Ten Sightings over the Gulf

One squadron, seven reports

The ten sightings come from seven documents: six mission reports and one Range Fouler debrief form, the form an aircrew fills in after an encounter with an unknown contact. All seven name the 482nd Attack Squadron, as the originator of the report (“482ATKS”) or as the pilot’s squadron (“482 ATKS”). USCENTCOM declassified them between January and March 2026.

The missions follow one pattern. The aircraft, its type and tail number blacked out, “TOOK OFF FROM OKAS”, Ali Al Salem air base in Kuwait, and “SUPPORTED NAVCENT”, the US Navy’s command in the region, “IVO ARABIAN GULF, STRAIT OF HORMUZ AND GULF OF OMAN”. The purpose is the same throughout, the pattern of life of vessels, “UAS ACTIVITY” and “ACTIVITY OUTSIDE OF PORTS”; from 27 August it reads “TO CHARACTERIZE IRIN\IRGCN VESSELS, UAS ACTIVITY, ACTIVITY OUTSIDE OF PORTS, AND TO ESTABLISH PATTERN OF LIFE”. IRIN and IRGCN are Iran’s navy and the navy of its Revolutionary Guard. Each mission lasted 20 to 21 hours and gives its primary sensor as “ANDAS4”. On every mission the aircraft was hailed on the guard frequency, twelve times in all; where the caller is not blacked out, it is “IRANIAN AIR DEFENSE”.

The publisher’s titles do not all fit the documents. The debrief form of 31 August is titled “Range Fouler Debrief, Japan, 2023”; the form is dated “08/31/20” and its grid reference lies in the Persian Gulf. The report of 2 November is titled “Iran”; its grid references lie over open water north of Qatar.

What the reports say

Time (UTC)	Report	What it says	The aircraft
16 July, 18:30	D65, p. 6	“OBSERVED AN UNIDENTIFIED AERIAL PHENOMENON”; “Observed Activity Description: UAP”; “Method of Observation: FMV”	FL 200, 98 KIAS, heading 152 M
16 July, 19:20	D65, p. 6	the same words	FL 190, 90 KIAS, heading 34 M
16 July, 23:45	D65, p. 7	the same words	FL 191, 115 KIAS, heading 331 M
8 August, 07:26	D60, pp. 4–5	“OBSERVED 1X UAP … NO IMPACT TO MISSION.”; “Observed Activity Description: TRANSITTING”; FMV	not given
27 August, 15:27–15:29	D61, pp. 6–7	“OBSERVED A FORMATION OF FLYING OBJECTS TRAVELING NE-NW ALONG THE COAST”; “FORMATION OF UNK FLYING OBJECTS”	not given
31 August, dusk	D42	“there were three on the screen at the same time moving amongst each other”; contact at 18,000 feet, “Direction/Speed” “150/230”	not given
16 September, 17:32	D62, p. 8	“OBSERVED A UAP”; “Observed Activity Description: UAP”; FMV	FL180, 90 KIAS
2 October, 18:29	D63, p. 1	“OBSERVED 1X UAP, SEE OBSERVATION LINE 1.” The line is not in the file.	not given
2 November, 21:43	D64, p. 6	“OBSERVED 1X UNIDENTIFIED AERIAL PHENOMENON … ALTITUDE UNKNOWN WITH A BEARING OF 080 T.”; FMV	FL220, 105 KIAS, heading 110 T
2 November, 21:48	D64, p. 7	“OBSERVED AN ADDITIONAL UAP TRAVELING NW”; FMV	FL220, 107 KIAS, heading 110 T

The formation of 27 August is the fullest account among the mission reports: “FROM 1527Z TO 1529Z, […] OBSERVED A FORMATION OF FLYING OBJECTS TRAVELING NE-NW ALONG THE COAST IVO 39RVM81[…]99[…] […] WAS TRACKING THIS FORMATION FOR APPROXIMATELY 2 MINUTES BEFORE PID WAS LOST IN CLOUD COVER. AIRCREW WAS UNABLE TO GAIN PID AGAIN ON THIS FORMATION”. Its weather line reads “LIGHT CLOUD COVERAGE PREVENTED THE CONTINOUS TRACKING OF THE FORMATION”. On 2 October the weather was “HEAVY HAZE PRECLUDED IMINT ANALYSIS”, on 2 November “HAZE PRECLUDED IMINT COLLECTION”; the observation lines of the other reports say “WEATHER WAS NOT A FACTOR”, although on 8 August another line of the same report has “DENSE CLOUD COVERAGE INTERMITTANTLY IMPACTED FMV COLLECTION”.

