The map we hold is ISPRA 2017. Newer data exists. None of it is monitored locally, or quickly.
The layer every surface here renders is the EMODnet seabed habitat polygon sourced from ISPRA, surveyed 29 November 2017. On that extent Pollica's mapped meadow is about 4,879 ha.
ISPRA has surveyed since. We do not hold those maps, and nothing newer is reachable in the open layers we can pull. So the honest statement is not that 2017 is the newest survey in existence. It is that 2017 is the newest survey this coast can put its hands on, and that even a fresher federal map would arrive on somebody else's schedule.
Professor Scardi's point is the operational one: a professor's calendar is the binding constraint, so the cadence a coast needs has to be run by the coast. A survey that depends on a distant specialist happens when the specialist is free, which is roughly never in the season you need it.
Habitat layer: EMODnet Seabed Habitats, ISPRA source polygons, EUNIS A5.535. Mapped area from the same extent, measured in Tessera.
Cheap enough for a small boat.
Simple enough for a working morning.
Held by whom, in five years.
A temperature, light and position package small enough to mount on a working fishing boat costs less than a single professional survey day. We state that as a comparison rather than a number because our own hardware bill of materials is still pending, and a placeholder that reaches a careful reader costs more than a gap does.
The method is the harder half. It has to be sound enough that a marine scientist will accept the output and simple enough that a fisherman can run it before the morning's work starts. That is a design constraint, and it is where most community-science kits fail.
Ownership is the half almost nobody writes down. Who holds the raw readings, who signs them, who archives them, and who can produce the 2029 record in 2034 without calling the consultant who has since moved on. Left unanswered, the programme ends when its funding does.
Ecosystem programmes are usually built from the top. A coast can build the same capability from the bottom.
Top down. An originator assembles the capital, the method and the verification, then places a working programme on a coast and runs it to a standard. It moves fast, it carries real expertise, and it is how most restoration at scale has actually happened. The mangrove programme in Madagascar is that done properly.
Bottom up. The coast holds the instrument and the record. The method is taught rather than delivered, the measurement is produced by the people whose sea it is, and the municipality takes custody of what comes out. It is slower to stand up and it does not leave when the grant does.
The two are the same thesis run from opposite ends, and they need each other. A top-down programme with no local measurement capability is verifying itself. A bottom-up programme with no capital and no standard is a hobby. The Mediterranean question we are putting in front of you is whether a comune can run the loop itself, and what it takes to find out.
A boat runs an assigned line on a day it was already working.
1. Plan. The comune sets the window. Lines are drawn against last season's coverage, prioritising ground nobody reached.
2. Survey. Boats run assigned lines across the season, on days they were already working. Coverage accumulates rather than arriving in one campaign.
3. Sign off. A director reviews each run and signs it, or sends it back. Doubtful readings are down-weighted, never silently discarded.
4. Publish. The season's sessions are versioned and frozen. Once published, a record cannot be quietly altered.
5. Archive. The comune takes custody as the official record, and can produce it on request without asking us.
6. Compare. This season against last, with uncertainty stated. A cell that changed because the weather was worse is distinguishable from one where the meadow moved.
7. Act. The comune routes what the comparison found: an anchoring proposal, a restoration priority, a practice change, a budget line.
Steps 3, 4 and 5 are the ones with no tool behind them yet. See A1. Walkable at /pollica/cycle and /pollica/operating-model.
A cell outside the 2017 extent is drawn as disagreement, not discarded.
the surfaceBlank means nobody went there. It never means fine.
Every cell carries three separate facts and they are kept apart on purpose. Covered: a boat ran this cell this season, with a timestamp, a vessel, a route and an operator behind the reading. Known: what the 2017 extent says is here, drawn as a prior rather than as the truth. Disagreement: where the two do not match.
The obvious way to build the screen is to test each reading against the 2017 polygon and drop whatever falls outside it. That treats the old map as ground truth, and until last week the code did exactly that. It was deleting the evidence the survey went out to find, which made it the worst bug in the system. A reading beyond the mapped extent now stays, dated and attributed, rendered as a disagreement between two sources.
We are not claiming the 2017 map is wrong. We are claiming it is old, and that a community can now check it.
The readings driving the surface today are synthetic. The full state of what is real and what is not is in A4, without the softening.
| Who | Holds |
|---|---|
| Community and fleet | runs the survey, holds raw data |
| The comune | archives the record, decides |
| Island Lab | method and quality control |
| Sign-off | nobody formally holds it yet |
We name that last row ourselves, before due diligence does.
