# Anse Cochon inquiry: identifying the vessels

Saint Lucia Policy Analysis · Coverage of the first public sitting, 13 October 2026

## SLPA’s working identification

Ocean Angel #3 is SLPA’s **highly likely working identification** of the Ocean Angel casualty vessel. The assessment draws on the visible name and #3 markings, open-boat layout, single-outboard views and registration-number strings, alongside additional research supplied by Kevin L. Michel. It is a reasoned editorial identification, not an authenticated registry finding or a Commission conclusion.

The photographic and documentary evidence answers different questions. The inquiry notices name Ocean Angel but print different registration digits. Photographs show fleet markings and particular configurations at unknown dates. Testimony describes the collision vessel and its operation. These sources can inform one another without proving that every photograph depicts the same hull or that its appearance matches the collision night.

Cazorla Jr. has a separate strong photographic lead. The unedited source photograph associated with a Mango Tours listing visibly carries **SL078180**, matching the number in the government inquiry notice. The connection is stronger than a resemblance alone, but dated owner or registry authentication remains outstanding in the materials reviewed.

## What the selected views contribute

| Photo | Visible evidence | Limit |
|---|---|---|
| 08 — Ocean Angel beach view | Ocean Angel name; apparent SL077875; light hull, red interior, blue canopy and one visible outboard | Lettering is soft. No #3 numeral is asserted. Photographer, dates and originating URL are unknown. |
| 01 — Mango Tours | SL078180 was read from the unedited Mango Tours source photograph | The public copy is an AI-assisted privacy edit, including regenerated pixels outside the softened faces. Its fine details are not independent authentication. Image date and registry cross-reference to the physical hull remain unverified. |
| 09 — Dockside view | Light hull, red interior, red-and-white benches, pale canopy, aft white console and partial Ocean Angel lettering | No readable registration is established. A filename is not evidence of fleet number. |
| 10 — Empty interior | Red-and-white benches, aft white console, canopy framework and one visible Yamaha outboard; a life-ring edge bears a 3 | A number on equipment is not a hull identifier. No readable hull name or registration establishes identity or engine power. |
| 11 — Two-boat photograph | The background boat appears labelled Ocean Angel #3 and has an apparent registration ending 857 | The number is partly obstructed. The foreground boat is not the identification subject. Nobody is identified from appearance. |

The four contributed photographs have no EXIF date fields. Their photographers, original capture/publication dates and originating URLs are unknown. The newly supplied empty-interior WEBP and an older operator-page JPG (Photo 02) show the same composition; different encodings are not independent corroboration. A repeated beach upload was byte-identical and is not counted twice.

## The 875/857 discrepancy remains unresolved

The government’s [8 September commissioners notice](https://www.govt.lc/news/commissioners-sworn-in-for-anse-cochon-maritime-collision-inquiry) prints **SL077875**. The [9 October public-sitting notice, available in this reproduction](https://stlucianewsnow.net/public-sitting-of-the-commission-of-inquiry-into-the-anse-cochon-marine-collision-to-commence-october-13-2026/) prints **SL077857**.

The beach photograph apparently supports the first string; the background #3 boat apparently has a number ending 857. This correspondence strengthens association with the same inquiry record. It does not independently establish that two official registrations belonged to one hull.

The available evidence does not resolve an erroneous visual reading, a published-number error, changed painted markings, a registration history or different hulls associated with the operator’s numbering. A registration change has not been established. Nor has a finding that the two photographs depict different hulls. Calling the discrepancy a settled transposition error would choose an explanation the evidence has not resolved.

Registry extracts and historical entries for both number strings, a hull identifier or construction record, inspection photographs and dated owner records could settle the cross-reference. A clearer image of the complete number on the background #3 craft would improve the visual reading, but would still need date and registry context.

## Configuration is useful evidence, within limits

The dock and empty-interior photographs help explain an aft helm/console and transverse passenger benches. They do not establish exact dimensions, manufacturer, model, capacity, engine rating, loading or the collision-night position of each passenger.

Console fittings, cushions, paint and canopy colour differ among the selected photographs. Such features can change. Differences therefore do not automatically disprove a common hull, and similarities do not prove one. Perspective, lighting, age and refits are explanations to investigate, not established histories.

One outboard is visible in the empty-interior view. Its appearance does not independently establish its horsepower or that the same motor was fitted on 18 July 2026. The testimony concerning machinery and helm operation retains its own source references and evidential status.

## Fleet comparisons

Ocean Angel 2 carries **SL077776**, different from both incident-number strings in the notices. Photographs of Ocean Angel #1 show its name and twin outboards. Those images are retained as fleet comparisons, outside the casualty-vessel selection. They are not displayed as the incident boat.

The selected photographs do not establish a complete fleet inventory. The existence or absence of a fourth numbered boat is not used to identify #3 by elimination. A historical reference to “Ocean Angel 3rd” is fleet context, not a registration cross-reference or proof of 2026 passenger capacity.

Other interior/profile views remain unconfirmed. A readable business name, dated review or life-jacket sign does not by itself establish the vessel’s identity, the photograph’s capture date or what occurred on the collision voyage.

## Source photographs and illustrations are separate

Ten selected source photographs are published without cropping, retouching or generated detail. Photo 01 is published as an AI-assisted privacy edit with occupants’ faces softened. Pixels outside the faces were also regenerated; the published copy cannot independently authenticate lettering, fittings or other fine detail. The SL078180 assessment was made from the unedited source, retained in the non-public research record.

`photos.json` records the public files’ dimensions and SHA-256 hashes. For Photo 01 it separately records the unedited source’s hash and dimensions. Its display and download use the privacy-edited copy; the legacy `original` field is a download alias, not an assertion that this copy is unedited. A hash verifies file integrity, not identity or truth. No person is identified or connected to the incident from appearance.

The separately labelled boat illustrations are AI-generated interpretations of the source photographs. Hull proportions and unseen fittings are synthesized. They can explain layout but cannot supply measurements or determine collision geometry, impact point, speed, capacity, machinery or individual passenger positions.

## What would permit firmer claims

A registry-backed number-to-hull history, dated originals and confirmation of the collision-night configuration would strengthen or change the identification. Qualified reporting can describe the present assessment while these records remain unresolved. None of the photographic work determines fault or establishes a precise collision track.

Public catalogue: https://saintluciapolicy.org/research/anse-cochon-marine-safety/vessels

Source metadata and disclosed public downloads: https://saintluciapolicy.org/documents/anse-cochon/2026-10-13/photos.json