The Wizards of Langley

by Jeffrey T Richelson

  MELODY also provided new intelligence on the dispersion of Soviet air defense radars, although it did not help OXCART planners to identify all such radars. Determining the range and power of the radars was the objective of the aerial ELINT operations of the Quality ELINT program, but the program also had the more secret element of determining their sensitivity. That element was known by the code word PALLADIUM.

  OXCART planners wanted to determine the sensitivity of Soviet radar receivers and the proficiency of their operators. With the help of scientific consultants, Poteat came up with a scheme to electronically generate precisely calibrated false targets and insert them into Soviet radars—deceiving the Soviet radar operators into seeing and tracking ghost aircraft.39

  The Soviets would expect a radar signal to return after bouncing off an aircraft. The first step in the Quality ELINT deception was to capture the radar’s signal and feed it into a “variable delay line” before sending the signal back to the radar. By smoothly varying the length of the delay line, CIA technicians could simulate the notional aircraft’s range and speed. The knowledge gained from the Quality ELINT power and coverage measurements enabled program staffers to simulate an aircraft of any radar cross-section—from an invisible stealth airplane to one that would create a large blip on a Soviet radar—and fly it along any path, at any speed and altitude.40

  The smallest blip would provide a measure of the sensitivity of the radar and the skill of the operators. The key was to find a way of determining which blips the Soviets could see on their radar screens, and which they could not. Poteat and his colleagues considered either monitoring other Soviet radars when a ghost airplane was projected to see if they were switched on in an effort to pick up the imaginary intruder detected by the target radar, or intercepting the pertinent Soviet communications link. With the assistance of NSA, it was possible to intercept and decrypt the relevant communications.41

  Every PALLADIUM operation involved three teams—a CIA team with equipment to generate ghost aircraft, an NSA team with special COMINT and decryption gear, and a military operational support team. Covert PALLADIUM operations were directed against a variety of Soviet radars around the world using ships and submarines as well as ground bases. The logistics of some operations were particularly difficult. One winter, heavy snows closed all the airports in northern Japan. Because of the small rail tunnels, one CIA officer “spent about three weeks in northern Japan, in the dead of winter, hauling his van of PALLADIUM equipment off trains and trucking and sledding it over the mountains—and putting it back on another train on the other side.”42

  The October 1962 Cuban missile crisis provided an opportunity to measure the sensitivity of the SA-2 radar. During the crisis, a PALLADIUM system was mounted on a destroyer operating out of Key West. The destroyer stayed just out of detection range of a Soviet early warning radar near Havana, and its PALLADIUM transmitting antenna was just above the horizon. The signals from the antenna made it appear that a U.S. fighter plane out of Key West was about to overfly Cuba. Meanwhile, a Navy submarine slipped in near Havana Bay, with instructions to surface just long enough to release a timed series of balloon-borne metalized spheres of different sizes.43

  Poteat and his colleagues expected the radar to track the false aircraft, while the submarine released the spheres into its “path.” They hoped that after the initial detection, the Soviets would turn on the SA-2 target tracking radar in preparation for firing missiles at the aircraft, and then report seeing other targets—the spheres. The size of the smallest spheres detected would establish the smallest radar cross-section that could be detected and tracked.44

  As Poteat recalled, “We got the answers we went after, but not without excitement and entertainment.” In the middle of the operation, Cuban fighter planes began circling over the area where the submarine had surfaced, while another fighter started chasing the ghost aircraft. The PALLADIUM system operated smoothly to keep the ghost aircraft always just out of reach of the Cuban planes. When the Cuban pilot told his controllers that he had the intruding aircraft “in sight” and was about to make a pass and shoot it down, Poteat and the others on the scene simultaneously had the same idea. The technician moved his finger to the switch, Poteat nodded yes, and the technician switched off the PALLADIUM system—causing the ghost aircraft to disappear from the Cuban pilot’s radar screen.45

  The data produced by the PALLADIUM operations, which continued for several years afterward, enabled analysts at OSI and elsewhere to identify which Soviet radars had low power or maintenance problems, which performed below expectations, and where U.S. aircraft might safely penetrate the Soviet border in wartime. OSI analysts also concluded, undoubtedly to the disappointment of the OXCART program office, that Soviet radars could detect and track an OXCART as soon as it came over the horizon.46

  SPECIAL ACTIVITIES

  Of the three groups forming the Directorate of Research, the Office of Special Activities could, during 1962 and 1963, lay claim to conducting the most extensive and important operations. From February 1962 through June 1963, U-2 and CORONA missions yielded significant intelligence on the Soviet Union, China, and a number of other targets.