The full matrix runs eleven functions across community and fleet, comune, Island Lab, FFI and an external verifier: set the survey schedule, plan the lines, run the survey, sign off a run, hold the raw data, archive the record, method and tooling and quality control, teach the method, convene and broker, verify a credit claim, decide and act.
The unresolved row is the load-bearing one. The premise is that an untrained operator produces defensible data because a named reviewer checks and signs the run. That role carries authority in the data schema and does not exist in the data model today: the operator is currently a hardcoded provenance string. Until a named person with a named permission sits in that row, the quality claim rests on nothing, and it is the first thing a verifier will pull on.
The decision is whether the director is a comune employee, an Island Lab role or a university secondment. Each answer changes who is liable when a run is wrong, which makes it a governance question before it is a technical one. It is one of the things we would like your view on.
Full matrix: /pollica/operating-model.
Protection costs 38 to 58 EUR per hectare-year. Restoration costs 516k to 2.5M.
A community that measures its own sea is what still protects it in 2050.
Passive conservation costs 38.10 to 57.80 EUR per hectare per year. Active restoration is quoted two ways in the same report and we print both rather than averaging them: 516k to 942k EUR/ha (Bayraktarov et al. 2016, explicitly at 2010 prices) and 2.5M EUR/ha for a 2024 Balearic project, which is the recent Mediterranean figure. Protecting a hectare for a year costs roughly one forty-thousandth of restoring one.
Which is why the order matters. If you do not protect what is there, more restoration does not help. Restoration comes after, on its own costed line, funded by capital or grant rather than by a credit that has not happened to anyone yet.
A 2024 French court ruling valued Posidonia ecosystem services at 86,000 EUR per hectare per year, reported via WWF MMI 2025 in ARTEMIS. A court has already used that number to price harm, which is a different kind of evidence from a valuation study. It is a gross ecosystem-service value: it is not capturable revenue and it is not a credit price.
Mediterranean Posidonia conservation needs 336M EUR a year and receives 17M, a 319M shortfall. Received funding is 5% of need. Worldwide, only eight seagrass projects have ever had private-sector funding, five of them in Europe and the Mediterranean, about 15.6M EUR in total. There is no crowded field to be late to. There is also no established playbook, and both halves are true.
The claim that community stewardship is the durable ecological outcome has a Mediterranean precedent, not just a theory. At Torre Guaceto, under fisher co-management of a protected area, catch per unit effort rose 2 to 5 times, with a Slow Food Presidium price premium averaging about +81%. Fishers complied because it paid them to. That is the mechanism this model depends on, and it is the reason the social layer is not a co-benefit bolted onto the ecological one: it is what makes the protection hold after the funding stops.
Four lines are pending, not estimated: hardware bill of materials, platform running cost, survey day rate, and training. We would rather show you a gap than a placeholder, and we will have them before due diligence.
All figures from ARTEMIS, Investment Opportunities in Seagrass Restoration, October 2025, read in full and traced to primary sources. the report
SeaForest is the Italian track, and we are verifying it. Bas-Carbone is the French precedent.
An EU LIFE project, 2018 to December 2025, run by ISPRA with CNR, D.R.E.Am. Italia and Carbonsink. It built a standardised Posidonia blue-carbon methodology and a national Italian voluntary credit platform, and measured accumulation of 7.13 to 217 gC/m2/yr. One of its three study sites was the National Park of Cilento, Vallo di Diano e Alburni, which is Pollica's own park. ISPRA is also the source of the 2017 extent we render.
What we do not know. Whether that platform is live today, and whether any credit has ever issued through it, is not established. We are verifying it and we are not asserting it. ISPRA
France approved "Protection des herbiers de posidonie" on 21 March 2023, the only approved Posidonia method anywhere. Its lever is avoided anchoring damage, which is the protection-first position written into a standard. Method paper: Comte et al. 2024. It is a French national scheme, so for an Italian site it is a precedent rather than a route. Bas-Carbone
VM0033 v2.1 for carbon alongside the SD VISta Nature Framework for biodiversity, which opened to all projects on 1 January 2026 after a pilot-only period since its 2024 launch. It stacks on the same hectare with Verra's own VCS programme, which is what VM0033 is, and not with other carbon standards. VM0033 Nature Framework
The first three methodologies were adopted on 10 July 2026 and all three are terrestrial. Seagrass is in scope with no methodology yet. This is the track an obligation would eventually run through, and it is the reason the record has to be built to a standard before there is a standard to build it to. CRCF adoption
No Posidonia carbon credit has ever issued, anywhere. Every published blue-carbon price is a mangrove price, which your Madagascar programme means you know better than most rooms would. ARTEMIS projects credits at 34,033 EUR/ha in 2035 and 41,221 EUR/ha by 2050, and those are modelled projections at those dates rather than prices anyone has paid.
