SO FAR IN 1962, EVEN AS THE UNITED STATES CONTINUED TO RAMP UP ITS nuclear testing activities, the Soviets had been unexpectedly quiet. They had conducted an extremely low yield test at the beginning of February, but nothing since. No one expected this state of affairs to continue for very long. Obviously they were simply analyzing data from their 1961 shots and making preparations for a fresh round of shots.
Which didn’t mean that the Russians were ignoring US activities. They might be unable to snoop around underground tests in Nevada, but DOMINIC in the Pacific provided a new opportunity to monitor American activity. “There was a Soviet ship, well known to us,” Mustin recalled. “It wasn’t a trawler; it was a big ship that cruised slowly through the Pacific … it was an intelligence-gathering ship with massive facilities for electronic interception and so on.” The vessel had been tracked by Navy patrols at least from the vicinity of Midway Island, and Mustin’s planes kept watch on her as she approached Johnston Island. The Soviet ship, however, was careful to stay outside of the restricted waters surrounding the testing sites, an easy task since the boundaries had been made public to all marine traffic as a safety measure. “They stayed out there for a while,” Mustin said, “long enough to confirm, to their own satisfaction, that we were detonating devices over there. She was about 500 miles away.” At least one of Mustin’s P2V aircraft stayed in sight of the Russian spy ship continually, rotating with other planes based on Christmas Island.
It was a hazardous duty, stretching the P2Vs to their limits of range and endurance. One plane was almost lost at sea when it experienced an engine fire. “Those men flying those long over-water patrols were really facing a very real risk every minute they were airborne,” Mustin remembered.1
JTF-8’s scientific deputy, William Ogle, noted that while the British were permitted to make scientific measurements of the Christmas Island shots as part of the agreement for letting the United States “borrow” the island, no such arrangements existed with the Soviets. “The Russians also made measurements with no formal agreement,” he wrote later. “The Task Force and the Commission discussed the subject, but there was really nothing we could do about it except watch and keep fairly close by to make sure that they did not come within the danger area.” Sometimes it was difficult to be quite so tolerant. At one point during a stopover in Fiji, JTF-8 commander General Starbird was dining at a small restaurant and happened to notice the captain of the Russian ship relaxing nearby, apparently taking a break with some of his officers. “Starbird did not want to chance an international incident, and therefore, did not have a discussion with the Captain,” Ogle noted.2
Unwanted audience or not, the Fishbowl part of DOMINIC had already begun unofficially on May 2, with the launch of an unarmed Thor missile from Johnston Island. It had essentially been a full dress rehearsal for all the various tracking, monitoring, and experimental stations on land, sea, and air. One of the primary reasons that the Thor missile had been selected, aside from its supposed reliability, was that it had the ability to carry pods which could be released from the missile at predetermined altitudes. Each Thor would carry three of the two-meter-long pods, which contained instrumentation and experimental equipment. After ejection from the booster, the pods would make soft landings by parachute in the open ocean, where they would be retrieved by Navy ships and Marine helicopters. At least, that was the plan.
The May 2 rehearsal, named Tigerfish, went reasonably well, although two of the three instrument pods encountered problems with their recovery systems and were damaged. More dry runs for the various support forces, without any missile launches, followed over the next several weeks. Finally, on June 2, the first nuclear shot, Bluegill, was ready to go.
Shortly after midnight, the Thor missile bearing the Bluegill device was launched. All was going perfectly well at first. The main radar tracking facility for the Johnston Island launches was a ship with the appropriate if not picturesque name Range Tracker, which was operated by the Air Force with a civilian crew as part of its Pacific Missile Range. Moored to the Johnston Island pier less than half a mile from the Thor launch pad, the Range Tracker had to follow the missile, which was shot into the skies nearly vertically, from almost directly beneath it. “This meant that as the missile climbed on up to hundreds of miles in the sky, with horizontal separation at the ground of a couple of thousand feet, it really required the radar to track very near the zenith, which is a tough problem for a radar,” Mustin explained. Somewhere along the way, as Bluegill approached its highest altitude, the Range Tracker lost it.
