Although the May 1 shootdown ended U-2 overflights of the Soviet Union, the CIA still had a valuable asset. In early 1958, a U-2 spotted what the CIA would subsequently conclude was construction of Israel’s Dimona nuclear reactor.84 In August of that year (by which time the program had been renamed CHALICE), in the wake of China’s shelling of the Taiwanese-held islands of Quemoy and Matsu, U-2s from Detachment C flew missions over China to monitor People’s Liberation Army troop movements—and found no signs of preparation for an invasion.85
Still, it was decided that the CIA would “maintain a greatly reduced and redeployed U-2 capability.” Approximately half of the CIA’s inventory of twelve U-2s would be turned over to the Air Force, and the detachment in Turkey would be closed down. In addition, it was “strongly believed that an appropriate way should be found to use the U-2 to complete the coverage of primary targets in China before air defenses there have been further improved.” The project would also receive a new code name—IDEALIST.86
U-2s were not the only planes overflying China in the 1950s under the direction of the CIA. In 1954, a small group of CIA and Air Force officers met with Navy officials to arrange for the purchase of seven P-2V7 Neptune maritime patrol aircraft, which had a range of about 4,000 miles. The CIA had chosen the plane because of its capabilities as well as the existence of Navy airfields worldwide. In addition, the small number of CIA Neptunes could easily be hidden in the much larger group of Navy Neptunes deployed throughout the world.87
Conversion of the planes from Navy patrol aircraft to CIA covert collectors was conducted first under the designation Project CHERRY (and later WILD CHERRY). The result was seven planes equipped with advanced cameras for low-altitude photography and electronic intelligence (ELINT) gear. The planes were also given a new designation—the RB-69A.88
Initial operations with the modified Neptunes began in Europe in 1955 and continued into 1956. Wiesbaden, which became a U-2 base in 1956, received two of the planes, which were often flown by qualified Polish or Czechoslovakian defectors due to their knowledge of the local terrain and East European languages. In addition to flying peripheral missions, the planes were used to overfly parts of the western Soviet Union to collect data on power grids.89
In 1957, under a program designated ST/POLLY, RB-69A/Neptunes began flying out of Taiwan and near and into Chinese airspace, primarily to gather electronic intelligence on Chinese radars. The two Wiesbadenbased planes subsequently joined the five originally sent to Taiwan. A crew of twelve was standard for ELINT missions—pilot, copilot, flight engineer, radio operator, and eight ELINT system operators. The crews were handpicked by the Taiwanese Air Force and trained in the United States by the CIA. Missions were launched from Taiwan and, on occasion, South Korea. In addition to the safer peripheral missions were the overflights—which sometimes took the crews to Beijing or Canton, flying below 1,000 feet. The planes were also employed to drop espionage agents, leaflets, and supplies to agents.90
BEYOND THE U-2
Richard Bissell and Kelly Johnson had no illusions that the U-2 would be perpetually invulnerable to Soviet countermeasures. All they hoped for was a couple of good years. The Soviet ability to detect and track the plane from the beginning was not expected, but they began thinking about a successor plane long before the May 1, 1960, Powers incident.
