Descent: Surface to Twilight
Close your eyes for a second, or don't.
Picture the surface. Light breaking apart into blue. The sound of the world above already fading.
You're going under.
Descent: Twilight to Abyssal
Down here, sound moves differently. Farther. Slower. Truer.
A whale is calling to another whale hundreds of miles away, and the water carries it there whole.
Something clicks, sharp and quick, then goes quiet. A current shifts, and the whole reef seems to lean with it.
You are not the loudest thing in this water. You are not even close.
The light is nearly gone now. What's left comes from the water itself.
The Turn: Abyssal Floor
Almost none of this has ever been seen.
Less than 0.001 percent of the deep ocean floor has been visually observed by anyone, ever. An area smaller than Rhode Island, in a space that covers most of the planet.
We have mapped less than a third of our own seafloor at any real resolution. We have not finished mapping the planet we live on.
This is not a knowledge gap. It is a blind spot, at the scale of a planet.
Atlantis Pads puts people and instruments in the water that isn't being watched. Research residencies. ROV operations. Data, where there was only guessing.
Ascent 1: Abyssal to Deep
The fish stocks that once fed whole coastlines are collapsing, quietly, one closed season at a time.
Warming water is moving faster than the species that depend on staying cold.
None of this waits for a five-year plan.
Atlantis Pads grows what the ocean can still carry. Seaweed, shellfish, species chosen not just because they survive here, but because growing them cleans the water and gives the reef something to rebuild on.
Ascent 2: Deep to Mid
The reefs that took thousands of years to build are dying in decades.
Coral reefs cover a fraction of a percent of the ocean floor and hold up a quarter of all marine life. When they go, they don't go quietly, and they don't come back on their own.
Atlantis Pads restores what's collapsing, actively, in the water, not from a lab.
Okinawa: Arrival
Before dawn, the dock at Onna is already working.
A mozuku line is checked by hand, the way it has been checked for generations. The man doing it has done this since before some of the machines that could replace him existed. He does it anyway, because no machine does it right yet.
A few hundred meters up the coast, an OIST researcher logs coral data before the water gets too bright to read it clearly.
This coral keeps this seaweed alive. This seaweed feeds most of Japan. Neither fact is abstract here. They are the same tide.
The reef is under stress. The people who know this water best are aging out faster than anyone is training to replace them.
This is not a place asking for sympathy. It is a place that still works, that still produces, that is running out of time to pass on what it knows.
Okinawa: Discover the Site
Five places worth a closer look. None of them are required.
With partnership, this is what Onna could become: the cooperative fully resourced, the coral monitoring funded before the reef needs saving instead of after, the knowledge that took a lifetime to earn actually passed to someone young enough to carry it another fifty years.
None of that is guaranteed yet. All of it is buildable.
This is a place that needs someone to grow alongside it.
Now, here is what happens when protection comes early instead of too late: the largest network of marine protected areas in the North Atlantic, built with the people who live on this water, not around them.
Traveling to Azores, Portugal
Azores: Arrival
5am, and the whale-watch boats are already leaving Horta in the dark.
Pico is lit up against a sky that hasn't decided yet if it's night or morning. An OKEANOS researcher is on the dock before the first boat clears the breakwater.
The channel is a mirror. The cetaceans come daily, on schedule, as if they know something the rest of the world hasn't caught up to yet.
This is not a place proving it can survive. It's a place proving the whole idea works, without asking for anything extra to make the math close.
Azores: Discover the Site
Four places worth a closer look. None of them are required.
Okinawa is what happens when protection comes late.
Azores is what happens when it comes early.
Same ocean. Same kind of place. Different timing, different outcome.
This is not luck. It is not geography. It is a choice, made in time or made too late, and the numbers follow the choice.
Every site in this network gets measured against the same question: has this water been protected, or does it still need to be?
The Wider Network
Act 6
Model the Network
Not a static ledger. An interactive scenario simulator, built on the same line-item data the rest of this appendix cites. Pick which sites are active, which funding route applies, and which ramp scenario to model. The headline numbers recompute live off those choices, plus a Risk and Seasonality panel that layers on real seasonal variance and disaster stress tests, site by site.
