On the night of 7 February 1953, personnel at the United States Air Force’s Nemuro detachment in northeastern Hokkaido watched a bright light to the east. Some saw rapid red, green and white changes; others described something more like a star. They did not agree about whether it moved. During the same episode, an operator also saw a brief radar return that did not reappear.
The original file does not turn those observations into a secure radar-and-visual track of one object. Its summary card treats the visual light as probably astronomical and the radar event separately. Another assessment in the packet favours a lighted weather balloon. The surviving documents therefore preserve competing explanations, differing witness impressions and a timing problem, rather than a single fully demonstrated identification.
What the detachment reported
The eight-page released packet includes Air Intelligence Information Report IR 5-53, dated 26 February 1953. It describes five detachment members seeing a light which changed colour and appeared brighter than the evening star. The formal description places it roughly fifteen degrees above the horizon on a true bearing of 91 degrees, and gives a visual duration of about thirty-five minutes.
That bearing was more than a casual guess. Lieutenant B. D. Burnett described directing that the CPS-5 radar antenna be sighted towards the visible light so that its direction could be read. He reported a bearing of 91 degrees with an allowance of about two degrees either way. By contrast, he expressly described the light’s elevation as a rough estimate made without appropriate measuring tools.
Using an antenna to establish a visual bearing is not the same as detecting the light by radar. The distinction is essential here because the report also contains a separate, much more distant radar return. Treating the antenna-bearing procedure as a positive radar lock would join two different kinds of evidence without establishing their connection.
A controller who doubted the apparent movement
Burnett took binoculars to the top of the radar tower. He described an initially rapid alternation of colours, sometimes red and green together, followed by white with intermittent coloured flashes and eventually steady white. Those changes were visible to the naked eye and more distinct through seven-by-fifty binoculars.
He was more cautious about motion. The observers disagreed: some thought the light circled, others thought it stayed still, and others described orbital movement. Burnett’s own impression was that it did not move relative to nearby stars. He also said that a single light provided no dependable way to determine its range. This is a valuable contemporary qualification from a participant, not an explanation invented by later commentators.
The individual statements reflect that disagreement. Radar mechanic Thomas J. Reynal said the object looked very much like a star, but had an unusual sector of light extending from part of it. He did not personally see the colour changes reported by colleagues. Another mechanic described apparent orbiting and a cone of light, while operator William W. Nichols reported an impression of gaining altitude but could not estimate range or height.
The return that appeared only once
Radar operator Harvey W. Arnold described a pip at about 85 degrees and 165 miles. It appeared on a single sweep and was not strong or persistent enough to establish a height. Searching a wider sector did not recover it, so it was regarded as interference. Burnett’s statement similarly says he learned of the return on reaching operations, but that it never reappeared during the subsequent search.
The formal account uses nautical miles for the return’s distance. Whatever the exact range notation in a particular statement, the important limitation is unchanged: this was not a continuous track with repeated positions. The report does not demonstrate that the radar pip was the visible light. Its visual bearing was close enough to invite attention, but a similar direction alone cannot establish a shared identity.
The radar feature is retained in this case because an actual reported return forms part of the original evidence. It should be understood with the single-sweep and interference qualifications attached. The file does not support the stronger assertion that radar established the light’s altitude, speed or distance.
The times do not form a clean chronology
The covering description and formal time field give an observation beginning at 21:30 local time, with a thirty-five-minute interval. Several personal statements instead refer to about 22:30. A copied controller’s log records a notification at 22:20. A faint concluding note discusses the delay in gathering statements and indicates a correction towards 22:30 rather than 21:30.
These differences should be carried into any reconstruction. They do not create a second sighting on a different date, and they are not a reason to invent an exact start time. They do limit the precision of a later comparison with a particular star position or balloon launch. The date, location and broad night-time setting are firmer than a minute-by-minute sequence.
A star on the card; a balloon in the appraisal
The Blue Book summary card judges the visible light probably a star, referring to its position, apparent behaviour and intensity. The card names Deneb in the typed discussion and also contains handwritten Regulus annotations. These are proposed astronomical comparisons in the record, not independent observations of a named star being positively recognised by the witnesses.
A separate Far East Air Forces comment argues for a lighted weather balloon. It notes that the incident was near the usual evening period for optical weather-balloon launches, describes good visibility on clear nights and suggests a launch from stations on islands off Hokkaido. That is an assessment based on timing and appearance. The packet does not provide a matching launch register, a recovered balloon or a measured trajectory that confirms the proposed source.
The star and balloon assessments are both conventional proposals, but they are not interchangeable confirmations. The disagreement about motion matters to both; so does the inconsistent time. A reader should not have to choose between erasing those complications and treating an unidentified light as evidence of an extraordinary vehicle. The surviving file supports neither shortcut.
Map and documentary limits
The map uses a rounded official Nemuro city reference, not a radar tower or a reconstructed object position. The original material includes a regional sketch of the eastward line of sight; it is contextual evidence, not a modern survey of the observation point. No access to a former or current military installation is implied by the locality marker.
The lead image is the original government summary card. It is not a photograph of the light and cannot show whether a star or balloon was present. All eight scanned pages were reviewed, including the statements, the repeated regional diagrams and the separate appraisal. Private service numbers on statement sheets have not been reproduced in this article.
The outcome is recorded as disputed because materially different assessments remain in the reviewed record. That label does not suggest that either conventional explanation was disproved. It preserves the narrower conclusion the evidence allows: a changing light was reported, a fleeting radar event was recorded separately, and no independently demonstrated identification connects all the observations.