A second coast is a config swap, not a rebuild.
Neya originates top-down. Island Lab is the bottom-up instance of the same thesis, in the Mediterranean.
The South Cilento Masterplan covers thirteen coastal municipalities along one stretch of water, with the same habitat, the same kind of fleet and the same basemap problem. The region configuration already travels: a coast is a JSON file with its own geometry, its own fleet, its own record and its own provenance declarations, running the same seven steps. That is engineering that exists rather than a promise about a second phase.
What does not travel is the local arrangement. Every coast has its own authority, its own fleet and its own reason to care, and the sign-off question has to be answered again each time. Carrying a method means each coast ends up with a capability of its own. Replicating a model means exporting a programme that belongs somewhere else.
That is where the two directions meet. Origination brings the capital and the standard. Bottom-up brings a measurement the standard can trust, produced by people who are not paid by the outcome. For an obligation that has to survive an auditor, the independence is the asset.
What is built and what is configured: /product-map.
The specific list of what you need to see.
Early September, in your due diligence.
Island Lab · sean@islandlab.dev · v0.1
The four pending cost lines with real numbers behind them: hardware bill of materials, platform running cost, survey day rate, and what the teaching layer costs to run.
A named answer on sign-off, which is the governance decision on slide 6, and the vehicle question that goes with it. Island Lab is a Washington State LLC founded by Sean Kolk and Kate Schnippering, which is a foreign entity to an Italian municipal programme. Whether funds could flow to it or whether an Italian or EU vehicle is needed is better answered before September than during it.
A verified position on the SeaForest platform, and the first readings that came off the water rather than out of a generator.
Everything in this deck is live and open: the protocol deck, the operating model, the comparison surface, the coast as currently known.
| Step | Tool | State |
|---|---|---|
| Plan | Portolan | mvp |
| Survey | Periplus | experimental |
| Sign off | none | absent |
| Publish | none | absent |
| Archive | none | absent |
| Compare | Tessera | partial |
| Act | Kairos | partial |
Three of the seven have no tool yet. They are the build.
Portolan plans the lines against last season's coverage. MVP. Periplus runs them: it models a working day honestly, with quality degrading with speed, depth and sea state, coverage accumulating, fuel and time draining. It persists nothing, so a run currently evaporates on reset. That is the active build.
Sign off, publish and archive have no tool at all. Persistence exists as a written migration that has never been applied, so nothing today makes a record immutable or transfers custody to anyone. Those three steps are the custody half of the argument, and they are the half that is not built.
Tessera is the per-cell comparison against the known extent, in build now. Kairos routes what the comparison found to whoever can act on it. Monitoring nobody acts on is a hobby, and that is the layer that stops it being one.
Statuses come from the platform registry rather than from ambition: /product-map.
L1 Instrument, built. The fleet carries three sensors: water temperature, light and GPS position. That is the whole physical capability today.
L2 Habitat sensing, absent. Side-scan or single-beam returns are what turn a temperature-and-position survey into a habitat survey. Designed, costed, and not built. There is no sonar ingest anywhere in the system and no log reader for any sonar format. Every depth currently shown is either an open bathymetry grid or a synthetic distance-to-coast stand-in, and is labelled as such at the point of display.
L3 Telemetry and ingest, built. Device to cloud under a versioned contract, so a field change cannot silently break a consumer: columns first, mapping second, app version third.
L4 Trust and provenance, built. Every reading carries where, when and how well it was taken, combined into one score. Weak readings are down-weighted rather than deleted, a lone weak reading is flagged as needing replication, and a bad reading collapses toward zero instead of poisoning the average. Provenance is an explicit vocabulary, so a modelled number can never be shown as a measured one.
