Iraq.Iraq 2022 · two crews, one day, 'UAP' in the narrative and 'UAS' in the form

On 5 May 2022, an MQ-9 with callsign CDR74 took off at 06:14 Z, late because of weather at its launch and recovery element, and came on station ten minutes later. The mission was reconnaissance, inside Operation Inherent Resolve, tasked by U.S. Central Command's 609th air operations centre and flown by the 196th Attack Squadron of the 163rd Air National Guard wing. The aircraft spent seventeen hours and fifty minutes airborne over Iraq: it collected signals intelligence with AirHandler version 2 between 06:40 and 00:05 Z, and from 10:54 to 21:00 Z supported a Marine aircraft group operation under the name PHANTOM FLEX, coordinated 409 minutes before takeoff. The report records that pre-coordination was satisfactory, that weather was not a factor and that the intelligence gap was filled. It landed at 00:39 Z the following day at Ali Al Salem, Kuwait, after eighteen hours and twenty-five minutes of mission.

At 20:26 Z, while scanning a point of interest in grid 38SMD38, the crew observed what the report describes thus: 'periodically observed UAPs flying sporatically within the target area of 92AS. According to the MTS, the UAPs were travelling at various speeds between 80-180 MPH. We were unable to determine altitude. The size of the UAPs were approximately 1-2 meters. The UAPs did not hinder the mission.' The form's observation block adds the aircraft's own state at that instant, which is what makes this document something other than an anecdote: flight level 200, one hundred knots indicated airspeed, heading recorded as '-75' — presumably 075 — a range of 4.98 nautical miles, killbox and keypad 91AS, and method of observation 'FMV', full-motion video. The primary sensor is the AN/DAS-4, the Reaper's electro-optical and infrared ball. The video was exploited by DGS 1, the distributed ground station.

The datum that reorders the reading is not in the narrative but in the form. Two lines above the prose that speaks of UAPs, the structured field 'observed activity description' contains a single word: 'UAS'. That is the regulation abbreviation for unmanned aircraft system. The same crew that wrote 'UAP' four times in the free paragraph coded the observation as drones in the box the system reads. It is not a careless contradiction: they are two records with different jobs — the box classifies for the statistics, the prose recounts what happened — and what it reveals is that the crew did not believe they were seeing anything extraordinary. The closing sentence confirms it in the language in which reports are closed: the objects did not hinder the mission, so the mission continued.

What was observed fits that reading without strain. Objects one to two metres across moving at between 80 and 180 miles per hour over an Iraqi target area in May 2022, at the height of Operation Inherent Resolve, describe a small drone with uncomfortable precision; drones of that size were routine in that theatre and that year. The inability to determine altitude is exactly what a single electro-optical sensor without ranging on the object yields: it measures angle, not distance, and without distance there is neither height nor true size — the '1-2 metres' is an estimate derived from assuming a range. What the document does not settle is whose they were: the report does not say, there is no identification, and in that theatre and that year there were drones from several sides in the air.

That was not the only crew to see it that day. The same release published DOW-UAP-D109, MISREP 7467030 from another MQ-9 — from the 20th Attack Squadron, under Central Command's air component rather than the previous one's ACC — which had taken off at 21:49 Z on 4 May and was on station from 00:57 to 15:52 Z on the 5th. Its tasking was target development over a compound, with the declared intent to 'conduct POL, identify PID of weapons and weapons/ammo storage areas, as well as identify personnel within the compound.' At 05:24 Z, fifteen hours before the other aircraft's observation, this crew recorded in the narrative: 'observed PROB UAPS traveling SW at roughly 80-180 MPH.' The observation block develops it in the same words as the other report — 'flying sporatically within the target area... travelling at various speeds between 80-180 MPH. We were unable to determine altitude. The size of the UAPs were approximately 1-2 meters... The UAPs did not hinder the mission' — and adds two things the first one lacks. First: the observation period runs from 05:24 to 07:04 Z, an hour and forty minutes, and the crew 'would reacquire the UAPs after returning the sensor back to the origional POI locations', meaning they were still there each time the camera came back. Second, in the weather field: 'due to dust storms some collection on UAPs was affected between 0524Z and 0704Z.'

