Analysts also examined telemetry to create a record of the liquid level at the bottom of the propellant tank as burnout neared. From that information, they sometimes could determine the shape of the bottom of the tank, a determination that could then be used in estimating missile size—which as the SS-8 debate indicated could be a key issue in assessing the missile’s capabilities. Telemetry data could be combined with other data to determine the velocity, acceleration, flight path, and angle of a reentry vehicle’s descent—significant data in establishing its vulnerability to missile defense systems.89
Turning telemetry and other technical data into estimates of the characteristics of Soviet missiles allowed the national estimates on Soviet strategic forces, such as the October 1964 NIE, to assess the probable range, accuracy, reentry vehicle weight, warhead weight, warhead yield, and type of propellant for each operational Soviet missile.90 Such data were important not only in assessing the Soviet threat and U.S. strategic requirements but also, a few years down the road, in developing arms control strategies.
EAVESDROPPERS
A significant part of the data that FMSAC and other elements of the intelligence community relied upon in attempting to decipher the Soviet missile and space programs was provided by the directorate’s own collection activities—particularly those of the Office of ELINT (OEL).
The CIA-funded Norwegian station at Kirkenes and its subsidiary METRO outpost at Korpfjell continued to intercept communications, telemetry, and other electronic signals. To enhance intercept capability, the CIA budgeted part of $104,000 to replace one of the principal ELINT receivers at the Kirkenes site during the 1966–1967 fiscal year.91
The remainder of the money went to “activate an ELINT boat operation in the Barents Sea,” which was targeted against Soviet naval operations. In 1965, the Globe XIV, a whale catcher, was purchased and converted into the ELINT ship Marjata I. The Marjata I replaced the Eger in 1966. Soon after the Marjata’s first operations, it became the subject of intense Soviet interest, and some “incidents” followed, but the Norwegians did not consider them sufficiently serious to halt the operations.92
According to John McMahon, the Norwegians were “not intimidated,” and the boat operation produced “great intelligence.” Included were data on launches out of the White Sea, on air-to-air and air-to-ground missile launches, and on Soviet practice firings from the Barents Sea. The operation also provided “good COMINT coverage.”93
OEL also sought to improve its ability to monitor missile tests emanating from Tyuratam and antimissile activity at Sary Shagan. In 1965 and 1966, OEL established a second telemetry intercept station in northeastern Iran at Kabkan, forty miles east of Meshed. Code-named TACKSMAN II, the station was only 650 miles southwest of Tyuratam.94 As with the TACKSMAN I facility at Beshahr, it was a strictly U.S. operation, with no Iranians permitted inside the facilities. It also had, as did the Beshahr site, a communications intercept capability to permit monitoring of test range communications.95
TACKSMAN II was located in a remote mountainous area inhabited by nomads, and although the station became home to advanced electronic equipment, living conditions were primitive for those on the site survey team and the initial permanent contingent.96 Bob Phillips was among the seven people who established the site in 1965, and he returned in 1966 to spend a year as chief engineer. The nine or ten individuals who spent that year at TACKSMAN II had to dig a slit trench to serve as the latrine, carry water up the mountain, and have their supplies flown in from Tehran. It was “like camping out for a year,” Phillips recalled, except camping out usually does not involve “sitting on a slit trench [in freezing weather] in the middle of the night.” The site was devoid of trees, a factor Phillips believed influenced his later decision to buy a house in an area of northern Virginia that had “trees everywhere.”97
But the hardships endured by the CIA’s personnel on an isolated mountain in Iran paid huge dividends for the FMSAC analysts who were trying to crack the Soviet missile and antimissile programs. At their peak, the Iranian stations provided about 85 percent of the hard intelligence on the Soviet ICBM program. The sites could do what no other U.S. intercept sites could do—monitor the last moments of the firing of the missile’s first stage, which meant a greater degree of confidence in determining missile dimensions and throw weight. The material, according to Phillips, came in “pure” and required no exotic processing. To FMSAC chief Duckett, it was “pure gold.”98
