and encroachment all remove delivered benefit, sometimes in a single event, and the risk is not symmetric across the asset life: a young intervention is more vulnerable than a mature one.

The scale a compiler should plan for is not marginal. The 2015 dieback along the Gulf of Carpentaria killed of the order of 7,400 hectares of mangrove, some 6 per cent of the Gulf’s mangrove extent, synchronously across roughly 2,000 kilometres of coastline, in conditions of extreme temperature, drought and a temporary decline in sea level associated with the 2015 to 2016 El Nino. An event of that kind will exceed the extent of any restoration programme in the same region, which is why reversal is treated here as a recording problem for the ecosystem asset account and not as a project-level contingency.

3.5.2 Recording a reversal

The recording rule is established and is applied here. Change attributable to human activity is recorded as degradation, whilst change arising from natural processes is recorded in the other changes in volume account. A cyclone destroying a restored mangrove is therefore an unmanaged reduction, and failure of the planting itself is not: the first is a volume change, the second a shortfall in the enhancement that was recorded.

The attribution carries a practical consequence. An account that records all reversal as degradation will misattribute climate-driven loss to management failure, and an account that records all reversal as an unmanaged volume change will conceal establishment failure. Where the cause cannot be reliably attributed, the total should be recorded without attribution and the limitation documented.

Catastrophic loss follows the general framework for unforeseen asset value changes in the national accounts. This is the ecosystem-side counterpart of the stranded produced asset treated at TG-6.11 Section 3.5.1, and neither Circular repeats the other’s mechanics.

3.5.3 Risk adjustment, and what the account publishes instead

Buffers are close to universal in NbS crediting, and are correct practice there. Verra’s Verified Carbon Standard requires the AFOLU Non-Permanence Risk Tool, and VM0033, the methodology for tidal wetland and seagrass restoration, requires the current version of that tool including for erosion and submergence risk from sea level rise. Withholding percentages are set in the tool and adjusted periodically, and a project whose assessed risk exceeds the highest available percentage is ineligible for crediting. Other standards operate comparable regimes, with some now admitting alternatives such as long-term conservation easements, transfer of reversal liability to a permanence trust, and cancellation of credits of negligible reversal risk.

The ecosystem asset account therefore records the unadjusted value. In place of a buffer, the compiler publishes the parameters a user needs to construct their own adjustment. For each intervention and each service:

1. the persistence or survival probability applied, or an explicit statement that none was;
2. the maturation profile assumed, and its basis;
3. the counterfactual tier under Section 3.3.2, with its uncertainty band;
4. reversal events in the period, with the managed against unmanaged attribution at Section 3.5.2;
5. the discount rate and the sensitivity range, per TG-1.9 Section 3.1.2;
6. the valuation method and its preference level under TG-2.4 Section 3.10.2, service by service.

A user holding those six can construct a buffered figure, a risk-adjusted figure or an insurance-priced figure. A user given only a buffered total can construct nothing. One compilation then serves the national accountant, the crediting user and an insurer, without the account taking a position that belongs to none of them.

The requirement sits inside existing practice, not beside it. TG-0.7 Quality Assurance Principles already requires scope, definitions, methods and assumptions to be documented, and supplies quality flags for account tables. Where a risk-adjusted presentation is wanted for policy purposes, it belongs in a complementary presentation, following the treatment of welfare values in TG-1.9.

3.6 Full Lifecycle Cost

3.6.1 Cost over an asset life

An NbS cost is not its establishment cost. Recurrent maintenance, monitoring and adaptive management continue over decades, and the restoration cost synthesis distinguishes capital costs from operating costs covering maintenance, monitoring, and equipment repair and replacement. Compilers should assemble the cost stream over the intended asset life, and where only capital cost is available should say so, and should not present it as a lifecycle figure.

Compilation cost is itself material, and constrains how many interventions a jurisdiction can account for and at what frequency. Australia’s blue carbon programme committed AUD 30.6 million over 2021 to 2025 across on-ground restoration, project-level environmental-economic accounting and knowledge exchange, and awarded separate measurement and accounting contracts for five restoration sites. Per-site compilation cost is not published, so compilers should budget from their own procurement and should record the cost of compilation alongside the cost of the intervention, since a jurisdiction that cannot afford to account annually will be reporting on a longer cycle than its interventions change.

