AARO Resolves Al Taqaddum 'Jellyfish' as Balloons

Brief

The report, dated 8 September 2025, resolves a 2017 sighting in which an IR sensor aboard an aerostat blimp at 2,700 feet recorded an unidentified floating object for 17 minutes and 30 seconds over Al Taqaddum Air Base, Iraq. AARO assessed with high confidence that the object was a cluster of fully and partially inflated balloons traveling 4–14 mph at an altitude of 850–2,200 feet, consistent with local wind patterns running east to west. An alternative hypothesis, that the object was a camouflage-draped quadrotor UAS, was rejected because the object drifted passively with the wind and produced no heat signature detectable by the IR sensor. Video quality degraded as the object moved farther from the sensor, but AARO concluded the data were sufficient for a high-confidence resolution.

Metadata

Agency

DoD / All-domain Anomaly Resolution Office (AARO)

Release

2023-01-01

Type

PDF • .pdf

Length

3 pages

Classification

UNCLASSIFIED

Tags

balloon cluster, infrared, Iraq, 2017, aerostat, Al Taqaddum, force-protection blimp, wind-drift

Key points

• AARO assessed with high confidence that the object was a cluster of fully and partially inflated balloons; the case is fully resolved.p.1
• The sighting occurred on 23 October 2017; an IR sensor aboard an aerostat at 2,700 feet recorded 17 minutes and 30 seconds of footage.p.1
• Assessed object altitude: 850–2,200 feet; assessed speed: 4–14 mph east-to-west, consistent with wind data.p.2
• The object's reporter is listed as 'Unknown'; the originally reported shape was described only as 'Abnormally shaped.'p.1
• Dangling strings below the object changed shape and number in the footage, consistent with balloons rotating relative to the sensor's point of view.p.2
• The fluctuating IR return was attributed to the sensor auto-adjusting grayscale pixel values against a changing background, not to any anomalous thermal behavior.p.2
• A quadrotor UAS hypothesis, proposed by one of AARO's partners, was discarded on two grounds: passive wind-drift behavior and absence of motor heat signatures.p.2
• Video graininess increased toward the end of the clip, assessed as a function of the object's growing distance from the sensor, not a sensor malfunction.p.2
• AARO explicitly noted it is not a member of the Intelligence Community and that its reports should not be treated as finished intelligence.p.3

Verbatim

• AARO assesses with high confidence that the Al Taqaddum object did not exhibit anomalous behavior or capabilities.

  p.1

• AARO assesses with high confidence that the object was consistent with a cluster of fully and partially inflated balloons (Figure 1).

  p.1

• AARO assesses with moderate confidence that the object's altitude was between 850-2,200 feet and that the object was moving at 4-14 miles per hour (mph), due to the variability in both historical and real-time wind data.

  p.2

• The fluctuating IR return of the object is a result of the sensor constantly adjusting to assign grayscale values to every pixel, which maximizes the visual dynamic range in a diverse and changing background.

  p.2

• A quadrotor UAS would be unlikely to drift with the wind, as this object does.

  p.2

• The motors of a quadrotor UAS would generate heat visible to an IR sensor, and the video does not indicate the presence of such heat sources.

  p.2

• AARO is not a member of the Intelligence Community.

  p.3

Most interesting

• The aerostat (blimp) carrying the IR sensor was itself a force-protection asset stationed at 2,700 feet, the object it recorded was assessed to be significantly below it, at 850–2,200 feet.
• Despite 17 minutes and 30 seconds of footage, the original reporter listed is simply 'Unknown,' and no altitude or speed was reported at the time, both had to be reconstructed analytically.
• The balloon-cluster identification relied partly on visible dangling strings that changed shape and count across frames, a morphological tell that was recoverable even from grainy IR video.
• The alternative quadrotor-UAS hypothesis originated from one of AARO's own interagency partners, indicating that trained analysts initially disagreed on the attribution before the IR heat-signature argument settled the question.
• Color temperature was artificially assigned to the still frames in Figure 1 to aid visualization, the raw sensor output was grayscale IR, not color imagery.