2013 Aguadilla UFO Video Incident

Case dossier

The underlying event is not a rumor or a recreated clip. An infrared sensor aboard a U.S. Customs and Border Protection De Havilland Canada 8 aircraft recorded small airborne targets while the aircraft climbed and circled Rafael Hernandez Airport near Aguadilla. DVIDS now hosts the official 3:54 file, and AARO has published both a case-resolution report and a digital reconstruction. Those records authenticate the video and its government origin; they do not, by themselves, make the targets extraordinary.

The date requires an explicit caution. The SCU report writes that the encounter began at 21:22 local time on 25 April 2013, corresponding to 01:22Z on 26 April, and the sensor overlay displays 26APR2013 with a UTC notation. AARO and DVIDS instead identify the event as 26 April and AARO describes 21:22 as local time on that date. No public acquisition log resolves whether the difference is a time-zone conversion or a later metadata error. This archive keeps 25 April as the local-date anchor while displaying both official versions.

The Scientific Coalition for UAP Studies published a detailed analysis in 2015 and later deposited version 2, together with the AVI file, on Zenodo. Its authors identified the sensor as a Wescam MX-15D and reconstructed a low, fast track. They argued that one object crossed the airport without a transponder, entered the Atlantic with little deceleration, emerged, enlarged and split. That report explains why the Aguadilla video became a leading transmedium claim, but these are the authors' analytic conclusions rather than findings issued by CBP, FAA or another government investigator.

AARO presented a different reconstruction at a Senate hearing in November 2024 and issued a seven-page case resolution on 20 March 2025. It combined the aircraft's recorded position with sensor elevation, azimuth and slant-angle information in Systems Toolkit, then used line-of-sight and minimum-separation analysis. AARO assessed the targets at about 656 feet and smaller than one meter, while the observing aircraft climbed roughly 1,725 feet and the range to the targets nearly tripled. The report calls the target shapes indistinct and acknowledges that image quality worsened with distance and cloud cover.

In AARO's model, the two objects moved southwest in a straight line at approximately 3.6 metres per second, or 8 mph, close to the recorded east-northeast wind of 4.4 metres per second. The aircraft's own speed, circling path, sensor zoom and changing geometry produced motion parallax: slow nearby targets appeared to race across the background. The reconstructed line of sight placed the objects over land from first detection near the airport's northeast side until sensor contact ended above the central parking apron. This removes the need for extraordinary acceleration or an underwater trajectory.

The apparent split and water entry also have geometric and sensor explanations. AARO identified moments at about 29.56, 40.76 and 47.00 seconds when both objects were already separately visible. As the aircraft climbed, the view changed from shallow and side-on to steeper and more top-down, increasing their apparent separation. Near the end, the Atlantic formed part of the background while the small warm targets lost contrast. Thermal crossover, increasing range and scattered clouds could make their signatures fade, disappear and return without either target touching the water.

AARO attributes the objects to a pair of sky lanterns with moderate, not high, confidence. It cites their sub-one-metre estimated size, movement close to wind speed, variable flickering thermal signature and statements from local hospitality vendors that lantern releases are common at celebrations. The report considered marine birds unlikely because wings or wingbeat pulsation should have remained visible at the modeled distances. It also discussed tied Mylar balloons but rejected reflected moonlight as the proposed infrared mechanism. Poor video quality prevents a categorical identification.

SCU did not accept the public resolution without further data. In April 2025 it said AARO had acknowledged questions about the three-dimensional coordinates, uncertainty margins, cloud-layer source, tower or military radar records and the pixel analysis used to reject a utility-pole crossing, but had not yet supplied answers. These are legitimate reproducibility questions. They limit independent verification of the model, yet they do not restore evidence of high speed or water entry. The current record most strongly supports two slow windborne objects whose extraordinary appearance arose from geometry and thermal imaging; exact sky-lantern identification remains probable rather than proven.