Two reports call the same moment by two names. On 8 August the narrative has “OBSERVED NOTEWORTHY ACTIVITY” and the observation line “OBSERVED 1X UAP”; on 16 September the narrative has “OBSERVED A UAP” and the intelligence line “OBSERVED NOTEWORTHY ACTIVITY”.

The report of 16 September also records two losses of the data link earlier that day, at 12:48–12:59Z and 14:14–14:41Z: “EXPERIENCED LOST LINK TO POSSIBLE […]”, “Type of EMI: UNKNOWN”, “EMI Impact to System: Complete”, “Mission Changed: Flight Path Deviation”. It does not connect them with the sighting three hours later.

Where they were

The reports give positions as grid references in the military grid (MGRS), partly blacked out, and they are blacked out in two ways. In the reports of 8 August, 27 August and 16 September each black box carries a small label “1.4a”; we measured that each box covers three digits, so the two digits left before it give the easting and the two after it the northing, to 1 km. In the other four documents the label “(b)(1)1.4a” is printed over the neighbouring digits. Checked against the one grid reference the report of 16 July gives in full, only the first digit of the easting can be read with confidence; a 2 under the label there reads as a 3. Those positions are therefore known to a band 10 km wide from west to east, and not at all within 100 km from north to south. Every conversion to latitude and longitude below is ours.

• 16 July, 23:45Z. The only full grid reference in the series, “IVO 39RUN6234236874”: 29.253°N 49.583°E, to the metre, in open water about 70 km west of Kharg Island. The map pin stands here.
• 16 July, 18:30Z. The object’s grid gives 28.0–28.9°N, 50.5–50.6°E, a 10 km band; the text line reads “IVO 9RVM5[…]”, without the 3. The aircraft’s grid gives 26.2–27.1°N, 52.2–52.3°E. The two lie between about 190 and 350 km apart.
• 16 July, 19:20Z. The object’s grid gives 28.9–29.8°N, 49.7°E, a 10 km band; of the aircraft’s only the 100 km square is legible.
• 8 August, 07:26Z. To 1 km: the aircraft at 26.70°N 51.96°E; the “Observed Activity Location” at 27.61°N 50.90°E; the text line “IVO 39RWL08[…]52[…]” at 27.60°N 51.09°E. The aircraft lies 145 km from the first of these and 132 km from the second, and the two lie 18 km apart. The report does not say how any of the three was obtained.
• 27 August, 15:27Z. To 1 km: the formation at 28.92°N 50.81°E, about 6 km from the centre of Bushehr, along the Iranian coast; the aircraft at 28.56°N 50.37°E, 59 km away.
• 31 August. “MGRS Grid of initial contact: 39RWL2[…]”: 27.1–28.0°N, 51.2–51.3°E, a 10 km band.
• 16 September, 17:32Z. To 1 km: 28.66°N 50.50°E, about 47 km west of Bushehr, given for the aircraft and the object alike.
• 2 October, 18:29Z. No position.
• 2 November, 21:43Z and 21:48Z. 10 km bands, all between 26.2° and 27.1°N: at 21:43Z the aircraft at 51.7–51.8°E and the object at 51.5–51.6°E; at 21:48Z the aircraft and the object at 51.6–51.7°E. This is open water north of Qatar.

By our reading of map data, the positions known to the metre or the kilometre lie over open sea 40 to 70 km from the nearest land, except the formation of 27 August, whose square sits on the coastline at Bushehr. The 10 km bands, which run 100 km from north to south, also lie over the sea, but the band of 31 August comes within about 5 km of the Iranian coast at its northern end, and those of 16 July at 18:30Z and 2 November at 21:43Z reach inside twelve nautical miles of the coast at one corner.

The first object of 2 November is given “A BEARING OF 080 T”. On the grid references it lay 10 to 30 km west of the aircraft, which puts it somewhere in the western half of the compass as seen from the aircraft, never at 080°. So 080 T is more likely the direction in which the object was moving than its bearing from the aircraft. That is our inference; the report does not say which it means.