The produced-asset side of a hybrid intervention is costed under TG-6.11 Sections 3.2.1 and 3.2.2, using the perpetual inventory method and its depreciation treatment.

3.6.2 Opportunity cost of space

An intervention occupies land or sea space that had, or could have had, another use. Foregone aquaculture production, foregone development value and lost access are costs of the intervention even where no transaction records them. Where tenure is contested the opportunity cost may be borne by parties who are not the implementing entity, which is a distributional question as well as a costing one.

3.6.3 Non-market and unpaid stewardship labour

Community monitoring, patrolling and locally managed enforcement are frequently the mechanism by which an intervention persists, and are frequently unpaid. The restoration cost synthesis records in-kind volunteer contributions and disaggregates funding across government, non-governmental organisation, private and volunteer sources, which provides a basis for estimating the share of intervention effort that sits outside market transactions.

Treatment follows TG-3.5 Social Accounts for unpaid labour, TG-2.3 Social and Livelihood Dependencies for livelihood dependency, and TG-4.4 Citizen Science and Community-Based Monitoring for community monitoring as a data source. An account that omits this labour understates the cost of the intervention and, by omitting the contributors, misrepresents who produced the benefit.

3.6.4 Transaction, safeguard and precondition costs

Free, prior and informed consent processes, safeguard design and grievance mechanisms carry real cost, treated under TG-3.6 Traditional Knowledge. So does the legal and tenure work that determines which interventions are feasible at all: tenure clarity frequently decides which intervention categories can proceed at a given site, and areas of contested tenure are commonly documented and excluded, not modelled.

3.6.5 Why capital-cost comparisons mislead

Claims that nature-based options are cheaper than engineered ones often compare establishment cost against construction cost. That comparison omits the recurrent and monitoring cost of the nature-based option, the opportunity cost of the space it occupies, the unpaid labour sustaining it, and the survival probability discounting its delivered benefit. It also omits the maintenance cost of the engineered option, which cuts the other way. A lifecycle comparison on equivalent service units, as at Section 3.8, is the only form in which the question can be honestly put.

3.7 Multiple Benefits Without Double Counting

3.7.1 The benefit bundle

The protective service is the largest single item in several published assessments, which is relevant to Section 3.8: mangrove flood protection benefits have been estimated to exceed USD 65 billion annually, with some 15 million more people flooded each year in their absence, and coral reefs have been estimated to avert flood damages of the order of USD 4 billion annually, with expected damages roughly doubling in their absence. These are global estimates produced for advocacy and prioritisation, not account entries, and TG-1.9’s preference order governs whether any of them may enter an account.

3.7.2 The recording protocol

Valuation method selection and compatibility are governed by TG-2.4 Section 3.10 and no separate treatment is offered here. Three of its rules bear directly on interventions. Replacement cost sits at preference level 5 in most ocean applications, because the replacement structure is hypothetical. Restoration costs must not be used to value an ecosystem asset. Stated preference estimates of total willingness to pay are not appropriate direct account entries. Values derived from methods with different uncertainty characteristics should not be summed without documentation and sensitivity analysis.

3.7.4 Two cases the existing rules do not reach

Upstream intervention, downstream benefit. Where a catchment intervention improves the condition of a downstream marine ecosystem, the intervention and the benefit sit in different ecosystem types whose accounting units do not share a boundary. The service is recorded against the ecosystem that supplies it, which is the downstream unit, whilst the expenditure is recorded against the intervening entity. TG-6.13 governs the attribution, and the metadata must state it explicitly, since the alternative is either a service with no expenditure or an expenditure with no service.

Services are attributed to beneficiaries through the supply and use framework, and the distribution matters independently of the total. An intervention whose protective service accrues to a port authority and whose foregone access falls on a fishing community has a different distributional profile from one where both fall on the same population, even at identical net value. TG-3.5 and TG-2.3 govern the treatment.

3.8 Comparing Nature-based, Engineered and Hybrid Options

3.8.1 Physical service first

TG-1.9 Section 3.2.1 recommends that physical accounts always be compiled and published, with monetary valuation considered as a complement and not a replacement. TG-2.4 Section 3.6 states the same, requiring physical accounts as the primary output. A comparison expressed only as cost per unit therefore inverts the publication’s own ordering, because it makes money the comparator.