In August 1957, the Scientific Engineering Institute (SEI), a Boston-based CIA proprietary that had been working on ways to reduce U-2 vulnerability, began to investigate the possibility of designing an aircraft with a small radar cross-section. SEI soon discovered that supersonic speed dramatically reduced the chance of detection by radar. As a result, the CIA focused on designing a successor to the U-2 that would fly extremely high and fast and would employ radar-absorbing or -deflecting techniques. Both Lockheed and the Convair Division of General Dynamics were informed of the conclusion to guide their research on a possible U-2 successor.91
To assist him in evaluating proposals, Bissell once again called on Edwin Land to serve as chairman of an advisory group. Other members included TCP veteran Edward Purcell, Allen F. Donovan of the Cornell Aeronautical Laboratory, and Bissell assistant Eugene P. Kiefer. The group often met in Land’s Cambridge office, at times with representatives of Lockheed, Convair, the Air Force, or the Navy attending.92 In September, the group rejected a Navy-Boeing proposal for a 190-foot-long, hydrogen-powered inflatable aircraft, as well as a Lockheed proposal for a hydrogen-powered aircraft, code-named SUNTAN. Two additional ideas from Kelly Johnson were also rejected—a tailless subsonic aircraft with a very low radar cross-section and a supersonic design designated A- 2. The group did approve Convair’s continuing to work on a ramjet-powered Mach 4 vehicle, code-named FISH, that would be launched from a specially configured Convair B-58B Hustler bomber.93
At a late November 1958 meeting, the panel reviewed the FISH program and Lockheed’s newest proposal, the A-3, and agreed that it was feasible to build an aircraft capable of flying fast enough and high enough to make radar detection exceedingly difficult.94 Eisenhower was already aware of the project via James Killian, who was serving as the first presidential science adviser. The President gave his approval for exploratory work after a December 17 briefing by Dulles and Bissell, with Land and Purcell in attendance. Lockheed and Convair were asked to submit detailed proposals, and they were provided funding for their research. The effort was designated Project GUSTO.95
At a July 1959 meeting, the Land panel rejected both the Convair and Lockheed proposals. Convair offered a FISH vehicle that would fly at Mach 4.2 at 90,000 feet, with a range of 3,900 miles. However, the technology of ramjet engines was unproven, and the B-58B had been canceled in June. Meanwhile, the susceptibility of Lockheed’s newest proposal, the A-11, to radar detection was considered too great.96
The following month, on August 20, the two contractors provided a joint Defense DepartmentCIAAir Force selection panel with the specifications for their proposed aircraft. One notable difference was in length, with Lockheed suggesting the 102-foot-long A-12, while Convair’s solution, the KINGFISH, was 79.5 feet. Projected speed—Mach 3.2—was identical for both planes. Lockheed’s proposal did promise better performance in terms of cruising altitude during the middle and final portions of the flight and range at cruising altitude (3,800 versus 3,400 nautical miles). The A-12 promised to reduce substantially the radar return through an additive in the fuel, which decreased the ability of a radar to detect the afterburner plume. A second novel feature of the design was the plan to use titanium rather than steel in parts of the aircraft in order to reduce its weight.97
Some CIA representatives at first favored Convair’s KINGFISH design due to its smaller radar cross-section, but they were eventually convinced to side with Lockheed by Air Force members, who had visions of B-58Blike delays and cost overruns. In addition, because of Lockheed’s work on the U-2, the company had in place employees with the proper clearances and the security arrangements that would be required for a project that Bissell and Dulles insisted be, if possible, more secret than the U-2.98
Because of lingering concern about the A-12’s radar cross-section, Lockheed emerged as only the provisional winner—with the requirement that it demonstrate the A-12’s reduced vulnerability to radar by January 1, 1960. Project GUSTO was terminated, and Project OXCART was born.99
Lockheed’s antiradar studies of that summer resulted in the eventual cobralike appearance of the OXCART aircraft. Edward Purcell and John Parangosky, the CIA’s program manager for OXCART, had theorized that a continuously curving airframe would be difficult to track with radar because it would present few corner reflectors or sharp angles from which radar pulses could bounce off in the direction of the radar. The studies also resulted in a contract, signed in February 1960, that called for Lockheed to receive $96.6 million in return for twelve aircraft.100
Of course, at the time the contract was signed, Gary Powers had yet to be shot down. Whether the project would survive, in a clima
te where its use over the Soviet Union might be considered too risky, remained to be seen. In late May, Eisenhower told his military aide, Andrew Goodpaster, that he believed the project should go forward on low priority, for Air Force use in time of war. In a memo, Goodpaster noted that Eisenhower “did not think the project should now be pushed at top priority. In fact, they might come to the conclusion that it would be best to get out of it if we could.”101
CORONA
In early August 1960, the U-2 was no longer flying over the Soviet Union, and the OXCART’s future was uncertain. Even if Kelly Johnson delivered the plane promised, it was far from clear that a president would permit it to overfly Soviet airspace. But another CIA-managed project was about to pay huge dividends and revolutionize U.S. intelligence capabilities.