  Nineteen CORONA missions were flown during that time period, all but three of which successfully returned film. There were an additional three successful ARGON mapping missions.47 Meanwhile, CIA U-2s closely monitored the growing Soviet involvement in Cuba, and U-2s flown by Nationalist Chinese pilots photographed Chinese nuclear facilities. CORONA missions also focused on Chinese nuclear and missile facilities as well as the primary Soviet target.

  Cuba, a difficult target for satellites, was frequently overflown by U-2s, as the CIA attempted to determine the nature of the relationship between Fidel Castro’s revolutionary regime and the Soviet Union. A particular fear was that the Soviets might be tempted to use an island ninety miles from the United States as a base for offensive weapons. In spring 1962, after receiving reports of increased Soviet activity in Cuba, the CIA requested permission from the NSC Special Group to conduct additional missions. The group authorized an increase of at least two additional flights a month, beginning in May, and NPIC began publishing a Photographic Evaluation of Information on Cuba series.48

  By early August, CIA analysts noted a substantial increase in Soviet arms deliveries to the Castro regime. An early August overflight was conducted too soon to detect the Soviet construction activities then getting under way, but an August 29 mission enabled the CIA to inform President Kennedy of at least eight SA-2 sites in the western half of Cuba. A mission on September 5 provided additional evidence, showing three more SAM (surface-to-air missile) sites as well as a MiG-21 at Santa Clara airfield.49

  The discovery strengthened John McCone’s belief, which was not shared by CIA analysts, including Ray Cline, that the Soviets were deploying Medium-Range Ballistic Missiles (MRBMs) in Cuba. The DCI argued that the only explanation for the extensive deployment of SAM sites was a Soviet desire to hide the missiles from U.S. reconnaissance.50 Bad weather prevented any missions until September 17, but heavy cloud cover that day prevented the mission from obtaining usable images. Finally, on September 29, an overflight of the Isle of Pines and Bay of Pigs area turned up another SAM site.51

  In gaining approval for the overflights, the CIA had to face reluctant senior Kennedy administration officials, including Secretary of State Dean Rusk, who feared a shootdown and an international incident. But over-flights continued to amass evidence of Soviet SAM deployments, with missions on October 5 and 7 bringing the total of sites discovered to nineteen, although there was still no evidence that offensive missiles had been deployed.52

  However, the most likely location for the MRBMs was western Cuba, which missions had largely avoided because of the concentration of SAM sites on that part of the island. On October 9, the Special Group met to discuss a recommendation by the Committee on Overhead Reconnaissance (COMOR), an interagency committee responsible for selecting U-2 and CORONA targets,
for a mission over the western part of the island. The committee also recommended that if the overflight did not provoke an SA-2 response, it should be followed by “maximum coverage of the western end of the island by multiple U-2s simultaneously.”53

  McCone brought along Col. Jack C. Ledford, who had become OSA director on September 4. A few months earlier, Ledford had expected that by September he would be assigned to a SAC air base in the Midwest. He had just graduated from the Industrial College of the Armed Forces, when a friend in Air Force personnel called to say he had a good assignment for Ledford in Washington, but the details were highly classified. Ledford, after four years in the area, was looking forward to a change of scenery. But when his friend told him that chief of staff General Curtis LeMay wanted him to take the job, Ledford, who had flown under LeMay in World War II, agreed.54

  Ledford suspected that Scoville and Giller, both of whom he knew, were instrumental in his selection. At a meeting with the two, they gave him some of the details—additional information would follow if he passed his polygraph exam. When that process was completed, he was officially assigned to an Air Force organization at Bolling Air Force Base in Washington—the 1040th USAF Field Activity Squadron, which served as an administrative office for Air Force personnel assigned to the CIA, NSA, and other intelligence agencies.55

