Twenty-five contactsover the Gulf of Oman.

A USSOF AC-130J crew. EO/IR observations. Roughly twenty-five reported UAP instances. Six released clips. And a native recording that did not survive.

READ THE FULL CASE FILE10 MIN · STORY, SENSOR RECORD & SOURCESEXPLORE THE EVIDENCE
08Sep 2021
~25observations
06released clips
EO/IRnative file lost
AUTHENTIC CASE MEDIA · PR121 · 8 SEP 2021
Official AARO/DVIDS secondary cellular capture of the AC-130J infrared sensor display. Genuine case material — not a reconstruction, and not the missing native EO/IR file.
RESEARCH & EDITORIAL NOTE

Built as a case file, not a verdict.

The page separates the contemporaneous intelligence narrative, the six secondary captures, the lost native EO/IR recording and the performance estimates derived by the crew.

ESTABLISHEDSupported by surviving official documents or released material.
REPORTEDContained in the intelligence narrative or attributed to the operational crew.
OPENClaims that cannot presently be reconstructed from the surviving public sensor record.
INTELLIGENCE STATUS — IMPORTANT

The IIR is first-party reporting. It is not a finished intelligence conclusion.

DOW-UAP-D101 is an Intelligence Information Report: a formal channel for disseminating reported information inside the defence and intelligence system. That gives this case unusually strong documentary provenance.

It does not mean every observation or performance estimate in the report had already been independently validated. The report is not finally evaluated intelligence, and AARO's video descriptions explicitly say they should not be read as analytical judgments or factual determinations about the event's nature.

The correct reading is therefore unusually clean: what the crew reported, what the government preserved, what the public can inspect, and what the missing native sensor record prevents us from testing.

THERMAL READING

“Cold orb” does not mean somebody measured a cold object.

WHAT THE SENSOR SHOWS

Lower thermal contrast

In the released imagery, the reported objects appear as darker or lower-radiance regions relative to the surrounding scene in the infrared display.

That is an imaging statement. It does not establish the object's actual physical temperature in degrees.

WHY IT MATTERS

Language can outrun the data.

“Cold orb” is a useful witness/reporting label, but readers should not silently convert it into “physically refrigerated sphere.” The sensor is showing contrast, not direct-contact thermometry.

EVIDENCE LOCK: preserve the crew's “cold orb” description, but translate what the infrared display can and cannot prove.
CONVENTIONAL RECONSTRUCTION

Could the first objects be flare-related smoke, steam or condensation?

WHY THE HYPOTHESIS DESERVES TESTING

The timing is mechanically relevant.

The AC-130J was operating in a live-fire context and had deployed a water-activated flare. That creates a specific local mechanism capable of producing smoke, vapour, steam or condensation near the water.

If that material were cooler in infrared contrast than the warm Gulf surface, it could appear as a darker region in EO/IR imagery.

WHAT IT DOES NOT YET EXPLAIN

A complete reconstruction still has to fit the sequence.

The hypothesis has to account for the reported persistence, motion, re-entry into the field of view, multiple observations and any apparent response to aircraft activity.

No public reconstruction presently reproduces the full event from the missing native data.

TEST STANDARD: a conventional explanation is strongest when it reproduces the observed geometry and timing, not merely when it names a plausible object class.
PR120 / SENSOR MODE TEST

When the sensor mode changes, the anomaly becomes less distinct.

OFFICIAL DESCRIPTION

The display changes with the modality.

AARO's description of PR120 notes that the area of contrast becomes less distinct against the background when the sensor changes modalities.

That is an important instrument-level clue because the apparent visibility of the target depends, at least partly, on how the sensor is rendering the scene.

THE QUESTION

Did the object change — or the representation?

This does not prove an artefact. It does mean any physical interpretation should account for the fact that target contrast is not stable across sensor presentation modes.

The missing native recording prevents a full frame-by-frame radiometric analysis.

THE PUBLIC RECORD GAP

Roughly twenty-five reported observations. About two minutes survive publicly.

~25reported UAP observations in the wider episode
121 seccombined duration of the six released clips

This is the central evidential imbalance in the Gulf of Oman case. The official report describes a much broader operational episode than the public can inspect directly.

That means the six clips are important, but they are not a complete visual archive of the encounter. They are surviving windows into a larger event whose native EO/IR recording was not successfully preserved.

READER TAKEAWAY: do not mistake the released clips for the totality of the observation record.
01 / THE MISSION

They were not searching for UFOs.

