Japan.Okinawa · a radar tracks something no one saw in the sky

In October 1951 Okinawa was under U.S. administration; today it is Japanese territory. On top of a 750-foot hill a few miles north of Kadena Air Base, radar Site 51 of the 529th Aircraft Control and Warning Group watched the sky and guided fighters toward whatever it detected. Its search radar sent out two beams, a high one and a low one, shown on separate screens: on the night of the 3rd the high-beam screen was set from 0 to 120 miles, and the low beam's long-range screen from 110 to 230. The antenna turned once every 15 seconds, and on each turn whatever the radar detected showed on the screen as a spot of light.

At 10:27 p.m. on October 3, 1951 (1327 Greenwich time), a sergeant watching the high-beam screen saw a blip appear at a bearing of 14° and 90 miles. He stated that it was easily seen but not brilliant, and that had it been brilliant it would have looked about three times larger than a B-29's; he described it as sausage-shaped and straight, some 2 miles wide by 4 long, and always set crosswise to the direction it was moving in. Between two turns of the antenna, within 15 seconds, it changed direction without the turn being seen, and moved away northwards. He noticed no change in speed, size or shape, and the last of the six positions, at 340° and 152 miles, was the brightest. The report warns that the time of each position was not written down ("No written record of the time of detection of consecutive plots was made"), so its table, which spaces the first five 15 seconds apart from 1327:00 and gives no time for the sixth, is a reconstruction.

When the blip left the high beam, a private apprentice operator picked it up on the low-beam screen, at 355° and 200 miles. He saw it four times, at 200, 210, 225 and 247 miles: brighter than an average blip, 2.5 to 3 times larger than a B-29's, solid, 2.5 to 3 miles wide by 6 to 8 long and also crosswise; at 247 miles it "disappeared into the second transmitted pulse". A sergeant radar mechanic watching the same screens confirmed what the two operators said and added that, about the time of detection, the permanent echoes, the blips that are always on the screen, began to brighten, particularly Yoron Island at 54 miles, and that shortly after the object vanished part of Okinoerabu Island, at 83 miles, showed on the low beam. The blip never appeared on the low beam's short-range screen.

That evening the two screens had been checked and oriented to true north. They showed no rain; before the detection two ships had been tracked at 55 miles, and returns from the sea reached out to 35 miles. During the same session the radar had followed an inbound B-29 out to 150 miles and an outbound C-54 to 165, both at 9,000 feet, and according to the report it was working normally. Nobody tried to see the object or send fighters after it. The attempt to pass the data to the Itazuke air defense center failed because of heavy radiotelegraph interference, recorded at 1350 (Greenwich time) on 5,850 kilocycles. Lieutenant Colonel T. D. Bradley, chief of air defense at Twentieth Air Force, was notified at 1335, and the radio alert to the intelligence staff of Far East Air Forces (FEAF), the U.S. air command in the region, went out at 0135 on October 4.

The site's report, dated October 8 and signed by the senior controller (name blacked out), estimated 4,800 miles per hour in the first minute, from five positions the high-beam operator had marked in grease pencil on the screen without noting the time of each. It adds that this operator did not see the blip between 112 and 152 miles, which he put down to readjusting the range of his screen, and that the low-beam operator did not see it between 225 and 247 miles, although the beam swept that area at least once; he too readjusted the range of his screen, which had been set from 110 to 230 miles when the blip appeared, while it was passing through that stretch. It rates the two sergeants "extremely reliable" and the private "reliable". It also records which radars were working at the neighboring sites: at Sites 52, 54 and 55 only one of their sets was working, and at Site 53 both were.

Twentieth Air Force held the report on October 9 pending an analysis that tried to support the theory that the track was that of a natural object rather than an aircraft or a missile; it required a special calibration of the radar that had not yet been done. Neither the calibration nor the analysis is in Blue Book's files. There are, however, two studies that disagree. Wallace W. Bush, of FEAF's Directorate of Intelligence, debriefed the personnel at Kadena and worked out 2,520 miles per hour on the high beam and 3,088 on the low beam, a figure he judged "probably close to being the actual velocity". He explained the width of the blip by a wide object, "possibly several objects flying in a wide formation", and calculated that it left the high beam at 129 miles ground range and 49,000 feet and entered the low beam at 194 miles and 45,000 feet. It was therefore descending about 3,000 feet a minute, and the turn would have been gradual, at 2 or 3 times the force of gravity. Bush also rejected the site's recent calibration chart, which would not allow anything to be followed on the high beam beyond 80 miles when this blip was followed to 152, and redrew it.