2 November: two objects, two recordings

On 2 November, from 19:10Z to 22:10Z, the crew “CONDUCTED OPEN WATER SCANS ISO UUV, NO IDENTIFICATION OF UUV WAS FOUND”: it was searching open water for an unmanned underwater vehicle. The two sightings, at 21:43Z and 21:48Z, fall within that search; the weather line reads “(U) HAZE PRECLUDED IMINT COLLECTION”.

Two recordings on the publisher’s video pages carry that date in the titles given by whoever uploaded them: “2 November 2020 [CALLSIGN] [CALLSIGN] Observes and tracks UAP 1 of 2” (DOW-UAP-PR077) and “… 2 of 2” (DOW-UAP-PR078), each 4 min 58 s. The publisher assesses that each is “likely derived from an infrared sensor aboard a U.S. military platform operating within the United States Central Command area of responsibility in November 2020”, and adds: “DOW-UAP-PR077 and DOW-UAP-PR078 are not duplicates. Both videos depict highly similar subject matter, but are distinct.”

By our viewing, both show the sea from above in infrared, and every overlay that would give time or position is blacked out.

• PR077. A small bright point near the crosshair, clearly visible from about 1:30. At 3:16–3:17 a second point crosses the top right corner of the picture, as the publisher describes, moving left and slightly up, while the blue “N” marker stands almost straight above the crosshair: roughly west-north-west on the screen. A dark dot to the right of the crosshair never moves on the screen; it is a mark of the display or the sensor, not the object.
• PR078. A bright point held at the crosshair, a tracking box labelled “15M” around 3:13–3:25, and a change of contrast at 3:50–4:15. Nothing crosses the top right corner between 3:14 and 3:18.

The publisher’s descriptions of the two recordings are identical word for word, including “03:15-03:17: A second area of contrast transits the top right corner of the sensor field-of-view”, which PR078 does not show. The description of PR078 repeats that of PR077. Both recordings also carry the same double smudge at the same place on the screen, which we found by extracting what stays fixed through each file; so they come from one sensor. Both findings are ours.

What joins the recordings to the report: the date in the uploader’s titles; the sea, which fits a search of open water; two objects, one after the other; and the second point’s movement on the screen, roughly west-north-west with north at the top, against the report’s “TRAVELING NW”. What separates them: nothing on the recordings gives a time, a position or a range; the report puts the two objects five minutes apart, and nothing says which object each recording shows; and movement on a screen mixes the object’s own motion with that of the sea and the sensor, so the direction agrees without being measured. It is the strongest pairing in the series, and still only a pairing made by us, from the date and the content; no document names these recordings.

31 August: three on the screen

The debrief form of 31 August was filled in by a pilot of the 482nd, grade “O-2”, on an “ISR” mission at “Dusk”; the time of detection is left blank. It gives the contact’s altitude as “18000”, constant, “Direction/Speed” as “150/230”, “Radar Equipped” as “Other”, and ticks two boxes, “Other Shape” and “Apparent Propulsion”. The pilot’s account in full:

“Saw the initial object fly through the screen and started tracking it. Initial object was surpassed by another object of same size and shape but much higher speed. At one point during tracking the objects, there were three on the screen at the same time moving amongst each other.”

Two recordings carry that date in their uploaders’ titles, “31 AUG 2020 [CALLSIGN] [CALLSIGN] Observes UAP” (DOW-UAP-PR088) and the same “part2” (DOW-UAP-PR089), both infrared by the publisher’s assessment. By our viewing:

• PR088. At 0:39–0:45 and again at 1:16–1:49 a dark shape moves with a long straight streak behind it. To us it looks like a boat and its wake on the water; it is the “area of contrast” the publisher places at 01:22–01:55. From 1:51 the picture shows the sea with small bright points moving about: two at once at 2:50–2:52 and again at 4:29–4:33, followed frame by frame.
• PR089. Small bright points entering and leaving the picture, and a tracking box at 0:25 and 3:50.

The two recordings carry the same fixed marks on the screen, so they too come from one sensor; this is our finding. What joins them to the debrief: the date, several small objects, two of them on the screen at once, and a sensor that moves on to a second object, which the publisher describes at 03:09–03:29 of PR088. What separates them: we found at most two on the screen at once, not the debrief’s three; the debrief gives no time, its height and “150/230” cannot be checked on the recordings, and a boat does not “fly through the screen”. The pairing is weaker than that of 2 November.