L5 Survey execution, partial. Portolan plans, Periplus runs, and nothing persists. L6 Map and comparison, partial. Tessera, with the comparison panel in progress. L7 Model, partial. The Posidonia model is a single-box model, not a per-cell simulation, so the honest interim product is a per-cell surface of observations rather than predictions. L8 Decision, partial. Kairos and the project console.
L9 Sovereign custody, absent. Publication, immutability and municipal archive. A published session that cannot be silently altered is what separates a record from a report. The persistence migration is written and has never been applied, so nothing today makes a record immutable or transfers custody. This is the layer the whole ownership argument rests on, and it is the one that is not there.
| Already there | Holds |
|---|---|
| Comune di Pollica | mandate, prior dive surveys, decisions |
| The national park | management plans for this coast |
| Paideia Campus | the teaching layer, run by FFI |
| The fleet | boats on this water most days |
None of these has been approached about this investment.
The comune is the local authority and the decision point, and it already holds prior dive surveys. The protocol adds a dated series to a record it already keeps, and step 5 hands custody to it rather than to us.
The Parco Nazionale del Cilento, Vallo di Diano e Alburni is the management frame for this coast, a UNESCO MAB reserve, and it was one of the three SeaForest LIFE study sites. Posidonia is an EU priority habitat under the Habitats Directive, so monitoring here is a legal obligation looking for an affordable method rather than a new idea looking for a mandate. The Santa Maria di Castellabate marine protected area sits immediately north, ending at Punta Licosa, and protects the same habitat.
The Paideia Campus is FFI's flagship in Pollica and the existing teaching vehicle. The Save the Ocean bootcamp is where this work already sits, which is why the method is taught rather than supplied. That layer is what makes it survive past the grant.
The fleet is a small-scale inshore fishery at Pollica and Acciaroli, and the meadow is the nursery that sustains the catch. A survey day is a second job on a day the boat was already out, and the boat gets a map of its own ground.
The standing constraint. The mayors and the public administration have not been consulted about this thread, by agreement on sequencing. Everything above describes what exists, not what has been agreed.
Written down and running.
Synthetic data through a real pipeline.
Nothing here came off the water.
Temperature, light and position today.
The protocol is designed, written down and implemented end to end. The seven steps are real, with the three custody steps standing on tooling that does not exist yet, per A1 and A2.
Every reading on every surface in this deck is synthetic data driving a real pipeline. The instruments have not reported since 8 July 2026, and the first real outing is the next stage of the work. Each surface says so at the point of display rather than in a footnote, and this slide is where it is said in full.
Sonar is designed and costed and not built. It is the difference between confirming the meadow's condition and confirming its extent, so today we can say the water was this warm at this position on this date, and we cannot say the meadow's edge has moved. There is no per-cell simulation either: the model is a single-box model, so what the map shows is observations rather than predictions.
None of the above is a reason to discount the protocol. It is the reason the first milestone is a real survey day with a signed run at the end of it, and it is the thing we would rather you hear from us than find.
| Open | Why it matters |
|---|---|
| Who signs a run | The quality claim rests on it. |
| Our four cost lines | Pending, not estimated. |
| SeaForest platform | Verifying whether it is live. |
| US LLC, Italian programme | A vehicle question. |
Sign-off. A comune employee, an Island Lab role or a university secondment. Each answer moves liability somewhere different. It closes with a named person, a named permission in the data model, and a signature on a real run. It is the first thing a verifier pulls on and the first thing we want settled.
Cost. Hardware bill of materials, platform running cost per year at Pollica's scale, survey day rate, and the teaching layer. The day rate is the hinge: it is what makes a fisherman a steward rather than a volunteer, and there is no number for it in our own repository. It closes with a quote and a conversation with a boat.
SeaForest. Whether the national Italian voluntary platform is live and whether any credit has issued through it. It closes with ISPRA, and until it does we describe it as a methodology with a Cilento study site and nothing more.
The vehicle. A Washington State LLC is foreign to an Italian municipal programme that runs under public procedure. Whether funds can flow to it or whether an Italian or EU entity is needed is a due-diligence question, and we would rather raise it than be asked it cold.
The 2026 season is the proof, and it has not happened. Custody is the load-bearing layer and it is unbuilt. No Posidonia credit has ever issued, so any carbon revenue is a hypothesis. And two founders is a small team against a September decision, which is a real constraint even though the pitch is method and partnership rather than headcount.