That was a major problem. Somewhere high above the Pacific Ocean with ships and airplanes spread out in all directions, a nuclear-tipped missile was reaching the peak of its ascent as its engine burned out, preparing to plummet back to Earth under the inevitable command of gravity. Without a radar fix, there was no way to know precisely whether the missile was still following its planned trajectory, and thus no way to know where the missile—and its live nuclear warhead—was going to fall. One does not take chances with live nuclear weapons, and there was no time to consider the problem at leisure. JTF-8 commander Starbird gave the only possible order. The range safety officer pressed the destruct button, and the Thor missile exploded, taking Bluegill with it. There was no nuclear explosion, so the entire experiment was a failure. For now at least, the Bluegill shot would have to wait.
Adding to the general consternation over the loss of a valuable missile and even more valuable (but more easily replaceable) nuclear warhead—not to mention the failure to collect any of the desired data—was the discovery that Bluegill had actually been on its proper trajectory, as subsequent data analysis proved. Although there had been no way for Starbird, the range safety officer, or anyone else to know it at the time, there was no need to press the destruct button. Starbird reported as much to Washington in a terse cable two days after the incident: “Preliminary evidence available indicates that the Bluegill missile probably flew a normal trajectory and, if we could have known this, the detonation could have been made to occur in a place to give a safe firing and successful data.” Although he noted that “the warhead-missile system had extensive protective systems in it to prevent any except a high-altitude [nuclear] burst,” he had decided that “if I could not know from satisfactory tracking that the trajectory could not target near the ship array, I would not authorize arming of the warhead. Unfortunately, the missile tracking system lost the track before we could secure guarantee of safe trajectory. The missile was not commanded to arm therefore but was commanded to destruct.” Looking on the bright side, he noted, “the fact that … the destruct worked in Bluegill as commanded, indicates the firing safety system is sound.”3
It was a chastening situation for all concerned. “There was plenty of agonizing going on, including some by me after I got into the act, to try and figure out what had gone wrong and what we could do to safeguard ourselves against this,” Mustin remembered. One obvious solution was to have more radar coverage. “It turned out that we had a number of resources, right there on the island, which indeed had been tracking this thing throughout its flight. If only they had been asked, they could have told us that it was exactly on course … but nobody had thought of that or made any arrangement of that nature.”4
Also disappointed were the people who had gathered on the beach at Honolulu watching for Bluegill’s nuclear light show. In any case, an official statement assured civilians that the mishap would not “cause hazardous levels of radioactivity in the water” or “constitute a hazard to human health.” Perhaps the best epitaph on the anticlimactic debut of Fishbowl came from Ogle as Bluegill’s pieces fell into the Pacific: “Best damn dry run we ever had.”5
While the three instrument pods were recovered from the missile (two of which, it was discovered, had failed to eject properly), nothing else was retrieved. There was nothing more to do but pick up the administrative and operational pieces and continue on. The next shot would be Starfish, planned for June 19. The intervening time would be spent in further rehearsals, including better tracking procedures to prevent a repeat of the Bluegill fiasco.
On the night of June 19, another Thor sat poised on the Johnston Island launch pad, ready to boost Starfish into space. Two of the three experimental pods were replaced by dummy RVs (reentry vehicles) intended to test the effects of X-rays from a nuclear blast on warhead materials, with an eye to decreasing the vulnerability of US warheads. Everyone was optimistic that the snafus of Bluegill had all been adequately addressed.
A few minutes before midnight, Starfish was launched. It flew straight and true into the night for just under a minute, then it all went wrong. The Thor missile sputtered and flared and began going out of control. This time, it was clear that something was dreadfully amiss, and at sixty-four seconds into the flight, the range safety officer again hit the button to destroy the missile and nuclear warhead.
It had still been early in the missile’s trajectory, however, and it had not yet moved very far from the vertical above Johnston Island, ending its existence at an altitude of only 30,000 to 35,000 feet. Debris from the explosion fell onto Johnston Island, causing no major damage but creating quite a mess. Parts of the missile, one of the RVs, and the remaining instrument pod were scattered over the island, and more debris fell into the lagoon and surrounding waters of Johnston Island. As the Navy recovery divers soon discovered, some of the junk was contaminated with plutonium from the Starfish warhead.
