On 22 September 1979, sometime around 3:00am local time, a US Atomic Energy Detection System satellite recorded a pattern of intense flashes in a remote portion of the Indian Ocean. Moments later an unusual, fast-moving ionospheric disturbance was detected by the Arecibo Observatory in Puerto Rico, and at about the same time a distant, muffled thud was overheard by the US Navy’s undersea Sound Surveillance System (SOSUS). Evidently something violent and explosive had transpired in the ocean off the southern tip of Africa.

Examination of the data gathered by satellite Vela 6911 strongly suggested that the cause of these disturbances was a nuclear device. The pattern of flashes exactly matched that of prior nuclear detections, and no other phenomenon was known to produce the same millisecond-scale signature. Unfortunately, US intelligence agencies were uncertain who was responsible for the detonation, and the US government was conspicuously reluctant to acknowledge it at all.

The United States established the Vela satellite network in the 1960s for the specific purpose of monitoring compliance with the 1963 Partial Test Ban Treaty. Though each satellite’s intended lifespan was only eighteen months, the units continued to detect detonations for years thereafter. Prior to the mysterious event of September 1979, the orbital surveillance system had successfully recorded forty-one atomic detonations, twelve of which were spotted by satellite Vela 6911.

Though the Vela satellites were bristling with atom-bomb sensing equipment, their most effective apparatus was each unit’s pair of aptly-named bhangmeters. These photodiode arrays were tuned to detect the one-millisecond burst of intense light created by a nuclear fireball, and the subsequent secondary light caused by the hydrodynamic shockwave of ionized air. The sensor’s engineers had been skeptical of its potential⁠— hence their decision to name it after the Indian variation of cannabis called “bhang”⁠— but the predictable pattern of bright flashes proved to be an extremely effective method for detecting atomic explosions from orbit. In over a decade of operation, the network of unblinking electronic eyes had yet to record a single false positive with the atomic-bomb signature.

Aerial photo of Bouvet Island
Aerial photo of Bouvet Island

Due to the satellites’ design and their distant orbit of 70,000 miles, technicians were not furnished with the exact location of nuclear events; the sensors could only narrow the area down to a 3,000 mile radius. Available data suggested that the 1979 Vela incident occurred near Bouvet Island, a frozen scrap of earth famous as the most isolated isle in the world. The tiny island was home to a Norwegian automated weather station, and in 1964 an abandoned lifeboat of unknown origin was found there, filled with supplies. But presumably the island was completely uninhabited at the time of the energetic event, meteorological automatons and enigmatic castaways notwithstanding.

When the technicians at the Air Force Technical Applications Center (AFTAC) first received the detection signal, they were not aware of the related observations from SOSUS and Arecibo. But the Vela report was strong evidence on its own: the signature was too unique to be explained by other phenomena, the flashes were orders of magnitude brighter than any non-nuclear source on earth, and the likelihood of both bhangmeters artificially producing the same specific pattern was vanishingly small. US intelligence concluded that a 2-4 kiloton nuclear device had likely been exploded between South Africa and Antarctica. No nations admitted responsibility for the covert test, but intelligence reports indicated that the most probable perpetrator was Israel, possibly working in cooperation with South Africa.

Upon receipt of the intelligence docket, President Carter called an urgent meeting in the White House situation room. His administration had placed considerable emphasis on nuclear non-proliferation, therefore the US would be expected to respond harshly to any confirmed atmospheric test. If Israel were linked to the covert explosion, the resulting trade sanctions⁠— or the refusal to impose them⁠— would be politically precarious for the President, particularly while campaigning for re-election. Though there was no reason to doubt the detection, President Carter ordered the creation of an advisory panel, with a special emphasis on seeking non-nuclear explanations.

In the subsequent weeks, the AFTAC findings and the resulting intelligence report were buried in a shallow grave of reasonable doubt. Although both bhangmeters on Vela 6911 had observed the alleged atomic event, they had recorded the flashes at distinctly different intensities. The elderly satellite’s electromagnetic pulse (EMP) detector had long ago failed, therefore it was unable to corroborate the observations. Vela 6911’s sister satellite hadn’t detected anything at all, though its working condition at that time was unknown.

The signature double-peak light pattern from a nuclear detonation (19Kt US test, 1952)
The signature double-peak light pattern from a nuclear detonation (19Kt US test, 1952)

In the days following the event, the US Air Force had deployed several research planes to scour the atmosphere over the Indian Ocean for telltale fission products. The mission didn’t detect anything unusual, but for reasons that are not entirely clear, the flights didn’t penetrate the low-pressure air mass where the explosion was thought to have occurred.

In spite of the lingering ambiguity, most experts still believed that a surface nuclear burst was the most probable explanation for the Vela alert. During the months of investigation, the committee was made aware of the SOSUS hydrophone recording of the blast, which had been found to be consistent with a small nuclear explosion at or near the Indian Ocean surface. Scientists at Los Alamos also made the connection between the Vela detection and Arecebo’s fast-moving ionospheric disturbance, though the researchers were not convinced that the coinciding events represented a nuclear test.

An additional item of interest was a flash of auroral light that appeared over Syowa Base in Antarctica a few seconds after the Vela event, reinforcing the possibility of an EMP burst. Nuclear bursts have been known to cause patches of artificial aurora, though these colorful displays are more often due to solar energy mingling with the atmosphere. Further circumstantial evidence appeared in the weeks that followed, including reports from a doctor in Western Australia who detected trace amounts of iodine-131⁠— a short-lived radioactive fission product⁠— in the thyroid glands of local sheep.

