Found twice, independently
In an exploratory reconstruction on the Maranhão coast, a systematic elimination of hypotheses ended at an uncomfortable wall: whatever carried the subject to the beach in the final stretch is not representable in any data the program can obtain. Months later, a kite incident on the Ceará coast — a rider separated from their kite roughly 2 km from the beach, rescued unharmed — re-demonstrated the same limit from a different direction: 100% of simulated particles terminated immediately, because the coastal cells of a 1/12° global grid (roughly 9 km per cell) do not resolve that strip at all.
Two coasts, two seasons, two very different cases, one conclusion. The zone where kitesurfing, surfing and swimming incidents actually happen is not covered by any existing public dataset.
Why this is a finding and not a complaint
Global current products are built for the open ocean, and at that job they are remarkable — the same series that exposed this gap also shows the engine matching real alongshore transport over six days. The blindness begins where the grid meets the beach: surf-zone dynamics, rip channels and the final wave-driven transport live below the resolution of every public product.
Naming the boundary precisely matters. Every drift estimate the program publishes declares it: the strip within roughly 2 km of the beach is not resolved by the underlying data. An estimate that hides its blind zone is more dangerous than no estimate.
What closes the gap
Not better software — measurement. The strip can only be characterised by instruments and tracked objects in the water on the coasts in question. That is the purpose of the program's planned small-scale release experiments with tracked equipment, run carefully from the beach with recovery in the same session. Until that data exists, the honest statement stands: models end 2 km from the sand.
This is research. Estimates are models, not guarantees; a model is never an observation, and a machine estimate is never “confirmed” — only a human reviewing it in the field makes it verified. Nothing on this page is an operational-capability claim, a safety promise, or a prediction of a specific event at a specific place.
References & data
- 1.Copernicus Marine Service — 1/12° global current products whose coastal cells define the resolution limit discussed here.
- 2.Diário do Nordeste — Public reporting of the June 2026 kite incident (ref. 1.3772763).
- 3.AYON research archive — Maranhão exploratory reconstruction report and Ceará case series (frozen-configuration hashes on file).