  At the October 9 meeting, Ledford presented a vulnerability analysis, produced with the help of OSI, concluding that there was a one-in-six chance of losing a U-2. As a result, the Special Group approved a flight over San Cristobal.56 However, notwithstanding the CIA’s role in providing initial evidence of the SAM buildup, the newly approved mission was to be flown instead under SAC auspices. As the Special Group meeting ended, Deputy Secretary of Defense Roswell Gilpatric questioned the plausibility of the CIA’s cover story for the missions, which involved alleged flights to Puerto Rico, and argued that it would be better to employ Air Force pilots. In the event of a mishap, officials could claim that a routine Air Force peripheral reconnaissance mission had gone off course. McCone asked Ledford, who agreed that the Air Force had a better cover story but noted that the CIA’s U-2s had superior electronic countermeasures and could fly 5,000 feet higher than SAC’s; Ledford suggested that the SAC pilots use the CIA’s planes. President Kennedy, after meeting with McCone and Gilpatric, approved the mission and the use of Air Force pilots.57

  On October 11, Department of Defense (DOD), Air Force, and CIA representatives, including Scoville, met to discuss the change in cover stories. The discussion then focused on who would run the next mission. The DOD representatives, who were strongly in favor of Air Force control of the missions, called McCone, who consented. The next day, Mc- Cone left for California and did not return until October 14. On October 12, Air Force control became official when President Kennedy transferred “responsibility, to include command and control and operational decisions, with regard to U-2 reconnaissance flights of Cuba” from CIA to Defense. The Air Force then requested the loan of two CIA U-2Cs.58 The mission flown on October 14 resulted in eight cans of films being delivered to NPIC the next day. The film provided the first hard evidence of the presence of MRBMs and set in motion the Cuban missile crisis, which led to the withdrawal of Soviet missiles from the island.59

  The transfer of the mission to the Air Force annoyed several CIA officials, both in the short and long term. Lt. Gen. Marshall Carter, the acting DCI, told national security adviser McGeorge Bundy, “I think it’s a hell of a way to run a railroad. It’s perfectly obviously a geared operation to get SAC in the act.” In conversations with senior Air Force and administration officials, Carter argued against the move, pointing out that OSA’s operation was up and running and working well. In contrast, the Air Force pilots had no experience with the U-2C. In addition, the Air Force had managed peripheral but not overflight operations. He told Gilpatric, “To put in a brand new green pilot just because he happens to have on a blue suit and to completely disrupt the command and control and communications and ground support system on 72 hours’ notice to me doesn’t make a God damn bit of sense, Mr. Secretary.” But Carter’s protestations were to no avail. Even a plea for a gradual transition to Air Force control of the Cuban overflights was rejected—with presidential national security adviser McGeorge Bundy telling Carter that “the whole thing looks to me like two quarreling children.”60

  In addition to monitoring Cuba, U-2s flown by National Chinese pilots, along with CORONA satellites, kept watch on China’s emerging nuclear and missile programs. A December 1961 CORONA mission returned the first photos of the Lop Nur nuclear test site. Overhead reconnaissance during the first half of 1963 provided images of the Lanzhou Gaseous Diffusion Plant, a nearby hydroelectric plant, and a nuclear facility at Baotao. The Lanzhou images led analysts to conclude that although the “nearby hydroelectric plant . . . has made some progress, . . . much work remains to be done.” Late 1962 CORONA missions provided NPIC’s analysts with photographs of a possible uranium mining activity at A-Ko-Su. Earlier in the year, in June, overhead photography of the Shuangchengzi missile test complex showed a large crater about 1,500 yards from one of the launch pads, indicating “one rather spectacular failure.”61

  An even higher intelligence priority than China was, of course, the Soviet Union. Overflights using U-2s were no longer possible, and OXCART was not, in 1962 and 1963, even an option. Until the launch of the Air Force’s high-resolution KH-7/GAMBIT satellite in July 1963, CORONA was the sole overhead means of monitoring the testing and deployment of ballistic missiles, the construction of submarines and other naval vessels, the types and locations of air defense missiles and radars, and other key targets.