The committee investigating the Vela incident absorbed numerous presentations from defense organizations and scientists. In the summer of 1980, after convening on three occasions, the panel produced their final report to the president. Owing to the lack of radioactive fallout and the inconsistent bhangmeter data, the investigators were unwilling to conclude that a nuclear bomb was responsible for the alert. Instead, they suggested that a micrometeorite had struck the satellite, dislodging particles which had reflected light back onto the photosensitive instruments. Another theory they considered was that a lightning “superbolt” had mimicked the distinct nuclear bomb pattern. They ultimately rejected this notion, however, since the Vela flashes had 400 times more energy and 100 times longer duration than the most intense lightning ever observed. The panel declined to address the Arecebo and SOSUS observations, executing a nimble leap of logic whose subtlety and elegance was second only to sticking one’s fingers in one’s ears and going “la la la la.”

Typical aurora
Typical aurora

Panel member Luis Alvarez, a distinguished physicist, later defended the panel’s reasoning in discarding data that corroborated a nuclear bomb:

“…a scientific detective’s main stock-in-trade is his ability to decide which evidence to ignore. In our [Defense Intelligence Agency] briefings we were shown, and quickly discarded, confirming evidence from a wild assemblage of sensors: radioactive Australian sheep thyroids, radiotelescopic ionospheric wind analyses, recording from the Navy’s sonic submarine-detection arrays that supposedly precisely located the blast from patterns of sound reflected from bays and promontories on the coast of Antarctica.”

As is often true when a committee is urged toward a particular outcome, it seems that the investigators may have exaggerated the evidence that supported their goal and ignored all else, an unfortunate human shortcoming known as confirmation bias. Subsequent analysis by Stanford Research Institute scientists found that the probability of a meteoroid impact mimicking a nuclear bomb flash pattern was roughly one in one hundred billion. In short, the president’s committee had reached a conclusion that was about as credible as the notion that a passing alien spacecraft had triggered the bhangmeter. The panel’s findings were accepted by the administration, however, and since trade sanctions were generally ineffective against rogue meteoroids, the US government was able to justify inaction.

In the intervening years, a few new Vela-related details have surfaced. With the collapse of the South African apartheid in the early 1990s, much of the information regarding their nuclear weapons program was made public. Among these revelations were documents indicating that their first functional nuclear weapon wasn’t constructed until November 1979, two months after the Vela incident. Some have hypothesized that France or Taiwan may have instead been responsible for the covert test, but evidence for either scenario is scant and circumstantial.

In 1994, convicted Soviet spy Dieter Gerhardt claimed that the flashes were the result of “Operation Phenix,” a joint Israeli/South African weapons test conducted under the cover of bad weather. “The explosion was clean and was not supposed to be detected,” Gerhardt claimed, “but they were not as smart as they thought, and the weather changed – so the Americans were able to pick it up.” He did not claim to be directly involved with the operation, stating instead that he had learned of it though unofficial channels. Gerhardt’s description of the explosion as “clean” suggests that, if his account is accurate, the device may have been a neutron bomb: an atomic device with increased neutron radiation and decreased fallout. Israel has never openly admitted to possessing nuclear weapons, but in 1986 a former Israeli nuclear technician named Mordechai Vanunu furnished a London Times reporter with photographs and descriptions of Israeli atomic weapons. Shortly before that article was printed, Vanunu was abducted by undercover Israeli Mossad agents, and imprisoned for his treason.

Mockup of an Israeli nuclear bomb, photographed by Mordechai Vanunu
Mockup of an Israeli nuclear bomb, photographed by Mordechai Vanunu

Today a mountain of Vela-incident intelligence remains classified, but a few heavily redacted reports have been released by the US government. Although these documents indicate considerable internal disagreement regarding the cause of the double-flash signal, they offer little new evidence. In his 2006 book On the Brink, retired CIA spy Tyler Drumheller wrote, “My sources collectively provided incontrovertible evidence that the apartheid government had in fact tested a nuclear bomb in the south Atlantic in 1979, and that they had developed a delivery system with assistance from the Israelis.” Unfortunately he does little to elaborate on the event or on his evidence, except to state that the South African bombs employed a “highly accurate delivery system using gliders.” One factor which casts doubt on the South African covert test theory is the conspicuous lack of South African scientists disclosing their participation, even after the fall of the apartheid.

Perhaps one day, when the redactions have receded and declassified documents are disseminated, further light will be shed on the Vela incident of 1979. If the distinct double-flash pattern was not a nuclear detonation, the Vela event would represent the only instance in history where a Vela satellite incorrectly identified an atomic blast⁠— in which case the true cause may forever remain unknown and/or irrelevant. In any case, the flurry of falsifications and artificial investigations churned up in the wake of the incident clearly demonstrated governments’ unwavering willingness to renegotiate reality for political purposes, even in the shadow of a mushroom cloud.

Update, September 2018: Scientists have conducted a modern analysis of records of radioactivity found in the thyroids of sheep in Australia during the time of the Vela incident, and the study “removes virtually all doubt” that the Vela detection was a nuclear explosion. Specifically, it was likely a small 2-3 kiloton nuclear artillery shell (compared to the 15-16 kiloton Hiroshima bomb), similar to the United States’ “Atomic Annie”.