Two and a half years earlier, on February 7, 1958, Killian and Land met with Eisenhower and Goodpaster at the White House to discuss the limited progress the Air Force was making in developing a photographic reconnaissance satellite.102 The primary objective of the program—first known as the Advanced Reconnaissance System (ARS), then as SENTRY, and finally as SAMOS—was to develop a satellite that would electronically scan the photographs obtained by its cameras and transmit the data back to a ground station, where it would be reconstructed into a picture. A subsidiary objective was to develop a satellite that would return its film back to earth in a canister.103
At that meeting, Eisenhower confirmed the decision to assign the CIA responsibility for developing a reconnaissance satellite that could eject its film for recovery on earth. That decision had been prompted by a late October 1957 report from the President’s Board of Consultants on Foreign Intelligence Activities (PBCFIA), which stressed the need for an interim photographic reconnaissance system that would be available before either SENTRY or OXCART. Bringing the CIA into the reconnaissance satellite effort had been proposed to General Bernard Schriever, head of the Air Force’s Ballistic Missile Division, by Colonel Frederic Oder, head of the SENTRY program. Oder believed that CIA funding would help to accelerate the program. He was also well aware of the CIAAir Force cooperation in regard to the U-2, inasmuch as Schriever’s deputy was Osmond Ritland, Bissell’s deputy on the U-2 project.104
Not surprisingly, given the success of the U-2, Bissell was assigned to manage the new satellite program, which would soon be designated CORONA, through his Development Projects Staff. Ritland repeated his role as deputy—for the Air Force would remain a key player in the program. Bissell exercised direct control of the program through monthly meetings with contractor representatives.105 His program staff included U-2 veterans such as Eugene Kiefer, his special assistant for technical analysis, and John Parangosky, who became deputy chief of the CORONA Program Office development staff. Out on the West Coast, Charlie Murphy, a longtime Air Force designee to the CIA, served as Bissell’s Field Technical Director at the Lockheed Advanced Projects (AP) facility in Palo Alto, where tests of the cameras and other elements of the payload were conducted.106
Five contractors played key roles in the development of CORONA. Itek, a Boston-based company founded by Richard Leghorn, and Fairchild Camera and Instruments were asked to develop camera systems. General Electric and Eastman-Kodak were, respectively, awarded contracts for developing the recovery capsule and supplying the film. The Lockheed corporation would have a dual role—in addition to building the upper stage that would propel the CORONA satellites into orbit, it was given the responsibility for integrating the entire effort.107
To provide cover for the launches—the portion of the effort that could not be hidden from public view—the DISCOVERER program was created. Ostensibly a scientific and biomedical research effort that would give some small animals a once-in-a-lifetime joyride into space, it would explain the repeated launches and recovery of payloads from orbit.108
As a result, a significant portion of the Air Force’s role, such as the procurement of boosters and launches, could be conducted as an overt effort. To carry out those responsibilities, a small program office, originally consisting of four or five people and headed by Lt. Col. Lee Battle, was established at the Ballistic Missile Division headquarters near Los Angeles. Work on the “black” side of the project was handled down the hall.109
Early on in the project, Bissell and Ritland made several key decisions concerning the camera and satellite. Originally, the camera to be carried on the first CORONAs, proposed by Fairchild, would remained fixed and scan the earth below as the satellite spun in orbit. In late March or early April 1958, they opted for an Itek design, in which the panoramic camera would scan while the satellite itself remained stable. Fairchild would remain as the camera builder. In addition, they decided that after the Agena second-stage had separated from the Thor rocket booster and carried the payload (consisting of the camera and recovery vehicle) to orbit, it would remain attached. The Agena would ensure that the payload remained stable in orbit, as well as provide power.110
On February 28, 1959, a little over a year after the CORONA program started, test launches began from Vandenberg Air Force Base in California. That first launch did not carry a camera, which was just as well, since the Agena rocket failed to reach orbit. Two more camera-less launches, in April and June, also produced failures. In the first instance, the reentry vehicle, instead of landing in the Pacific where it was to be recovered either in the air before final touchdown or from the ocean, landed in the vicinity of Spitzbergen, Norway, near a Soviet mining operation, and was never recovered—at least not by the United States.111
