On 9 December 1953, six aircrew members flying training missions around Thule Air Base in Greenland reported a sequence of brief luminous objects. Their accounts are unusually useful because the original intelligence report preserves both the descriptions and the witnesses’ differing interpretations. Some considered the lights meteor-like; others believed they had seen flying objects. The investigating officer favoured an astronomical explanation, subsequently endorsed by higher headquarters.
One short sequence, several accounts
Captain Harold R. Miller, the group intelligence officer, compiled report IR-10-53 on 10 December from interviews with six observers. He grouped the accounts because they occurred over a short period and might be related. The report contains seven numbered narrative entries, including further sightings by people already interviewed. Seven entries therefore do not mean seven independent sets of witnesses, nor do they establish seven separate mysterious craft.
The observers belonged to the 318th Fighter Interceptor Squadron and were flying F-94 aircraft. The reported sequence runs approximately from 1230Z to 1300Z on 9 December, with individual views generally lasting only one to three seconds. The heading and position data refer to different aircraft at different moments. Compressing the material into a single long pursuit would erase the short durations and the changing observation geometry preserved in the original.
The first two crews disagree
The opening account came from Captain Irvin and his radar observer, Lieutenant Roberts. Flying south of the base, they saw an oblong, bluish-white light which seemed at times to have light shining through it. They compared its apparent size with a half-dollar held at arm’s length. There was no recorded tail, exhaust trail or sound. Its apparent course was south to north, and it disappeared abruptly after roughly two or three seconds.
The pilot turned towards the light’s general direction, but the report records neither a successful interception nor airborne radar contact. The crew considered it unlike astronomical or meteorological phenomena they had previously encountered. The intelligence officer nevertheless recorded their view as an interpretation, rather than converting the description into a confirmed object type. Their apparent-distance and speed estimates were not measurements from a radar track.
Lieutenant Zook and radar observer Lieutenant Zeiglar, in a second F-94 a little behind the first, reported a light at about the same time. It was intensely white, brighter forward, and round or slightly oblong in appearance. They saw it for about a second, moving generally south to north before going out without an afterglow. The investigator considered it probably the same object reported by the first crew.
This second crew interpreted the appearance differently. Zook thought it was a meteor at relatively close range, and Zeiglar thought it looked more like a meteor than a flying object. The report regarded both crews as generally reliable, so the disagreement cannot be reduced to an official endorsement of one set of witnesses as truthful and dismissal of the other. They agreed that something luminous had been seen while disagreeing about what it was.
Further white lights
Irvin subsequently described another round white light, seen during the same flight at approximately 1234Z. It appeared to follow a course broadly similar to his aircraft’s and lasted about a second. He turned towards it and lost sight of it, again without radar contact. He thought this might be the same object as before and interpreted the sequence as something orbiting in the area. That connection was his inference, not a demonstrated continuous track.
Zook’s next account concerned a white globe of light, less brilliant than the earlier one. It appeared below and south of his aircraft, travelled across his view and went out after a slight downward trajectory. He estimated the observation at about one second and evaluated it as a meteor. A further observation by Zeiglar around 1240Z described a small white light which grew brighter before going out after approximately two seconds. He too said it gave him the impression of a meteor.
These entries preserve important differences of apparent brightness, direction and duration. They also show why a summary saying merely that military pilots saw high-speed objects would be incomplete. The witnesses themselves supplied ordinary astronomical interpretations for several episodes. Their estimated speeds depended on assumed distances, and the available report does not turn those assumptions into instrumentally established velocities.
The flame-like observations
The third crew, Lieutenant Green and radar observer Lieutenant Dungan, reported a different appearance at about 1245Z while south of the base. Their light looked more like a flame, yellowish orange and comparable to the visible blast of a jet afterburner. They described a brief apparent passage towards and over the base, lasting about one or two seconds. Both thought it was a flying object rather than identifying it as a meteor.
Dungan then reported another similar appearance at about 1300Z, with his aircraft east of the base. This view lasted about a second. The pilot turned to attempt an interception, but none was achieved. Dungan said the object did not look like a meteor. Those views remain part of the case and should not disappear simply because the intelligence assessment adopted an astronomical explanation.
The final narrative mentions radar interference resembling ground harmonics. It does not report a stable radar target corresponding to the light. Across the detailed observations, the report repeatedly records no airborne radar contact or no additional corroborating information from communications facilities. Interference is therefore distinguished here from a radar detection; the case is not tagged as a confirmed radar tracking episode.
The official assessment and its limits
Miller recorded clear weather, no clouds, an unlimited ceiling and visibility of about ten miles. His overall evaluation was that the sightings were probably astronomical phenomena. A message from higher headquarters concurred with the base intelligence officer’s assessment, using the contemporary expression “meteorite shower”. The later Project 10073 card labels its first, short observation as an astronomical meteor.
The card summarises an oblong bluish-white light lasting two or three seconds; it does not reproduce all seven narrative entries or the different opinions. Its speed shorthand is also expressed in miles per hour, whereas the detailed first account gives an estimate in knots. Neither figure should be promoted into a precise measured performance claim. Reading the complete report is necessary to understand what the summary compresses.
The outcome used here is possibly explained. The meteor interpretation has direct support from several observers and the contemporary investigation, but the investigating officer’s wording was qualified. The reviewed file supplies neither photographs nor a demonstrated instrument correlation establishing the identity of every brief light. It supports an ordinary astronomical explanation without requiring either certainty for every episode or a claim of unexplained advanced aircraft.
Location, sources and related reading
The map uses an approximate Thule air-base area reference, based on GeoNames 3831683 at the publicly recorded base location, rounded to 0.01°. It is an anchor for the reported training area, not a plotted route or a light’s measured position. The case belongs to Greenland (GL), with timezone America/Thule. The historical base name is retained because it appears in the 1953 reports; the installation was renamed Pituffik Space Base in 2023.
The complete ten-page original Project Blue Book file was reviewed, including IR-10-53 on PDF pages 7–10, its concluding weather and assessment page at PDF page 2, the headquarters message at page 3 and the summary card at page 1. The current name is confirmed by the U.S. Space Force’s original renaming announcement. Later scanning and release dates are not incident dates.
For a related example of why aircrew testimony, radar availability and later official assessments need to be read separately, see Goose Bay, 1952. That Canadian case involves different people and episodes and is not evidence for the identity of the Thule lights.