Skip to content
Coastal IntelligenceCase series

What the wind said: 12 kite incidents on the Ceará coast, matched to the wind measured that day

A research case series: we logged 12 kitesurf incidents on the Ceará coast from public reports, 2017 to 2026, and went back to the wind measured on each day. In the cases with usable data, two patterns appear. Injuries from the force of the wind fell on unusually strong days for the month, and the two riders rescued adrift were caught by an unusually sharp afternoon drop. The adrift deaths show no wind signal at all. The numbers are small, and we say so: this defines the test, not the rule.

Case reconstructions

Research film: the Ceará register and the wind on each incident day. Satellite imagery and real airport wind records; synthetic narration; no person or incident site is shown.

AYON Coastal Intelligence · contains modified Copernicus Sentinel data 2026 · wind data IEM/METAR · 96s

A register that did not exist

We found no public register that links kite incidents on this coast to the conditions on the day. The record is scattered across press reports and local rescue logbooks. This research started one: 12 incidents between 2017 and 2026, from Tatajuba in the west to Fortaleza in the east, 6 of them fatal. Each entry records the beach, the month, the outcome and the pattern. People are not named, and only facts already public in the press are used.

A state bill filed in July 2026 proposes an official register of adventure-tourism accidents, kitesurfing included. Until one exists, this is a working register, open to correction by anyone who was there.

Each day, against its own normal

For each incident we took the hourly wind measured at the nearest airport station: Fortaleza, 2005–2026, about 193,000 reports, and Jericoacoara, 2018–2026, about 8,800. The day's mean from 09:00 to 17:00 was ranked against every day in the same calendar month on the same instrument, so the comparison never mixes sensors.

A second pass used ERA5 reanalysis at each beach, 2000–2025, to rank the afternoon drop (the change from late morning to mid-afternoon) against the spot's own normal for the month. Airports read lower than beaches, and reanalysis under-reads coastal wind, so the ranks matter here, not the knots.

Pattern 1: the force of the wind

All three wind-force cases with a usable station fell on days in the windiest 10% for their month. Lagoa do Cauípe stands out: a low-season month, and a jump of about 6 knots overnight after three light days, the kind of day when riders may be on kites sized for lighter wind.

  • Cumbuco, October 2017, injury: day mean 20.0 kt against a month normal of 15.2 kt, the 98th percentile.
  • Lagoa do Cauípe, May 2024, fatality: 15.9 kt against 10.9 kt, the 97th percentile.
  • Barra do Ceará, October 2019, fatality: 18.6 kt against 15.2 kt, the 91st percentile.

Pattern 2: the afternoon drop

Both riders rescued adrift were out on ordinary or light days, and both were caught when the wind fell away in the afternoon. A drop that is unusual for that beach in that month is a signal a forecast can rank in advance; it is the signal this research will test next, on every new case.

  • Jijoca de Jericoacoara, August 2024, rider rescued: at the airport the wind fell to 6 kt at 14:00, when 13 is normal for the month. At the beach, the afternoon drop was bigger than on 91% of August days since 2000.
  • Icaraí, June 2026, rider rescued: an ordinary day, the 55th percentile at the airport, with an afternoon drop at the beach bigger than on 96% of June days.

What the wind does not explain

The adrift deaths share neither signal. Cumbuco (May 2026), Praia da Baleia (January 2026) and Paracuru (November 2026) had steady or strong wind and no unusual drop. Those cases point elsewhere: separation from the board, riding alone, no flotation, or a medical event. Wind data alone cannot explain them; the register's own fields will have to.

Three cases, Preá 2019, Jericoacoara 2022 and Tatajuba 2025, had too few airport reports on the day to rank. Swell was small on every incident day.

Where the wind is measured

Wind on this coast is measured, but nobody reads it together. Along the kite beaches the readings come from private stations run by pousadas and kite schools and published on wind-forecast sites (at Guriú, Ilha do Guajiru, Icaraí de Amontada, Mundaú, Paracuru, Taíba and Cumbuco, among others), from a research station near Pecém, and from the two airports inland. Many keep only days of history, none is archived beside incident records, and their siting and calibration are not published.

That is the next step of this research: an open daily archive of the coast's publicly visible station readings, kept beside the incident register, so the next case can be checked against a beach measurement rather than an airport inland.

Limits

12 incidents, most from a single press source; 8 with usable airport wind data. No cause is implied: this research reports what the wind did on the day, not why anyone was hurt. Any family can ask for a case to be removed from the register, and it will be.

Evidence discipline

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. 1.Iowa Environmental Mesonet — ASOS/METAR archive — Hourly reports for Fortaleza (SBFZ) and Jericoacoara (SBJE) airports, used for every day's mean and monthly normal.
  2. 2.ERA5 — Copernicus Climate Change Service — Hourly reanalysis wind at each beach, 2000–2025, for the afternoon-drop ranks.
  3. 3.Diário do Nordeste / O Povo — Public reporting of the incidents in the register; article references on file.
  4. 4.Assembleia Legislativa do Ceará — PL 465/2026 — State bill proposing a register of adventure-tourism accidents, filed 1 July 2026.
  5. 5.Windguru / Windfinder station listings — Private beach stations publicly visible on the Ceará coast, checked 13 November 2026.
  6. 6.AYON Ecosystem research archive — Incident register, analysis scripts and the spot-and-month normals table; on file.