The fourth launch, on June 25, was the first to carry a camera, which was subsequently designated KEYHOLE-1 (KH-1). The Itek-designed panoramic camera, manufactured by Fairchild, was capable of scanning 35 degrees in each direction from the line of flight. At the center of its swing, it would point straight down toward the earth.112
But that mission proved no more successful than the previous three. The next eight tries, seven of which involved camera-carrying satellites, also resulted in failure of one variety or another. The reentry vehicle would sink or be shot into a higher orbit, or it would fail to detach itself from the spacecraft, or its parachute would not open; other times, the rocket would be destroyed or head for the Pacific Ocean rather than outer space. Failure was constant, even if the cause was not. It was, Bissell recalled, “a most heartbreaking business.”113
Finally, on August 10, 1960, Navy frogmen recovered the reentry vehicle from DISCOVERER 13 after it had eluded the C-119 aircraft that was supposed to snatch it out of the air. There was no camera on board, but the recovery represented a major step forward for the beleaguered program—a significant enough triumph that James Plummer, Lockheed’s CORONA program manager, was thrown into a swimming pool during the resulting celebration.114 Others were not as elated. The cable from Bissell assistant Eugene Kiefer to Lee Battle read “Congratulations on a random success.” Battle was not amused.115
But just eight days later, full success arrived. A KH-1-equipped CORONA blasted off and, after some difficulty, attained a 116-by-502mile orbit, with an inclination of 80 degrees, allowing it to overfly all of the Soviet Union. As it began to pass over Soviet territory, its camera started operating as programmed. Its first photos were of the MysSchmidta air base in the Soviet Far East, about 400 miles from Nome, Alaska. On August 19, after seventeen orbits of the earth, which took it over the Soviet Union seven times, the satellite’s film-return capsule was ejected and then plucked out of the air off Hawaii as it descended toward the ocean.116
PHOTOINTERPRETATION
A few days after the capsule had been recovered and the film processed, Arthur Lundahl, director of the CIA’s Photographic Intelligence Center, addressed an auditorium full of his interpreters—the individuals whose skill and special training would begin the process of turning overhead images into intelligence. In the past, after film from U-2 missions was delivered to Washington, Lundahl’s photointerpreters were shown a map of the Soviet Union with a squiggly line indicating
the route of the U-2. In addition, they viewed blowup photos of particular targets.117
The CORONA briefing had a dramatic flair. After the photointerpreters were settled, Lundahl announced that it was “something new and great we’ve got here.” His deputy, Jack Gardner, opened a curtain to show a map of the Soviet Union. Instead of a single line across the map, there were seven vertical stripes emanating from the poles and moving diagonally across the Soviet Union. The interpreters knew the stripes represented the portions of the Soviet Union that had passed under the satellite’s cameras. Their immediate reaction was to cheer. After being briefed on what to look for, especially for missile sites at Plesetsk, they began work on OAK-8001, the first photointerpretation report based on satellite photography.118
They had an extensive amount of film to study. The first CORONA mission produced coverage of over 1 million square miles—greater than that produced by all of the U-2 overflights over the Soviet Union. There were 1,432 photos, including photos of the Kapustin Yar Missile Test Range, its presumed impact area, SA-2 missile sites, the Sarova Nuclear Weapons Research and Development Center, and several newly discovered airfields.119 The quality, however, was substantially lower than that achieved by the U-2. The KH-1 camera produced images with a resolution of about 40 feet—sufficient to permit identification of the missile sites, airfields, and facilities, but of little use in producing detailed technical intelligence.120
Still, the crude photographs returned by the August 18 mission, along with photos from four additional successful CORONA missions between December 1960 and June 1961, all of which employed the improved KH-2 camera with its 25-foot resolution, enabled Lundahl’s photointerpreters to shatter a significant myth—one that had played a major role in the 1960 presidential campaign.121
The Wizards of Langley Page 4