What the video records
A CBP Dash 8’s infrared sensor tracked one or more small heat sources near Rafael Hernández Airport. The image moves against terrain as the aircraft flies a curved path; later frames show fading and two visible thermal images. The footage records line-of-sight imagery, not a direct measurement of target distance, altitude, size or speed.
The original extraordinary reading
Civilian analysis interpreted the clip as one high-speed object crossing airport airspace, entering and leaving the ocean, then dividing. That conclusion follows only if the image track is treated as the movement of one distant object. It does not follow from the pixels alone.
AARO’s reconstruction
AARO used aircraft trajectory, target geometry, winds and clouds to assess two slow objects drifting together. On this account, aircraft parallax produces the apparent ground speed, clouds explain fading, the targets remain above the water, and initially overlapping heat sources later resolve into two. AARO assessed no anomalous or transmedium performance with high confidence and sky lanterns with moderate confidence.
What the record establishes
- CBP aircraft recorded infrared imagery of small thermal targets.
- Apparent motion in the image requires geometry before it can be interpreted as target speed.
- No public underwater sensor, recovered material or decisive radar track supports water transit.
- The official reconstruction strongly rejects the original anomalous-performance reading.
How the Aguadilla Story Developed
The event’s public importance comes from a later interpretation of video rather than a rich set of named witness descriptions. The case therefore turns on reproducible sensor geometry: distinguish what the clip shows from the Earth-fixed path inferred after assumptions about range and altitude.
CBP aircraft records infrared targets near Aguadilla.
SCU publishes a reconstruction interpreting the video as one fast object entering water and dividing.
AARO resolves the case as likely two sky lanterns, without anomalous or transmedium performance.
Evidence, Documents and Physical Record
Authentic thermal imagery
The footage is useful evidence of infrared targets, but a sensor image has no self-evident range or physical scale.
Aircraft motion matters
A turning airborne camera viewing a slow target against distant terrain can create an apparent rapid track through parallax.
Wind and cloud constraints
Wind alignment supports a slow drifting interpretation; cloud layers can hide or attenuate a thermal source without immersion.
No recovered or underwater evidence
No public physical sample, visible-light video, underwater track or decisive radar measurement confirms the extraordinary account.
Evidence Assessment & Hypotheses
Aguadilla demonstrates how a genuine recording can support sharply different physical reconstructions. The thermal images establish that the aircraft’s camera tracked targets; speed, altitude and behaviour must be inferred from the changing relationship between those targets, the sensor platform and the background. The dispute between SCU’s reconstruction and later conventional analyses turns on those assumptions, rather than on competing judgements about whether the footage exists.
High speed and movement through water require a range solution that places the targets at the distances used in the anomalous reconstruction. A nearer, slowly drifting source produces substantial apparent movement through parallax, while loss of thermal contrast can account for disappearance near the waterline. These alternatives undermine the claim that water entry or extreme performance can be read directly from the image sequence. The apparent separation also needs to be tested against the possibility of two overlapping targets.
AARO’s distinction between high confidence in non-anomalous performance and moderate confidence in a sky-lantern identification is consequential. A reconstruction may reject extraordinary kinematics more securely than it identifies the precise objects. The unresolved identification therefore supplies little support for a transmedium vehicle. Further analysis would be useful where it constrains geometry independently or tests competing models against the same frames, rather than reproducing a preferred trajectory from assumed distances.
Open Questions and Next Steps
Future work should publish and test the reconstruction inputs, not infer performance from the screen path alone.
Question
Can separate target centroids be resolved earlier?
Answer
Frame-level analysis with sensor metadata could further test the two-target account.
Question
What raw aircraft and weather records can be released?
Answer
They would allow independent replication of the geometric and cloud reconstruction.
Question
Can a specific launch source be identified?
Answer
It could raise confidence in the lantern identification but is not required to reject anomalous performance.
Witness Deep Dives
The available public record centers on the CBP crew’s sensor operation rather than detailed named-witness testimony.
Source Library
Official technical resolution and original sensor material should lead; competing analyses are valuable when their geometry assumptions are stated.
Tier 2 Institutional / Archival
Further references and records
These bibliographic references and record descriptions do not have a direct document link here. Where the discussion relies on a later account of a record, the linked author remains the source for that account.
- Original CBP infrared video, aircraft data and contemporaneous weather records.
- Airport operational records and any relevant radar material.
- Authentic CBP infrared video and associated metadata
- AARO formal assessment of the case
- Technical work on thermal imagery, parallax and airborne sensor geometry.
- Replicable critiques of both civilian and government reconstructions.
- Scientific Coalition for UAP Studies (SCU) 2015 kinematic analysis
- Arguments for a transmedium craft—useful for tracing claims, not substitutes for a calibrated trajectory.