The sixth release also published the footage for that second observation, which it did not do for the first. They are two clips that expressly cite D109 and repeat its verbatim figure —'traveling at various speeds between 80-180 miles per hour'— so the link between video and report does not have to be inferred. DOW-UAP-PR140 runs one minute forty-seven and DOW-UAP-PR141 two minutes thirty-six: four minutes and twenty-three seconds in all, against the hour and forty minutes the report says the observation period lasted. They are excerpts, not the full record.

What the two clips mostly teach is about the instrument. PR140 switches between electro-optical and infrared mode five times in under two minutes, and at each switch the area of contrast leaves or re-enters the frame; PR141 shows both channels side by side —infrared left, electro-optical right— and records that around fifty-two seconds a second area of contrast briefly appears, which fits a report that speaks of 'UAPs' in the plural. It also records that the sensor loses contact intermittently through the final minute. That set —mode switching, lost contacts, reacquisitions— is exactly the operational signature the report itself describes when it says the crew 'would reacquire the UAPs after returning the sensor to the original point-of-interest locations'. The video adds no shape and does not resolve the altitude the report declares it could not determine: it confirms the sensor's behaviour, not the object's nature.

Two crews from different squadrons, the same day, over the same area, describing the same thing with the same figures. That is corroboration and it shifts the weight of the reading, though not in the direction the word 'UAP' suggests: what two independent observations of one-to-two-metre objects loitering for hours around a compound under surveillance for weapons storage most naturally describe is a drone presence at that compound. The dust storm does little for the other reading — it degrades the imagery and therefore the size and speed estimates — and the fact that the objects were still there each time the sensor returned speaks of local persistence, not transit. One difference between the two reports is worth noting too: where the 196th coded 'UAS' in the observed-activity description field, the 20th's is redacted under the national-defence exemption, so we do not know what it wrote.

It is worth noting too how the paper reached us, because it is the archive's machinery working. The form itself sets its automatic declassification for 6 May 2047, twenty-five years after the flight. It did not wait: the document carries the stamp of Central Command's chief of staff, Major General Richard A. Harrison, declassifying it in June 2026 inside the mandatory declassification review batch USCENTCOM MDR 26-0084 to 26-0095 — twelve files — and annotated 'approved for release to AARO'. The redactions that remain correspond to categories 1.4(a) and 1.4(g) of the classification executive order — military plans and operations, systems vulnerabilities — and to the personal privacy exemption: the callsign, the aircraft type, the tail number, the departure base and the names of the four people who signed the report all vanish. Everything else survives, which is a great deal more than this corpus gets from a typical case.

This file is, above all, a lesson in how an archive is built. AARO assembles its collection by searching for records 'responsive' to UAP subject matter, and that search runs on text. A mission report in which the crew wrote 'UAP' in the free paragraph enters the collection even though the same crew ticked 'UAS' in the box beside it; an identical report in which someone had written 'drone' in both places would never have entered at all. The consequence is twofold and worth keeping in mind when reading any count from the collection: the archive inherits the vocabulary of whoever filled in the form, not that person's judgement, and that is why it contains cases their own authors considered closed. It is not a defect of the office — a keyword sweep is the only practical way to comb decades of reports — but it is a warning against counting files as though each one were an enigma.

The second thing this case does is mark the good end of the evidentiary scale, and the sixth release supplies the comparison on the same day. The Colorado file published alongside it has one witness with a phone and redactions that erased the time, the bearing and the place: nothing to check. This one brings time to the second, a grid, a flight level, indicated airspeed, heading, range, the sensor identified and the station that exploited the video. Both end with a low probability, and for opposite reasons: one because nothing can be checked, the other because nearly everything can be and what checks out points to a drone. That the corpus separates them by tier — A against B — and joins them on probability is exactly what those two axes exist to express.

There remains a residue the file does not close and should not disguise. The report does not identify whose the objects were, and in Iraq in May 2022 that question is not rhetorical: there were drones belonging to Iranian-backed militias, to the coalition and to civilian operators in the same airspace. That the crew coded them as UAS says what kind of thing it believed it was seeing, not whose it was. That is why this case's partition leans towards another state's technology without settling it, and leaves a long quarter split between the mundane reading — a friendly or commercial aircraft — and indeterminacy. 'Did not hinder the mission' settles the operation, not the attribution.