Japan.Haneda AFB · Radar-Visual over Tokyo

Around 23:30 local time on 5 August 1952, two operators of the control tower at Haneda Air Force Base (Tokyo's airport, then under American operation) noticed, while crossing the ramp on their way to their shift, an exceptionally bright light over Tokyo Bay, northeast of the field; once in the tower they pointed it out to two colleagues, and the four watched it through 7x50 binoculars for between half an hour and an hour. According to the Far East Air Forces intelligence report (IR-35-52), they described a circular blue-white light of constant brilliance that seemed to be part of a darker round shape about four times larger, with a few dim lights along the lower edge; but the report itself notes it could not be determined whether that dark area was due to a body or to shadow or a blinding effect from the light. The object faded twice to the east and came back; no sound was heard and no rotation seen. Tachikawa tower, some 30 miles to the west, called to point out the same light over the bay, and the pilot of a C-54 entering the Haneda pattern, asked to look at it, replied that it looked like a brilliant star and dismissed it as such.

The tower alerted the ground-controlled intercept (GCI) centre at Shiroi, of the 528th Aircraft Control and Warning Squadron. The first controller saw only 'three or four blips' on the indicated bearing, with no definite movement; his relief, Lieutenant Albert, tracked a target he described as small and relatively weak but sharply defined, in a right-hand orbit at about 100-150 knots with occasional stops and a maximum of perhaps 250-300 knots; around 00:12 the return broke into three smaller contacts. An F-94 night fighter was scrambled (pilot Lieutenant Holder, radar operator Lieutenant Jones) and, delayed by fuel-system trouble, became airborne at 00:03. Vectored by Shiroi, at about 00:15-00:16 its radar operator picked up a weak contact at 6,000 yards, 10° port; in his statement he wrote that the target was moving rapidly from port to starboard and that 'a lock-on could not be accomplished'; after a turn the blip appeared to accelerate and disappeared from the scope in about 90 seconds. The crew saw nothing at any time despite 60-70 miles of visibility, and both the F-94 and the target were then lost in the Shiroi ground clutter. The tower kept seeing the light until about 00:50, when it was lost among low clouds moving over the bay.

The officer who prepared the report judged all the witnesses serious and called it the first sighting in the Far East combining visual and radar contact over an extended period, but he himself listed possible explanations: full-moon reflections off clouds for the light, and a blinding effect for the dark shape. At Wright-Patterson, Blue Book's radar specialist, Captain Roy James, wrote on 15 September 1952 that the report did not allow a plausible explanation of the targets and that, without temperature and moisture data versus height, no anomalous-propagation estimate could be made, though he noted the targets appeared at times to be permanent echoes. The case card was marked 'Unknown'. The Condon Report (1968), which reviewed the file, argued that the light and the radar return were not the same object: the light remained visible for at least five minutes after the F-94's radar lost the target, and no one at Shiroi saw it although it should have been to their south. It proposed the star Capella, 8° above the horizon at 37° azimuth at midnight, distorted by the atmosphere, and anomalous propagation for the returns, inferred from their coming and going, the break into three and the lack of correlation between the ground and airborne radars. That argument rests on the tower's timeline, and the file itself is not consistent on that point: the summary of the FEAF report says that 'Soon after loss of radar contact by Shiroi and the F-94, the tower lost visual contact as the object either faded, moved away, or was obstructed by clouds', whereas the tower statements keep it in view until about 00:50. Condon, moreover, conceded that the precise optical mechanism that would have produced such a strangely diffracted image 'must remain conjectural' and that, either way, 'the phenomenon must be quite rare'. An ephemeris calculation for Haneda confirms Capella's position: it was rising from about 4° at 23:30 to about 12° at 00:30, on the 30-50° bearing the report gives, and one of the operators noted at the end that the light seemed 'somewhat higher in the sky'. The same calculation shows Venus was below the horizon that night, so the bright planet almost due east with which the witnesses compared the light, and which all of them took to be Venus, must have been Jupiter.

Haneda is often cited as a triple-channel radar-visual case (eyes, ground radar and airborne radar on the same target) and the primary file lets us measure how far that label holds. The three channels existed and were documented by witnesses the report itself judges serious and experienced; the case is still marked unknown and Blue Book's radar specialist found no explanation for the returns. But the file does not support the three channels coinciding: the F-94's radar never locked on, the crew saw nothing in exceptional visibility, the tower, by its own statements, kept seeing the light after the radar had lost the target (although the FEAF report's summary places the visual loss soon after the radar loss), and the light, low in the northeast and also visible from Tachikawa 30 miles away, sat in the position of a bright rising star.

That makes the case instructive rather than conclusive. It shows how a radar-visual can be assembled by adding channels that are not observing the same thing: a low star seen through binoculars across humid air, and weak, intermittent returns on a search radar on a night with no sounding that could rule out anomalous propagation. At the same time, the Condon Report's composite explanation is inferential (there is no temperature profile for that night, and the report itself admits the optical mechanism 'must remain conjectural' and that the phenomenon 'must be quite rare') and the 1952 radar specialist himself could not close out the returns, so a residue of indeterminacy remains in the radar channel. The context matters too: Tokyo in 1952 was the rear area of the Korean War, with air defences on real alert, and a Soviet reconnaissance aircraft was the first operational hypothesis; nothing in the file supports it, since the returns moved at 100-300 knots in tight orbits near the coast and no aircraft was seen.