The cleanup effort consumed the next several weeks. Analysis of the debris along with other data revealed that one of the dummy re-entry vehicles had been the culprit, disrupting the missile’s exhaust flow and weakening it until the Thor’s engine tore loose and collapsed into the fuel tanks.
The other DOMINIC shots were proceeding apace, but Fishbowl was beginning to look decidedly jinxed. Fixing the problem that had caused the destruction of Bluegill had done nothing to save Starfish from the same fate. And only a limited supply of Thor missiles, instrument pods, and other necessary equipment was available, so continuing to lose them at this rate for no appreciable return was obviously unacceptable.
The displeasure wasn’t simply felt by the test personnel, but extended all the way to the top. “These two failures presented us with the problem of what to try next at Johnston Island,” AEC chairman Seaborg remembered.6 Because Starbird had made it clear that to allow time for all the necessary preparations it wasn’t possible to conduct further missile shots closer than fifteen days apart, and because Fishbowl had already been operating for more than a month with nothing to show for it, President Kennedy was beginning to ask embarrassing questions, with the implied possibility that the rest of the high-altitude program might be curtailed or even cancelled entirely.
At the end of June, Seaborg visited the Pacific test sites along with several aides and JFK’s national security advisor McGeorge Bundy. They watched a shot at Christmas Island and then proceeded to Johnston Island to tour the facilities and, perhaps, to provide a little incentive and encouragement. “We saw the launching pad and the complicated diagnostic facilities operated by the Air Force and several laboratories,” Seaborg wrote.7 Upon his return to Washington, Seaborg sent a message to General Starbird, telling him not to rush the upcoming second attempt at Starfish. The luxury of failure, always a fragile thing in the best of circumstances, was gone. For the next shot, everything had to work.
As discouraging as they may have been, at least the failures thus far encountered hadn’t resulted in casualties to anything other than equipment. For all aspects of Operation DOMINIC, safety had been a prime consideration from the first planning sessions, and continued to be emphasized throughout all operations. No shot, whether on or near the ground, high above, or in space, would be fired unless all weather conditions were good and the potential affected area had been swept clear of any unauthorized or unneeded personnel.
That was relatively easy for the more conventional tests; whether dropped from aircraft, carried aloft by tethered balloons, or detonated underwater, no human being was required to be nearby when the weapon burst into life. But the Fishbowl shots posed a unique problem. Someone had to be there on Johnston Island to prepare and launch the Thor missiles needed to carry the nuclear devices into space. And liquid-fueled missiles such as the Thor were infamous for blowing up when no one wanted them to. Even the crew of a B-52 bomber, sitting mere feet away from a megaton weapon cradled in their bomb bay, were safer than a crew working a half mile away from a megaton warhead sitting atop thousands of gallons of volatile, highly explosive liquid fuel. The rain of debris that had fallen all over Johnston Island from the failed Starfish test was a grim reminder of the possibilities.
There was only one option. “Our practice was to evacuate the island, the late afternoon immediately before the shot, of everyone except those absolutely required for the conduct of the test,” Mustin said. “It soon turned out that the only way to do this was to evacuate them by helicopter to a ship.” Because there might be as many as a thousand people on the island during test activities, that required a large ship, preferably an aircraft carrier.
Mustin first secured the Iwo Jima, which was later replaced by the Princeton, both carriers that hosted Marine helicopter squadrons to handle the evacuations. “These Marines really made a precisely organized and executed operation of evacuating 2000-2500 [sic] men off the island in a couple of hours,” remembered Mustin. “Of course, it required that the people being evacuated respond precisely to organization and discipline. And this was not something they were accustomed to. They were a heterogeneous bunch of stevedores and cooks and bakers and scientists and engineers, a certain proportion of military, and so on. But the need was apparent and the mechanism was there. The Marines knew exactly how to do their part. It really went beautifully, and we never had an accident in quite a number of evacuations of the island.”8 Other documents cite the usual complement of evacuees as only around 800, but whatever the number, it was nonetheless an impressive operation.