On 8 September 2021, a U.S. Special Operations Forces AC-130J was conducting a live-fire exercise over the Gulf of Oman in darkness. A water-activated flare was being used as a mock target. The crew were operating an aircraft designed to find, track and engage targets through sophisticated sensors.

Then the EO/IR scan reportedly found something else.

The intelligence report that followed describes two approximately four-foot 'cold orbs' near the water. The pair were reportedly between zero and twenty feet above the surface. According to the report, they departed after the crew pulled the trigger for the aircraft's 105 mm cannon.

That was not the end of it. The report describes roughly twenty-five UAP observations during the wider episode, including formations, erratic manoeuvring and estimated speeds ranging from 250 to 1,300 mph.

Those are extraordinary claims. What makes the case important is that they were written into a contemporaneous military intelligence report rather than appearing only years later in a witness interview.

AC-130J and P-8A over USS Delbert D. Black in the Arabian Gulf
CONTEXTUAL IMAGE · U.S. Air Force AC-130J Ghostrider operating in the Arabian Gulf, 11 November 2022. U.S. Navy photo by Mass Communication Specialist 3rd Class Jeremy R. Boan / DVIDS. Public domain. Not the aircraft photographed during the 2021 incident.
02 / THE EVIDENCE GAP

The sensor saw more than the public record preserved.

What the crew reported

  • two approximately four-foot cold orbs near the water
  • roughly twenty-five UAP observations
  • formations and erratic manoeuvring
  • TACTOOL estimates of 250–1,300 mph
  • departure after the 105 mm trigger pull

What the public can inspect

  • six released secondary video captures
  • the contemporaneous intelligence report
  • reported crew-derived estimates
  • no surviving usable native EO/IR recording
  • no public raw telemetry package

The central problem is that the public record preserves two very different things.

First, it preserves the crew's reported observations and the intelligence narrative produced from them. That narrative contains the most dramatic details: the cold thermal appearance, the near-water behaviour, the reported formations and the TACTOOL speed estimates.

Second, it preserves six secondary video captures released publicly. These are not the same thing as the aircraft's native EO/IR recording.

AARO has made clear that the released material is secondary capture material rather than the original primary sensor file. That matters because a native sensor recording could contain metadata and image information that are critical for reconstructing range, field of view, angular movement and apparent speed.

The distinction does not make the report unimportant. It tells us exactly where the strongest claims live.

03 / THE SIX RELEASES

Six clips survived.

PR117 · 16 sec

PR117

An area of contrast remains generally within the centre of the sensor field-of-view.

OPEN ON DVIDS →
PR118 · 30 sec

PR118

An area of contrast remains generally within the centre of the field-of-view and briefly exits the bottom of frame.

OPEN ON DVIDS →
PR119 · 2 sec

PR119

A two-second secondary capture with an area of contrast generally near the centre of frame.

OPEN ON DVIDS →
PR120 · 13 sec

PR120

The tracked area remains generally centred; the sensor changes modalities twice during the final five seconds.

OPEN ON DVIDS →
PR121 · 34 sec

PR121

A larger area of contrast transits bottom-to-top, re-enters from the upper left and moves toward the lower left; a smaller area also exits and re-enters. AARO says this clip significantly overlaps PR122.

OPEN ON DVIDS →
PR122 · 26 sec

PR122

Closely overlapping material with PR121. AARO describes a larger contrast area crossing the field-of-view and a smaller area exiting and re-entering.

OPEN ON DVIDS →

PR117 through PR122 use the exact official DVIDS players published by AARO. They load only when you press play, so the page does not make six external video requests on opening.

PR121 and PR122 are official 7 August 2026 releases with resolved DVIDS pages. AARO notes that their content significantly overlaps, so six released files must not automatically be read as six wholly independent visual events.

04 / THE LOST RECORDING

The most important recording didn't survive.

No native EO/IR file to examine.

According to the intelligence report, the aircraft's DVR save attempt failed and the native EO/IR data became corrupted. That removed the record most capable of testing the strongest kinematic claims independently.

The most consequential sentence in the case may be the one describing what happened to the recording.

According to the intelligence report, an attempt was made to save the EO/IR footage from the aircraft's digital video recorder. The save failed, and the native data became corrupted.

That loss changes what can be demonstrated today.

If the native recording survived, independent analysts might be able to examine the original image sequence, sensor metadata, zoom changes, timing information and perhaps the assumptions underlying the speed estimates. Without it, the public is left with secondary captures and the report produced by the people who had access to the event closer to the time.