Bush added that Sites 52 and 55, closer to the path, detected nothing at that time, although "the Bogie passed over site #55 at approximately the same time the initial plot was obtained by site #51": the unknown passed over Site 55 at almost the moment Site 51 picked it up. Site 54, southwest of Site 51, did not see it either. Interference from an unknown source was frequent on that network, and he could not determine whether it was a deliberate attempt to block the report. No one among the local population had seen anything unusual, and a check the next day found the two beams to be off in bearing by "several degrees".

The second study, by Frank C. Anton, a technician from the Philco company, rejected the natural object because of the nearly constant speed, the turn of about 90° within 15 seconds, the change in the blip's shape when the heading changed, its lack of fluctuation and the fact that two closer sites did not see it. He concluded that the signal probably came from the object itself: a manufactured missile, "possibly guided", carrying a beacon, a transmitter that answers the radar's pulse with a signal of its own, in the S band. That would explain why Sites 52 and 55, working in the L or P bands, did not see it. He granted that these were "merely assumptions", but asked for the report to be studied in much greater detail ("are great enough to warrant further study of the report in much greater detail"). Bush found the beacon hard to justify, because the site's beacon receiver was not connected to the screens on which the object was marked.

FEAF gathered everything in report IR-25-52 of June 26, 1952, prepared by Captain Charles J. Malven, who asked to be sent any significant findings from its evaluation. On September 19 the Air Technical Intelligence Center (ATIC) passed it to its electronics branch as an "extremely detailed" report that should prove interesting "as a possible unidentified aerial object". On October 1, 1952 Lieutenant Colonel H. C. Johnston replied that, since two radar experts had advanced conflicting explanations, further theorizing would be "pure conjecture". The Blue Book record card lists it as a ground-radar observation from a military source and concludes "UNIDENTIFIED".

A military radar, with three men operating and watching it, left a track of a little over two and a half minutes, by the antenna turns Bush counted, with an untimed gap between 152 and 200 miles, and the Air Force studied it for a year without being able to explain it. There is no visual witness, no photograph of the screen and no time for each position, and the two technical studies reached opposite conclusions: a real object, perhaps a formation, descending at some 3,000 miles per hour, or a signal sent out by a missile carrying a beacon. The natural-object analysis that Twentieth Air Force announced, a third reading, is not in the files.

What weighs against an extraordinary object is in the same papers. No other radar saw it: not Sites 52 and 55, closer to it, although according to Bush the object passed over Site 55 at about the time of the first position, not Site 54, and not the height-finding radar at Site 51 itself; Anton puts that down to the frequency band. The speeds rest on pencil marks without times and on gaps of tens of miles in which the operators did not see the blip while readjusting their screens. The next day the two beams were found to disagree by several degrees in bearing, and the site's calibration chart was in dispute. The mechanic noted that the islands' fixed blips brightened at just that hour and that an island 83 miles away appeared on the screen; no document links this to the object.

Our own reading, which does not decide how the case is classified: islands showing up farther and brighter than usual is a common sign of anomalous propagation, when layers of air bend the radar beam and carry it beyond its normal range. Moreover, at 350 pulses per second the radar can only place correctly what lies within about 266 statute miles (own calculation): the echo of anything farther away comes back after the next pulse has gone out and appears folded in, at a shorter range than the real one. The blip was lost at the edge of that interval. But a folded echo keeps the speed at which something moves toward or away from the radar and, if anything, understates its sideways speed, so it does not lower the measured speed; and the folded blip of an island would not move (own calculation). Neither possibility on its own explains the shape or the turn the operators describe, and no one examined them in 1951 and 1952.