Once an evacuation was completed, the carrier would move away from the island and out to sea, clear of the trajectories of the warhead-carrying Thor as well as the various instrumented rockets that would be fired for the test. “The main danger, we were told, would not be from the nuclear explosion, but from the barrage of instrumented Nike missiles which would be launched to take readings on the detonation,” recalled one test veteran. “The impact points for these missiles were unpredictable.”9
Aside from observing the shot, there was little to do. “Many of the people brought to the ship were relatively highly paid scientists, engineers, etc., who had nowhere to go and nowhere to spend their salaries,” observed a website for Johnston Island veterans. “Rumor has it that many high stakes poker games could be found on the USS Princeton as these folks waited for the all clear signal to return to their work stations on the island.”10
From its safe vantage point, the carrier became the main observation point and command post, with constant secure communications channels to Washington. Before Fishbowl finally came to an end, the wisdom of the pre-shot evacuation policy would become even more evident than after Starfish.
The second attempt at Starfish, now called Starfish Prime, was ready to go on July 4, a coincidentally appropriate date to get Fishbowl back on track with a huge fireworks display. Unfortunately for happy coincidences, the winds were too high for a safe Thor launch, and the shot was postponed. For the next few nights, the weather was too cloudy over Johnston Island and vicinity, precluding the necessary observations and measurements of the high-altitude fireball. Postponements on account of weather were par for the course in nuclear testing, but in this case, they were becoming particularly frustrating.
Finally, just after 11 PM on July 9, Starfish Prime was launched. Starbird was reportedly so nervous about the outcome that he refused to watch the closed-circuit television of the launch at his command post. But for once, he had nothing to worry about. Everything went perfectly: the launch, the Thor missile, the trajectory, the radar tracking, the launch of the experimental rockets, the positioning and operation of the observation ships and aircraft. Precisely as planned, the Thor missile reached an apogee of about seven hundred miles, then nosed over and began its ballistic trajectory back to Earth. Finally the actual detonation of the 1.4 megaton Starfish Prime device took place at an altitude of about 250 miles.
What followed was perhaps the greatest light show ever seen over the Pacific Ocean. Starfish Prime lit up the entire sky from horizon to horizon, from Hawaii all the way to New Zealand, opening with a flash of white light brighter than the sun, which gave way to intense auroras, shimmering curtains of green, yellow, and red—an awesome, primal spectacle of energy on a scale that dwarfed into insignificance all those who witnessed it.
Mustin happened to be 1200 miles away on Christmas Island during Starfish Prime, but he didn’t miss a thing. “I just went out and sat on the edge of the sand dunes there and watched the sky,” he recalled, as above him was “a complete unbroken sheet of yellow light.”11
The civilian spectators back in Hawaii, who had been so disappointed when the previous shots had turned out to be duds, finally had their patience rewarded. Some had been waiting on the beaches, while others were guests of several hotels offering rooftop bomb-watching parties, just as some Las Vegas hotels had been doing for years with the Nevada shots. “Except those in a few isolated areas where rains blotted out the sight, residents on all the major islands of the state saw the explosion,” reported the New York Times. “The most fortunate observers here were those on the heights overlooking Honolulu. They had a clear field of vision when the blast illuminated the horizon south of the city. The brilliant flash sent yellow fingers stabbing through the broken clouds. Some observers believed the color was greenish. The sky turned pink and then a tomato red. It was Hawaii’s second sunset of the day and was visible for nearly seven minutes.” The paper also featured before and after shots at Waikiki Beach, showing clearly how the night had temporarily become day.12
At Kwajalein Island, 1400 miles west of Johnston, an Air Force major reported that “a brilliant white flash burned through the clouds rapidly changing to a green ball of irradiance extending into the clear sky above the overcast. From its surface extruded great white fingers, resembling cirro-stratus clouds … to be replaced by spectacular concentric cirrus-like rings moving out from the blast at tremendous initial velocity, finally stopping when the outermost ring was 50 degrees overhead. They did not disappear but persisted in a state of frozen stillness. All this occurred, I would judge, within 45 seconds. As the greenish light turned to purple and began to fade at the point of burst, a bright red glow began to develop on the horizon at a direction of 50 degrees north of east and simultaneously 50 degrees south of east expanding inward and upward until the whole eastern sky was a dull burning red semicircle … obliterating some of the lesser stars. This condition, interspersed with tremendous white rainbows, persisted no less than seven minutes.”13 Reports from the launch site back at Johnston Island were similarly vivid, describing multicolored glows, streamers, and discs in the heavens.