Recordings without a chain of custody

All four recordings belong to a collection the publisher describes in the same words on every page: “On March 6, 2026, eight members of the U.S. House of Representatives requested access to 51 potentially UAP-related records allegedly held by the Department of War and the Intelligence Community. The All-domain Anomaly Resolution Office (AARO) identified a collection of responsive materials held on a classified network. Many of these materials lack a substantiated chain-of-custody.” Their titles are, in AARO’s phrase, “uploader-defined”: given by whoever put them on the network, not by the unit that recorded them. Every pairing of a recording with a report on this page is ours, made from the date and from what the recording shows.

Two more recordings share a date with the squadron’s reports, and little else joins them. DOW-UAP-PR085, “16 Sept 2020 [CALLSIGN] [CALLSIGN] observes UAP”, from “a full-motion video camera” by the publisher’s assessment, shows the sea and a small bright point held near the crosshair; the report of 16 September says only “OBSERVED A UAP”. DOW-UAP-PR092, “08 AUG 2020 [CALLSIGN] [CALLSIGN] UAP observation”, shows the sea and, when the picture is enlarged, a point a few pixels across that the sensor keeps near the crosshair; the pale mark circling the crosshair is the overlay’s north marker, not an object. The report of 8 August says “TRANSITTING”. Both pairings are weak, and we do not cite either recording as a source.

A fifth recording of that summer, DOW-UAP-PR090, “24 AUG 2020 [CALLSIGN] (Mission) Observes UAP”, shares its date with a debrief from the North Arabian Sea filed by another unit (DOW-UAP-D56). Every frame of it, from 0:01 to 4:55, shows land: fields, a canal, walled compounds and trees. It belongs neither to that debrief nor to this series.

What the reports do not say

None of the ten sightings is given a size, a shape or a colour. Only the debrief of 31 August gives a height and a speed; a direction appears there (“150”), for the formation of 27 August and for the two objects of 2 November. “Relative Bearing or Clock Position” and “Range” are “-” in every observation line. The report of 2 October refers to an observation line that is not in the file.

No report of the squadron gives an explanation or an assessment of what was seen. Other reports in the same release do: one sorts its objects as “POSSIBLE BIRDS” (DOW-UAP-D10), others say a “BALLOON” (DOW-UAP-D7, DOW-UAP-D19). Each mission report says “FMV WAS EXPLOITED BY DGS1” or “DGS-1”: the video went to a ground station for analysis. What the analysts made of it is not in the release.

Why this scores where it does

Witnesses score 2: trained crews on seven missions, one squadron, and no one outside it. Radar scores 0: nothing measured a range, and “Radar Equipped: Other” on the debrief records no track. Optics scores 2: four recordings exist and show objects, but they are joined to the reports only by date and content, their overlays are blacked out, and the publisher itself says such material lacks a substantiated chain of custody. Physical trace scores 0.

Multi-channel scores 1: the reports and the recordings come from the same kind of sensor, flown by the same squadron. Documentation scores 2: seven formal reports with times and grid references, but the grids are blacked out to 1 km or 10 km, one observation line is missing, and two reports place the aircraft 130 to 350 km from what it observed. Anomaly scores 1: the reports record nothing unusual beyond the word “UAP”, a formation and one pilot’s “much higher speed”. Data today scores 2: the reports and the recordings are public, but without the times, positions and sensor data that would let anyone check them.

That puts the case at 1.2 overall, class D, with the status insufficient data.

What would settle it

The grid references without their redactions, and the missing observation line of 2 October. With them, the aircraft and the objects could be placed, and the distances between them that the reports of 8 August and 16 July leave unexplained could be checked.

The original sensor video of each mission, with its metadata: time, the aircraft’s position, where the sensor pointed and its field of view. With it, the four uploaded recordings could be matched to a mission or ruled out.

A record of who copied the four recordings, from which mission, and when. That is the chain of custody the publisher says is missing.

The ground station’s exploitation reports. Every mission report says its video was exploited by DGS; what the analysts saw in it would be the second pair of eyes this series lacks.