Puerto Rico UAP:
Sky Lanterns, Sensor Geometry and the Transmedium Claim

Managing Expectations Research Note · September 7, 2026 · AARO / Aguadilla / Puerto Rico Object / infrared video / evidence labels

The 2013 Aguadilla, Puerto Rico infrared video has long circulated as one of the stronger public UAP clips because it appears to show an object moving fast, splitting into two, and entering or exiting the ocean. The video is real. It was recorded by a U.S. Customs and Border Protection aircraft near Rafael Hernández Airport on April 26, 2013. But the existence of a real government video does not settle what the video shows.

AARO’s 2025 case-resolution report is useful because it attacks the strongest version of the public claim: speed, replication, and transmedium motion. Its conclusion is narrower and less dramatic. AARO assesses with high confidence that the objects did not demonstrate anomalous performance characteristics or transmedium capabilities, and with moderate confidence that the objects were a pair of sky lanterns.

What the public record supports

DVIDS hosts the public video under the title Puerto Rico Objects. Its description says an infrared sensor aboard a CBP aircraft captured footage over Rafael Hernández Airport near Aguadilla and that the footage appears to depict a UAP moving at high speed, splitting into two objects, and entering and exiting the water before disappearing off Puerto Rico’s northwest coast. That sentence describes appearance, not a final physics finding.

The same official description then gives the important correction: AARO assessed with high confidence that the objects did not show anomalous speeds or flight behaviors. A reconstruction by an Intelligence Community partner indicated that the video instead depicts two objects travelling near each other rather than one object splitting in two. The reconstruction also indicates that the objects moved in a straight line at wind speed and did not enter the water.

Why the “transmedium” impression matters

Transmedium claims carry weight in UFO culture because they imply a craft moving between air and water without losing performance. If true, that would be a serious aerospace and maritime problem. But a claim that important requires more than a difficult viewing angle and a low-context clip. It needs range, altitude, sensor look angle, target path, shoreline geometry, sea-surface context, and competing hypotheses tested against each other.

AARO’s reconstruction does exactly the kind of work the public often cannot do from a compressed clip. The DVIDS reconstruction page says the Systems Toolkit model shows the aircraft’s flight path, the infrared sensor’s look angle to the ground, and the objects’ assessed flight path. From above, the mystery changes shape: what looked like a fast object over complex background becomes objects travelling straight over land at wind speed.

Sky lanterns are a hypothesis, not a slogan

“Sky lantern” can sound dismissive if used as a shortcut. In this case, AARO’s confidence wording matters. The office gives high confidence to the negative finding — no anomalous performance or transmedium capability — and moderate confidence to the specific identification as a pair of sky lanterns. That distinction is responsible. It says the extraordinary behavior is not supported while leaving more uncertainty around the exact mundane object.

That is how evidence labels should work. Analysts can be confident that a video does not show the behavior claimed for it without being equally certain about every remaining detail. A drifting heat source, viewed obliquely by a moving aircraft over terrain and coastline, can generate misleading impressions. The identification question and the “did it do something extraordinary?” question are related, but not identical.

What this case teaches

The Puerto Rico Object belongs in the UAP conversation because it is a public, named, sensor-based case with a long afterlife online. It also shows why public-record literacy matters. A government video can be authentic and still be misunderstood. A dramatic appearance can weaken when reconstructed. A popular interpretation can persist long after better geometry is available.

NARA’s UAP Records Collection and AARO’s case-resolution releases are part of a healthier public process: put records, videos, summaries and metadata where people can read them. That does not mean every agency conclusion is beyond criticism. It means criticism should engage the actual records. If someone rejects AARO’s Puerto Rico conclusion, the burden is to show where the reconstruction fails, what alternative path fits the sensor angle and wind, and why that alternative explains the apparent split and water contact better.

Managing expectations

Read the Aguadilla video in layers. Layer one: a real CBP infrared recording. Layer two: a striking visual impression. Layer three: a reconstruction that changes the geometry. Layer four: an official conclusion that the objects did not behave anomalously and were likely sky lanterns. None of those layers requires pretending the original observers were foolish. It means difficult sensor views can mislead honest people.

The disciplined conclusion is not “all famous UAP videos are lanterns.” It is narrower: this famous video no longer supports the strongest claims attached to it unless someone can overcome the official reconstruction with better evidence. That is not a defeat for UAP research. It is the point of research.

Useful source links

• AARO: Puerto Rico UAP Case Resolution (official case-resolution PDF)
• AARO: UAP Case Resolution Reports (official index of AARO case resolutions)
• DVIDS: Puerto Rico Objects (public-domain DoD-hosted video page)
• DVIDS: 2013 Puerto Rico Object Reconstruction (Systems Toolkit reconstruction page)
• NARA: UAP Records Collection FAQ (official records-transfer and access context)
• NASA: UAP Independent Study Team Final Report (official data-quality framework)