  At least ten CORONA missions between June 1962 and March 1963 provided photographs of the Kapustin Yar and Tyuratam missile test centers, which revealed the major expansion of Tyuratam. Suspected chemical warfare production plants, such as the one at Dzerzhinsk, and facilities probably involved in the testing of solid propellants and related explosives, like the one at Krasnoyarsk, were also frequent targets of CORONA cameras between mid-1962 and mid-1963.62

  On April 30, 1962, Herbert Scoville could not be found at his desk. He was with several CIA officials in Nevada, at the top-secret Area 51, to witness the first official OXCART flight. Richard Bissell was also there, leading Kelly Johnson to note in his project log “I was very happy to have Dick see this flight, with all that he has contributed to the program.”63 Piloted by Louis Schalk, the unusual-looking plane flew at an altitude of 30,000 feet and a speed of 340 knots—only fractions of what the planes would prove capable of, but it was a start. A May 4 flight that reached Mach 1.1 brought Kelly Johnson a congratulatory telegram from Mc- Cone.64

  But McCone would not remain pleased. Near the end of 1962, flight tests of two A-12s demonstrated the capability to reach 60,000 feet and a speed of Mach 2.16. However, progress was still slow, due to delays in the delivery of engines and the disappointing performance of those that had been delivered. It had become evident that Pratt & Whitney was finding it far more difficult to bring United Aircraft’s J-58 engines up to OXCART specifications.65

  In early December, McCone complained to the president of United Aircraft after learning that there would be another delay in the delivery of engines. As a consequence, “by the end of the year we will have barely enough J-58 engines to support the flight test program adequately.” By the end of January 1963, ten engines were available. The first flight with two of them installed occurred January 15.66

  A few months later, in late May, the project almost was exposed to the public. On a routine training flight, pilot Kenneth Collins’ plane stalled and went out of control, due to a faulty air speed indicator, forcing him to eject from the aircraft, which crashed fourteen miles south of Wendover, Utah. Collins was unhurt and the wreckage was recovered in two days, while persons at the crash site were identified and requested to sign secrecy agreements. Two farmers, who arrived near the crash scene in a pickup, were told the plane had be
en carrying atomic bombs, which discouraged them from getting any closer. The press was given a cover story describing the crashed plane as an unclassified F-105 fighter. All A-12s were grounded for a week, until the cause of the faulty instrument, which was easily correctable, was discovered.67

  While one component of OSA was wrestling with OXCART’s problems, another was concerned with improving space reconnaissance capabilities. The most important aspect of that effort centered around a modification to CORONA, referred to as the CORONA-J or KH-4A. The camera system for CORONA-J, the J-1, was essentially identical to the KH-4 or MURAL camera, with only slightly improved resolution—usually between 9 and 25 feet, although it reached 7 feet at times. The major change involved doubling the size of the film-supply cassette, a change made possible and necessary by the addition of a second reentry vehicle to the spacecraft. KH-4A spacecraft, which began operations in August 1963, eventually carried up to 160 pounds of film, in contrast to the 39 pounds carried in the KH-3 missions. Mission lifetimes grew from a maximum of 7 days for KH-4 missions to up to 15 days for KH-4A missions. The combination of extra film and two recovery capsules increased the frequency with which targets could be photographed during a mission, and increased the probability of finding targets free of cloud cover. A KH-4A mission could produce 18,000,000 square miles of stereo coverage.68

  During 1962, OSA was also involved in a joint program with the Air Force, designated LANYARD, to develop a satellite intended to provide high-resolution photography of a target located at Tallinn in Estonia. In 1961, KH-4 CORONA photos showed what some analysts feared were antiballistic missiles. The photos were of poor quality, showing roads ending in a circular clearing, like “lollipops in the snow,” according to one pho-tointerpreter. NPIC’s interpreters concluded that the photos showed construction for the deployment of the SA-5 GAMMON interceptor missile, with three batteries of six launchers arranged around a single engagement radar. The exact purpose of the Tallinn line, as it was called, was to be hotly debated within the intelligence community for years to come.69