Sites
- Okinawa Startup $891K INTERNAL MODEL • Margin -37.1% to +18.5% INTERNAL MODEL
- Azores Startup $1.51M INTERNAL MODEL • Margin +12.7% to +49.4% INTERNAL MODEL
- South Africa Startup $888K INTERNAL MODEL • Margin -50.5% to +23.3% INTERNAL MODEL
- Norway Startup $1.34M INTERNAL MODEL • Margin -20.6% to +29.0% INTERNAL MODEL
- Kefalonia Startup $987K INTERNAL MODEL • Margin -6.0% to +42.1% INTERNAL MODEL
Funding Route
Ramp Scenario
Risk & Seasonality
Show seasonal range
Qualitative notes only, where a peak or low pattern is documented. No invented numeric spread.
Disaster stress test, per site
Okinawa
2026 compound heat and typhoon event
Aquaculture revenue applied at -28.1%, matching the documented 2026 mozuku harvest decline (combined heat and successive-typhoon damage).
Sourced: The Japan Times, "Okinawa prized seaweed under threat as oceans warm" (5 Jul 2026); atlantis-pads-numbers-justification.md.
Historical precedent, applied as an estimateHistorical precedent applied as an estimate, not a prediction.
Azores
Hurricane Lorenzo (2019)
$362M in archipelago-wide property damage is documented. No source quantifies revenue impact to this specific site. No quantified revenue impact available.
Sourced (event and damage figure only): atlantis-pads-deck-v6-script.md, [6.1b].
Flag only, numbers do not changeHistorical precedent applied as an estimate, not a prediction.
South Africa
Shark-absence scenario (orca-driven)
Cage-diving revenue toggled near zero, reflecting the documented shark-sighting collapse at this coast (300+ sightings/year in 2011 to zero every year since 2019, peer-reviewed).
Sourced: site-profile-south-africa.md; atlantis-pads-numbers-justification.md.
Historical precedent, applied as an estimateHistorical precedent applied as an estimate, not a prediction.
Norway
Storm Ingunn (Feb 2024)
Kefalonia
Wildfire and Mediterranean heatwave risk
Startup Capital
$1.78M
INTERNAL MODELAnnual Opex
$551K
INTERNAL MODELRealistic Revenue Range
$384K to $698K
INTERNAL MODELBlended Margin Range
-43.7% to +21.0%
INTERNAL MODELRange spans the documented conservative floor to realistic capacity cases already used throughout this deck, recomputed live off the sites, stress tests, and ramp scenario selected. No figure here exceeds what is already backed in atlantis-pads-numbers-justification.md.
Build Sequence
Bars reflect the network's existing Phase 1/2/3 rollout windows (network-overview.md), not invented month-level construction schedules. Okinawa and Azores each carry a separately sourced 8 to 18 month licensing timeline once their critical-path relationship or zone confirmation clears. See Act 7, Ready to Build.
- Okinawa
- South Africa
Grant Upside, not in the base case
Real, non-dilutive funding excluded from every headline number above. INTERNAL MODEL
$400K to $1.1M
Full sourcing: atlantis-pads-numbers-justification.md (line-item audit), network-overview.md and site-profile-*.md (per-site models), atlantis-pads-deck-v6-script.md (inline citations).
Act 7
Ready to Build
Everything before this was the case for the network. This is the start of building it. Two sites have a real, sequenced path to first operations.
Okinawa
The gating factor
A working relationship with the Onna Village Fisheries Cooperative (OVFC), established before any license application. Social and commercial, not legal or financial.
First move
An introduction to OVFC, ideally through OIST or a mutual local connection, before capital is deployed.
Timeline to first aquaculture revenue
8 to 18 months from cooperative engagement to license issuance.
Contact sequence
- JETRO (free FDI consultation)
- OIST Marine Science Station / Sea neXus hub (parallel track)
- OVFC (the critical path)
Open decision
Nakijin Village land (cheaper) versus Onna Village itself (pricier, co-located with OIST and the cooperative).
Azores
The gating factor
Aquaculture zone confirmation. Only three designated zones exist in Azorean waters. Outside them adds 1 to 3 years for re-designation.
First move
A query to the Regional Directorate of Sea confirming candidate parcels fall within or adjacent to a designated zone, before any property search begins.
Timeline to first aquaculture revenue
8 to 18 months if zone-aligned. Add 1 to 3 years if not.
Contact sequence
- Regional Directorate of Sea (the critical path)
- OKEANOS (parallel track, highest leverage)
- DGRM for formal licensing
Open decision
Purchase versus lease for the initial site, and which of the three zones to target.
South Africa, Norway, and Kefalonia don't have a getting-started guide yet. That gap belongs to each site's own "Critical Path: Not Yet Defined" placeholder, not to this act.