Tokyo, Japan, March 1955March 1955

On March 20, 1955, radar operators at Johnson Air Base near Tokyo detected something unusual. A controller working the CPS-1 radar system, which normally handled routine airway traffic control, spotted what he believed to be sixteen unidentified targets moving in formation north of the base. The targets appeared at ranges of roughly 20 to 25 miles and seemed to be drifting southeast. The controller notified another radar station called Butterfly and requested that an F-86D fighter be scrambled from Yokota Air Base to investigate.

The F-86D pilot, 1st Lieutenant C. G. Du Marrick, took off at approximately 1538 (3:38 p.m.) local time and was vectored by ground control toward the reported targets. Though Butterfly radar could not paint the original sixteen targets, the pilot's own airborne radar eventually picked up two blips at close range. He pursued one target, maintaining contact through a series of turns and altitude changes over roughly fifteen minutes. The pilot described the radar returns as sharp and well-defined, similar to what he would expect from another aircraft. He maneuvered to within minimum firing range, around 2.5 to 3 miles, and reported that the target's radar presentation behaved normally throughout the engagement. The target eventually broke contact and the pilot was vectored back to base.

The investigation revealed several complications. The ground radar observation of sixteen targets could not be correlated with the pilot's pickup of only two. Controllers working the ground radar were unfamiliar with the normal radar clutter (unwanted signals from terrain and weather) in the northern sector, having worked primarily with southern approaches. Weather data in the file showed a pronounced temperature inversion between 5,000 and 10,000 feet, which can produce false radar echoes from ground targets. An intelligence analyst noted that the apparent stationary position of the targets later in the sequence, combined with the sharp radar definition in an area not usually monitored, suggested the possibility of ground clutter misidentification. However, the analyst also observed that this was a detailed and well-prepared report documenting an unusually prolonged radar pursuit, and that the inconsistencies between the ground and airborne observations could not be satisfactorily explained with the available data.

The Air Force evaluation listed in the case metadata is "unknown." The full case file, comprising 10 pages, is reproduced below as held by the National Archives.