The Norwegian and Iranian stations (along with other stations operated by NSA or its military components) had an assortment of operational and test firings to monitor between 1964 and mid-1966. In 1965, the Soviets began test firing solid-fueled missiles, which the United States designated the SS-13 Savage, from Plesetsk out to the Kamchatka Peninsula, over 3,000 miles away. That same year they also commenced test firings of a solid-propellant missile from Kapustin Yar on journeys of about 1,100 miles. Test firings of the SS-10 from Tyuratam began with a failure in April 1964, but six successful tests followed by the end of September.99
Between mid-March and mid-April 1966, four different types of Soviet missiles (SS-7, SS-8, SS-9, SS-11) were test fired from Tyuratam to the Klyuchi impact zone on Kamchatka. One objective of the SS-9 firings may have been to test a major modification of the SS-9 reentry vehicle. Similarly, SS-11 tests appear to have been related to testing of a heavier version of the missile’s reentry vehicle.100
In 1965 and 1966, while Kirkenes, Beshahr, and Kabkan were listening to Soviet missile tests, another CIA facility was listening for signals from the moon. Out at Stanford University in Palo Alto, California, the CIA was employing a 150-foot dish antenna to monitor the signals of Soviet radars after they had bounced off the earth’s only natural satellite.101
The “moonbounce” phenomenon had been discovered in 1946, when scientists detected a man-made signal reflected from the moon. Experiments that followed revealed the extraordinary weakness of such signals. A typical signal received via moonbounce was a billion times weaker than if it were intercepted by an airplane ten miles from the transmitter. As a result, only very large antennae could effectively hear such signals and distinguish them from other signals.102
By the early 1960s, the possibility of exploiting the moonbounce phenomenon was being investigated by a number of agencies. N. C. Gerson of the National Security Agency used the Arecibo Ionospheric Observatory in Puerto Rico to intercept moonbounce signals from a Soviet radar operating on the Arctic coast. Along with a member of the Army Security Agency, he produced a three-volume study—Moonbounce Potential from Scooped Antennas. The Air Force also had a moonbounce project, FLOWER GARDEN, which relied on several antennae, including the 250-foot antenna at Jodrell Bank. Other moonbounce collectors were the antennae at the Grand Bahama tracking station, a Navy intercept site at Sugar Grove, West Virginia, and the Naval Research Laboratory’s Chesapeake Bay Annex.103
Among the intelligence issues the intercepts helped resolve was the Hen House radar. In January 1964, the NRL’s 150-foot antenna at Chesapeake Bay, programmed with information from other sources about the Hen House’s frequency, made the first intercept of the radar’s BUEB signal after it ricocheted off the moon—although it was not clear at the time whether the intercept was of a Hen House signal or a Hen Roost signal. Analysts were able to determine that the signal was from a phased-array radar and made considerable progress in determining the radar’s signal characteristics. They also determined that the signal came from Sary Shagan.104
By 1964, the CIA had received reports that Hen House radars were being deployed at several locations in the Soviet Union, but there was no corresponding evidence of Hen Roost deployment. If the intercepted signal did come from Hen House, the United States had acquired significant intelligence about the radar well before its full deployment. If the signal was from Hen Roost, then the CIA and other members of the intelligence community would have no clue about how Hen House worked. In early 1965, the issue was settled by a CIA con
tractor, ESL, Inc., founded by future Secretary of Defense Bill Perry. ESL proved with mathematical rigor that the BUEB signal came from Hen House—a radar that would be employed for ABM, early warning, and space tracking purposes.105