The comparison is accordingly layered.

Layer 1, primary. The physical service each option delivers, both expressed in the same stated unit. For coastal protection this is a measure such as wave-height reduction at a stated return period. Both options are measured in the same unit, so this is the only fully commensurable layer.

Layer 2, complementary. Cost per unit of that physical service over the asset life, drawn from the produced-asset cost stream under TG-6.11 for the engineered option and from Sections 3.4 to 3.6 for the nature-based option, each with an uncertainty range.

Layer 3, not commensurable. Benefits that cannot enter the comparison, treated at Section 3.8.3.

The service unit must be stated on the face of the comparison. A comparison that resolves cleanly should be treated as a signal that the unit has been defined too loosely, since the published evidence is conditional rather than general: saltmarsh and mangrove defences have been found to be materially cheaper than a submerged breakwater for wave heights up to about half a metre, and to become more cost-effective at greater water depths, which is a result bounded by wave height and depth, not a general ranking.

3.8.2 Building the comparison from account entries

The comparison uses figures the accounts already contain. From TG-6.11, the engineered option contributes its capital cost, depreciation profile and expected service life. From this Circular, the nature-based option contributes lifecycle cost under Section 3.6, survival-adjusted expected benefit under Section 3.4, and the counterfactual tier under Section 3.3.2, which sets the width of its range.

The nature-based range will usually be wider, because establishment risk and counterfactual uncertainty compound. That width is a result and should be reported. Where the ranges overlap, the correct statement is that neither option is established as cheaper per unit of service, not that the option with the lower central estimate wins.

3.8.3 What the comparison cannot carry

Most of the nature-based case sits outside Layers 1 and 2. TG-1.9 Section 3.2.3 records that monetary values from accounts do not capture all relevant considerations for decision-making, naming cultural values, equity, rights and biodiversity intrinsic value. Biodiversity is, under Section 3.1.1, a necessary condition of NbS status, so the single most defining benefit of an NbS is one the commensurable comparison structurally cannot express.

The remedy is not to monetise it. Several of these benefits cannot enter an exchange-value account at all, and the methods that would force them in are the ones TG-2.4 Section 3.10.5 rules out. Layer 3 is therefore recorded in its own accounts, in physical or qualitative terms, with the measurement basis stated for each benefit, and is never summed into the cost-per-unit comparison. Summing across bases would mix exchange values with physical and qualitative measures and produce a total that means nothing.

3.8.4 Presenting the comparison

A budget process requires the comparison in the form set out at TG-1.1, and an appraisal in the form set out at TG-1.6. In both cases the presentation should carry all three layers, the service unit, and the counterfactual tier. A comparison presented without its tier invites the reader to treat a Tier 1 estimate as a Tier 3 one.

3.9 Financing and Disclosure Interface

3.9.1 What the accounts supply

The instruments themselves are treated at TG-1.8 and TG-2.6. What a financing or insurance counterparty requires from the accounts is narrower: the physical service measure and its condition basis, the attributable change and its counterfactual tier, the lifecycle cost stream, and the six metadata items at Section 3.5.3. An account that supplies these supports due diligence without the compiler taking a view on the transaction.

3.9.2 Disclosure frameworks ask different questions

Alignment with the Taskforce on Nature-related Financial Disclosures and with IFRS S1 and S2 is treated at TG-1.11 and TG-6.11. Compilers should note that the wider standards stack around an intervention, including the IUCN Global Standard, the SEEA EA, taxonomy screening criteria and national green finance taxonomies, asks different questions of the same intervention. Meeting one does not constitute meeting another, and an account should not be described as taxonomy-aligned on the strength of SEEA EA compliance.

3.9.3 Credit stacking and the account

TG-1.8 Section 3.4 raises the stacking of carbon credits with payments for other services. The accounting risk is narrower and should be stated plainly: where a service is recorded as enhancement in a national ecosystem asset account and simultaneously sold as a credit, the same benefit supports two claims. The account should record the existence and coverage of any credit issuance over the accounting area, so that a user can identify the overlap. This is a metadata requirement, not a netting rule: the account does not deduct sold credits from recorded enhancement.