Other reports came from civilian planes, such as a Canadian Pacific Airlines in flight to Sydney, Australia. Seeing the sky to the north come alive with fiery auroras, the captain turned the plane to give his passengers a better view. In New Zealand, Fiji, and all over the Pacific, observers marveled at the brilliant flash and the active colorful auroral display that followed. “An interesting side effect was that the Royal New Zealand Air Force was aided in anti-submarine maneuvers by the light from the bomb,” noted a later technical report.14
There were also official technical observers spread across the Pacific, manning over two hundred stations in Samoa, Fiji, Okinawa, Wake Island, and other remote spots. A New York Times reporter was stationed with the Samoa group. “Word of the explosion was greeted jubilantly—and with a sense of relief,” he reported. The hours leading up to the test were lazy and relaxed, as the men lounged about house trailers and tents with their instruments. “As the hot and humid afternoon wore on, scientists, technicians and military observers performed the few routine tasks in preparation for the big show, as they had done twice before in vain. Then most just sat and talked or read or had a beer or two as the countdown sounded over the shortwave radio.”15
And Starfish Prime was felt and observed far beyond the Pacific. In Boulder, Colorado, needles on scientific instruments monitoring magnetic atmospheric and earth currents at the National Bureau of Standards were pegged violently when the device detonated. Similar effects occurred in other labs across the country. “Many American scientists … were amazed at the intensity of the long-range effects,” noted Walter Sullivan.16
Unlike the situation during Argus, when the number of satellites circling the planet could be counted on one hand, the orbital population had increased considerably during the past four years. Among them was Ariel I, the UK’s first satellite (though launched in collaboration with the US at Cape Canaveral), and Injun 1, the latest satellite from James Van Allen and his crew at the State University of Iowa. Injun 1 would follow in the footsteps of its predecessor Explorer 4 to provide data on Starfish Prime, while Ariel, as well as several other satellites, would soon suffer a less dignified fate.
Not everyone who witnessed Starfish Prime that night responded with awe and wonder. Many natives on various Pacific islands reportedly feared that the sky was literally falling on them, fleeing into churches or other refuges. “Crazy white man!” one Samoan exclaimed. Military and test personnel were supposedly obliged to reassure panicky people. Noted the Times reporter on Samoa, “Observers not connected with the nuclear test program felt that much of the commotion among the natives could have been avoided if the Samoans had received more advance information on the nature and purpose of the nuclear experiment.”17
Some reactions were less primal and far more political. The American embassy in London was again besieged by demonstrators from the Campaign for Nuclear Disarmament, as it had been since the opening of DOMINIC. The President of Ghana protested to Kennedy in a formal diplomatic note: “Is there no way wherein you … can resolve the present cold war anarchy? Is it not real wisdom to suspend this dangerous arms race until efforts being made for peace … can result in a general disarmament treaty?” Moscow, predictably enough, condemned Starfish Prime as a “crime” committed by US “atom-maniacs.” Proclaimed the Soviet news service TASS, “The United States exploded a nuclear device in space despite strong protests of all mankind … [and] confirmed yet again that it is following the course of whipping up the nuclear arms race.” Meanwhile, noted the New York Times, “In contrast to the Johnston Island nuclear shot of 1958 [a reference to Teak], which took Hawaiians by surprise, there was little criticism today of the hydrogen bomb explosion [in Hawaii].”18
Regardless of whether the rest of the world was delighted by the spectacle or outraged by its political ramifications, the US test authorities, military, scientists, bomb designers, and support personnel were all enormously pleased. There had been a few minor glitches, of course; a couple of instrument rockets had failed, and one or two experiments hadn’t worked here and there. But all of the major elements had gone well. The smoothness of the entire operation seemed to wipe out all the frustration and disappointment of the previous failures.
It soon became evident, however, that smooth and spectacular as it was, Starfish Prime had been something far more than just another nuclear test shot. Even as the magnificent visual displays diminished and then disappeared from the night skies, and all those involved in the operation congratulated one another and basked in success, the effects of Starfish Prime were not about to leave the stage quickly or gracefully.
Nor were the demons of Fishbowl banished for good. They were, in fact, about to become far worse.