The CIA’s Palo Alto facility, which had been chosen because of its potential with regard to westward-looking Soviet radars, enhanced understanding of the radar system. The facility consisted of “quite sophisticated collection equipment, including two unique receivers,” which were built particularly for the moonbounce mission. In August 1965, Palo Alto made its first intercept of a Hen House radar signal, a signal it was able to observe for thirty-eight hours a month.106
The data collected at Palo Alto, added to those obtained by the Defense Department’s antennae, led to three major conclusions: First, the Hen House signal had a “spread-spectrum” mode—the frequency spread of the signal could be deliberately broadened to increase the radar’s range or its accuracy in reading the target’s speed. Second, Hen House relied on an advanced scanning system that enabled the radar not only to search for a target but also to dwell on it for a short time. The brief look allowed the identification and measurement of the radar’s signal parameters. Finally, the moonbounce data led to estimates that the peak power of the Hen House transmitter was twenty-five megawatts, making it one of the highest-powered radars in the world. These findings led to a fourth conclusion, which proved correct—the Hen House was a new, sophisticated ABM radar. On the basis of that conclusion, the United States could begin developing countermeasures and tactics to reduce its effectiveness.107
The Office of ELINT was also pursuing its Quality ELINT program in support of current U-2 and future OXCART operations—flying its Power and Pattern Measurement System (PPMS) on various Air Force planes in order to determine the power and coverage of the Soviet radars designed to detect penetrating aircraft.108
Choosing which radars to target involved balancing intelligence priorities, air access, and eavesdropping conditions. Preference was given to targets in isolated areas where signals from other radars would not interfere. Radar signals were identified in advance, using direction finding equipment that was part of the airborne system. Special navigational instruments recorded the aircraft’s position and altitude during each collection operation so that analysts would know the exact geometric relationships between the radar and the measurement system. Several projects were completed in six missions or less, whereas others required more than forty flights.109
Seventeen Quality ELINT missions were flown between August 1963 and September 1966 (see Table 3.1), with the PPMS being carried on Air Force RB-47H, C-97, and C-135 aircraft. On the September–October 1963 New Breed III mission, an RB-47H flew over the Arctic north of the Soviet Union to monitor Tall King and Spoon Rest radars. In October, an RB-47H mission, designated Iron Lung, monitored signals from a Spoon Rest radar in Cuba. Between May and October 1966, a C-135, as part of Operation Briar Patch, flew over the Barents Sea intercepting signals from a Hen House radar.* During that same period, a C-97 flying over the Gulf of Tonkin intercepted the emanations from a Fan Song radar during Operation See Top.110
TABLE 3.1 Quality ELINT Missions, August 1963-October 1966
CHECKROTE
In addition to trying to unravel the secrets of Soviet radars, the Office of Research and Development (ORD) continued to employ over-the-horizon (OTH) radars to help decipher the mysteries of the Soviet and Chinese missile programs. By September 1965, the EARTHLING radar in Pakistan had detected sixty-five missiles launched from Tyuratam; these accounted for 82 percent of the missiles known to have been launched from that site when EARTHLING was operational. A few of the detections had not been noticed by any other collection system, possibly the result of some combination of aborts not picked up by any line-of-sight collection system or false alarms.111
In May 1965, as U.S.-Pakistani relations deteriorated and the United States faced loss of all its intelligence facilities in Pakistan, ORD began to install an OTH radar system called CHECKROTE on Taiwan—but only after Wheelon had exerted considerable effort to move a graveyard inconveniently located on the intended site. The radar’s primary function was to monitor missile launches from China’s Shuangchengzi missile complex. By August 1, 1966, CHECKROTE was up and running. Almost a year earlier, in September 1965, Pakistan had closed the EARTHLING installation.112
EMPLACED SENSORS
In 1965 and 1966, the DS&T made at least two attempts to collect intelligence on the Chinese nuclear and missile programs using emplaced sensors—sensors placed at a strategic location and, DS&T hoped, undiscovered by the target.