3.9.4 What the accounts cannot supply

Accounts establish the quantum. Whether it can be transacted is determined by legal and institutional preconditions that sit outside them, among them the definition of carbon rights, the legal enablement of payments for ecosystem services, operational protocols for free, prior and informed consent, and tenure clarity. A Circular presenting accounts as sufficient for financing would mislead, and compilers asked whether an intervention is bankable should answer that the accounts inform the question without settling it.

Step 1: Build the intervention register. Record site, area, date of completion, technique, implementing entity, funding sources, and land or sea tenure status. One row per intervention, with a stable identifier that later periods reuse.

Step 2: Screen eligibility and classify. Apply the instrumental test at Section 3.1.4 and record the determination with its reasoning. Classify by IUCN category and by intensity type (Section 3.1.3), and record the benefit bundle expected (Section 3.7.1). Record the IUCN Global Standard assessment, where one exists, as quality metadata.

Step 3: Establish the baseline and select the counterfactual tier. Compile the pre-intervention state. Select the highest tier the data support under Section 3.3.2 and record it, together with the control or model used. Where Tier 1 is the ceiling, record that attributable change is not established.

Step 4: Assemble lifecycle cost. Capital, recurrent, monitoring and adaptive management over the intended asset life; opportunity cost of space; non-market and unpaid stewardship labour; transaction, safeguard and precondition costs (Section 3.6).

Step 5: Measure attributable extent and condition change. Compile extent and condition following the owning thematic Circular’s methods, and difference against the counterfactual, not against the baseline alone. Attribute managed and unmanaged change under Section 3.2.3. A compiler holding a two-period extent account, plot-based condition data and a protection-status stratification has what this step requires.

Step 6: Apply survival and maturation, then value. Adjust design capacity to expected delivered benefit under Section 3.4.2, documenting the survival source and the maturation assumption. Value service flows under TG-2.4 Section 3.10, recording the method and preference level for each.

Step 7: Route the entries and run the double-counting check. Post to each applicable entry point in Table 3.2.1. Confirm against Section 3.7.2 that each service is recorded once, and record the routing, including any staggered or upstream-downstream attribution under Section 3.7.4.

Step 8: Assemble the combined presentation. Bring the expenditure, produced asset, ecosystem asset, service flow and social entries into a single presentation following TG-3.8 Combined Presentations, with the account-tier boundary preserved.

Step 9: Quality assurance and documentation. Publish the six metadata items at Section 3.5.3. Apply the quality flags and uncertainty categories at TG-0.7, noting that ecosystem dynamics uncertainty and valuation uncertainty will usually dominate.

The illustration that follows is stylised. Figures are constructed to exercise the distinctions this Circular draws and are not drawn from any jurisdiction. Empirical application belongs in the Section 5 country case studies.

Three interventions in a single coastal accounting area, over a five-year accounting period.

Intervention A, Type 2 restoration. 400 ha of mangrove replanting for coastal protection, carbon and nursery habitat. Baseline and matched control sites available, so Tier 2.

Intervention B, Type 1 protection. 1,200 ha of existing mangrove brought under area-based protection, producing avoided loss against a modelled counterfactual of continued conversion. Tier 2, model-based.

Intervention C, Type 3 creation. A constructed wetland in the catchment, with engineered inlet works, reducing sediment load to a downstream seagrass meadow.

Table 3.11.1: Intervention register extract

Three features of the routing are worth drawing out.

Intervention C splits across the account-tier boundary. The inlet works are a produced asset under TG-6.11, at cost less depreciation. The wetland’s ecosystem entries are recorded under this Circular. The two are never summed. The service, sediment reduction benefiting the seagrass meadow, is recorded against the downstream seagrass unit under TG-6.13, whilst the expenditure is recorded against the intervening entity, per Section 3.7.4.

Intervention A’s benefit is not fully delivered in the period. With a five-year period and a mangrove maturation profile extending beyond it, the account records the change in asset net present value arising from the intervention, with the maturation assumption documented, and does not record design capacity as delivered service.

Table 3.11.2: Reversal in year 4

A cyclone in year 4 removes an estimated 30 per cent of Intervention A’s replanted area.

Had the loss instead arisen from planting in unsuitable substrate, the reduction would be a shortfall in the enhancement previously recorded, not an unmanaged volume change, and the earlier enhancement entry would be revised.

3.12 Summary and Cross-References