This is not a reason to discard the case. It is the reason the evidential hierarchy has to stay visible.

05 / THE SPEED CLAIM

250–1,300 mph — but how certain is the range?

1,300MPH · UPPER REPORTED ESTIMATE

The report gives estimated speeds ranging from 250 to 1,300 mph using TACTOOL.

Those numbers are among the most dramatic claims in the entire Gulf of Oman record. They are also among the claims most dependent on assumptions that the released material does not allow the public to reconstruct fully.

Speed derived from imagery depends heavily on range. A small error in assumed distance can produce a large error in calculated velocity. Without the native telemetry and full geometry, the published figures should therefore be described as reported estimates, not as independently verified performance.

That does not make them meaningless. The crew were operating in an environment where range, targeting and motion matter professionally. But professional estimation and recoverable raw telemetry are still different categories of evidence.

06 / EXPLANATIONS

What could produce the observations?

Several conventional possibilities remain serious.

Birds have been proposed for some infrared UAP videos, although the report's description of these objects as thermally cold and the reported performance would need to be reconciled with that explanation. Drones, balloons, classified systems, sensor artefacts, geometry errors and incorrect range assumptions also remain candidates.

The released secondary clips alone are not rich enough to eliminate all of those possibilities. Nor does the existence of a possible mundane explanation for one clip automatically explain approximately twenty-five reported observations.

The whole-case question is therefore not 'can something in an infrared video look strange?' It plainly can. The harder question is whether one conventional explanation can account for the reported thermal appearance, repeated observations, formations, near-water behaviour, speed estimates and operational response together.

At present, the public record does not allow that question to be settled cleanly.

Birds

Potentially relevant to some infrared imagery, but must be reconciled with the reported thermal description, behaviour and speed estimates.

Drones / balloons

Possible candidates for some observations. The question is whether they fit the reported formations, near-water behaviour and performance.

Classified systems

Cannot be excluded publicly, but invoking an unknown system is not itself an explanation of the released record.

Geometry / range error

Especially important because speed calculations depend strongly on distance. The missing native telemetry limits independent testing.

WHAT SURVIVES / WHAT IS MISSING

A strong official record with one enormous hole.

What survives

  • contemporaneous intelligence report
  • multiple trained military observers
  • six secondary video captures
  • operational live-fire context
  • reported TACTOOL estimates

What remains unresolved

  • Native EO/IR/telemetry if recoverable
  • range assumptions behind TACTOOL speeds
  • identities/independent statements of crew
  • radar or RF correlation
  • exact sequence mapping of PR117–122 to the 25 observations
  • final intelligence/AARO assessment.
HELP IMPROVE THIS CASE

Can you help reconstruct what happened?

We would like to hear from people who can improve the public understanding of the sensor record, geometry, released clips or conventional alternatives.

Useful contributions can include imagery analysis, AC-130J/EO-IR expertise, range and motion analysis, identification of ordinary objects, source corrections, or public documents we have missed.

PRIMARY SOURCE ROOM

Start with the government record.

AARO release index

The official AARO page lists PR117–PR122, their durations, secondary-capture limitations, contemporaneous relationship and AARO descriptions.

OPEN AARO RELEASE INDEX →

DOW-UAP-D101

The accompanying Intelligence Information Report is the source for the 'cold orbs', approximate four-foot size, 0–20 ft reported altitude, wider observation count and TACTOOL speed estimates. It is reported intelligence information, not a final analytic verdict.

THE UAP MAP ASSESSMENT

Six clips survived. The most important recording did not.

Gulf of Oman is powerful because the extraordinary parts of the story were not invented by later popular retelling. They appear in a contemporaneous official intelligence narrative generated inside the military reporting chain.

But the case also contains a major evidential fracture.

The report preserves the dramatic performance claims. Six secondary video captures preserve pieces of the visual record. The native EO/IR recording that might have allowed those claims to be tested most rigorously did not survive in usable form.

That leaves an unusual modern UAP case: stronger than a story told from memory, weaker than a complete raw sensor package, and important precisely because the gap between those two things is so clear.

Six clips survived. The most important recording did not.

GULF OF OMAN · FLAGSHIP PROTOTYPE V2 · THE UAP MAP
GULF OF OMAN AC-130J FLAGSHIP REVIEW V2 · COMPETITOR-UPGRADED · 8 SEPTEMBER 2026