In 1965, the Indian government gave the CIA permission to plant a device on the summit of Nanda Devi in the Himalayas to monitor telemetry from the Chinese missile center at Shuangchengzi. The device, which was developed by the Office of ELINT in response to FMSAC’s requirement, would unfortunately be swept away by an avalanche.113
In 1965 or 1966, a U-2 carried a spearlike device on an eleven-hour trip into China and ejected it. At 3,000 feet, its parachute was to open and carry the device to earth, where it was to stick in the ground. The mission was appropriately named Project JAVELIN, and the device was code-named TOBASCO. ORD and Sandia Labs had developed it for OSI. From its intended location near, but not too near, the Chinese nuclear test site at Lop Nur, the device, equipped with airwave and ground motion sensors, was to provide the CIA with data on the explosive power of Chinese nuclear tests.114
TOBASCO had been tested in Nevada and New Mexico during U.S. nuclear tests. But it was never heard from again after being ejected from the U-2. Whether it hit a rock, was discovered by nomads, failed to implant itself because the terrain was harder than expected, or just didn’t work, was never discovered. Perhaps it was an omen that upon his return the pilot who handled the aerial javelin toss was forced to crash-land short of his base and came down in a rice paddy.115
AERIAL PROJECTS
In addition to javelin tossing, U-2s had continued to fly over China, photographing nuclear and missile facilities. In 1962, Sino-Indian relations deteriorated into war, giving the United States and India a common enemy. The CIA had already provided India with U-2 photographs of the Chinese border.116 In 1963, the CIA suggested establishment of a temporary U-2 detachment in northern India. From there, the nuclear test site at Lop Nur and other targets in Xinjiang province could be reached on U-2 missions.117
The White House approved the idea, which was followed by protracted negotiations with the Indian government. In spring 1964, India agreed to deployment of a U-2 detachment at Charbatia, an old wartime base near Cuttack on the east coast. Two or three missions followed, with the Indian government receiving up-to-the minute intelligence on Chinese military deployments along the border, while the CIA obtained pictures of Lop Nur and other Xinjiang targets.118
In 1964, as the United States pondered the status of the Chinese nuclear program, Texas Instruments Corporation developed an infrared scanner, the FDD-4, that could be employed to determine whether nuclear facilities such as Lanzhou were operational. The expectation was that the heat generated by active facilities would show up in infrared photographs.119
In 1964 and 1965, the Nationalist Chinese Black Cat squadron made several attempts to fly a U-2 equipped with the FDD-4 scanner over the nuclear facilities at Baotou and Lanzhou to determine if they were active. Previous tests of the scanner on missions over U.S. nuclear facilities indicated that it produced its best results at night.120
The first target was the uranium gaseous diffusion plant at Lanzhou. However, the November 1964 mission was aborted due to an electrical failure. The second try, this time targeted on Baotou, was also aborted when a test of the plane’s defensive systems failed. The third time appeared to be the charm, at least for a while. The plane got within thirty miles of Lanzhou when its SAM warning device started blinking, indicating a SAM site almost directly ahead and forcing an abort.121
Another mission, this time flying ou
t of Korea, also had to be aborted. Finally, on January 8, 1965, a Black Cat pilot took off from Taoyuan on a flight that lasted a little over seven hours, almost six hours of which was spent over the mainland. The pilot made it to Lanzhou without equipment failures or SAM attacks. The infrared camera demonstrated that the facility was operational. Unfortunately, two days later a mission over Baotou resulted in disaster when the plane was shot down.122
Indeed, the pilots who participated in Project TACKLE, as the U-2 missions over China were code-named, were in constant danger from China’s air defense system. Nationalist Chinese pilots flying from Taiwan as well as U.S. pilots flying from a base at Ban Takhli, Thailand, all faced the prospect of death or incarceration.
During a July 1966 flight, Major “Spike” Chuang Jen Liang ran into a SAM battery when turning over Kunming air base; eight missiles were fired at him in two salvos. Two were photographed as he put his plane into an evasive turn and escaped death. On another overflight targeted on Lanzhou, Major Billy Chang Hsieh spotted two SAMs fly by as he passed over the nuclear facility. As he revisited the site shortly afterward, the camera photographed missile crews scrambling to reload the launchers and fire another salvo.123
A device instrumental in saving Chuang’s life was the Oscar Sierra warning system, which had been installed in the U-2s in spring 1965. The key to the system was its detection of the change in pulse repetition frequency of the SA-2 radar when the system shifted into firing mode. The change caused a red light labeled “OS” to go on in the U-2 cockpit, giving a pilot flying at 72,000 feet about forty seconds to get out of the way. Pilots soon came to translate the “OS” as Oh Shit!124
The Wizards of Langley Page 13