The Wizards of Langley (15 page)

Meanwhile, Soviet lunar probes were launched to orbit the moon or achieve a soft landing and in each case also to send back photographs. A string of failures in 1964 and early 1965 was followed by the success of
Zond 3,
which sent back pictures of the lunar surface taken during a flyby. That was followed by several more failed soft landings before
Luna 9
was successful on January 31, 1966.
⁷³

Probes fired at Mars or Venus included the April 1964 mission of
Zond
1,
which the Soviets apparently lost contact with before it passed by Venus. The
Venera 2
and
Venera 3
missions in November 1965 were par
tially successful but did not transmit any data because the first flew by and the second crashed into the planet.
⁷⁴

One component of FMSAC, the Activities Interpretation Division, served as twenty-four-hour watch center. The time difference between Washington and Soviet launch sites, along with the urgency in determining the mission associated with each launch, required FMSAC to disrupt the sleep of key personnel. Evan Hineman remembers being called in during the middle of the night on several occasions to examine data on trajectories and orbits so that the White House, NASA, and other agencies could be told whether an earth satellite, interplanetary probe, or lunar mission had been launched.
⁷⁵

Among the important elements of these inquiries was trajectory analysis. Establishing the launch time to the nearest minute, through tracing the trajectory back to the launch point and incorporating data on velocity, permitted a great deal to be inferred about the objectives of the mission as well as the technology employed. Thus, the launch time of
Luna 1
on January 2, 1959, had enabled U.S. analysts to conclude that the Soviet claim that it was a solar satellite was most likely an attempt to cover up a failed attempt to send the spacecraft crashing into the moon’s surface.
⁷⁶

Quick assessments were only part of FMSAC’s work. Wheelon and Graybeal noted in their 1961 article that although “gross features of a Soviet space shot can usually be . . . established within the first few hours by an experienced technical man,” the “variations and nuances of a given flight, however, which can be equally important, may require weeks of concentrated effort by a team of subsystem specialists working together.” The results of such analysis would be a clear picture of mission performance and the system’s technical features.
⁷⁷

Thus, the CIA was able to report to the White House on June 1, 1964, that
Zond 1,
which had been launched two months earlier, would reach the vicinity of Venus on July 20. The report noted that “we cannot yet tell whether it will impact on Venus or fly by, perhaps ejecting an instrumented probe to explore the planet’s atmosphere as it goes by.” (
Zond 1
came within 62,000 miles of Venus, but the failure of its radio prevented any data from being returned.)
⁷⁸

FMSAC personnel also sought to unravel the failure of the unmanned
Cosmos 57
mission, which was placed into orbit on February 12, 1965. Analysts examined the data from radars that tracked the spacecraft’s movements in space and scrutinized intercepted telemetry in their attempt to understand the purpose of the mission and why the spacecraft burned up not long after its first orbit.
⁷⁹

The space walk of Alexei Leonov a few weeks later pointed FMSAC analysts in the right direction. By comparing the telemetry from
Cosmos
57
and
Voshkod 2
, they were able to determine that
Cosmos 57
was a test for the automated system that operated the airlock Leonov needed to pass through. They also determined, from intercepts of signals to and from
Cosmos 57,
the key channels on which commands were transmitted to the spacecraft from the Soviet control station, as well those from the spacecraft to ground controllers that told how it was responding.
⁸⁰

The explanation analysts pried from the intercepted signals was that while the spacecraft was in range of signals from one transmitter, it received a duplicate set from a second transmitter that was intended to pick up communications with the spacecraft when it flew out of range of the first. The double signals were merged into a single signal that instructed the spacecraft to fire its retro-rockets in preparation for descent. Possibly because of the mass of deployed airlock, the spacecraft then began tumbling some seventy-eight times a minute. Only because the airlock operations could be carried out manually was the Leonov space walk able to proceed as planned.
⁸¹

The
Luna 9
mission of early 1966 was also of great interest to FMSAC analysts because a successful soft landing would have returned the Soviet Union to leadership in the lunar race. U.S. collection sites, including the Army’s Sinop facility in Turkey and the NSA’s STONEHOUSE facility in Ethiopia, enabled U.S. analysts to monitor launch, orbit, and ejection and to determine that
Luna 9
’s trajectory was on target.
⁸²

From the intercepted telemetry, analysts determined when the spacecraft’s engine had been fired to send it cruising toward the moon. On February 3, it was oriented for retromaneuver while STONEHOUSE, Jodrell Bank, and other sites listened in order to provide FMSAC, DEFSMAC, and other interested parties with data to analyze.
Luna 9
landed softly that same day. Its first signals from the moon included telemetry as well as what was soon recognized to be a fax transmission. In the United States and England, fax machines were modified to convert the signal into pictures—with the result that some of the pictures obtained at Jodrell Bank were published before the Soviets officially released them.
⁸³

Of greater concern than the Soviet space program was the Soviet missile program—particularly the ICBMs and submarine-launched ballistic missiles (SLBMs) that could hit U.S. territory. From 1964 through mid-1966, the Soviets began testing three third-generation ICBMs—the SS-9 Scarp, SS-11 Sego, and SS-13 Savage. They also conducted tests of their SS-7 Saddler and SS-8 Sasin missiles.
⁸⁴

Some issues about those programs, such as the numbers produced and deployed, the locations of deployment sites, and targeting policy, were considered outside of FMSAC’s charter. The technical intelligence analysts at FMSAC worried about key characteristics of the missiles and their warheads—missile size, yield, accuracy, range, throw weight, vulnerability to defensive systems, whether the warhead could be airburst, and whether the missile was liquid- or solid-fueled.
⁸⁵

Such technical details were not merely of academic interest to missile designers, for they could have significant strategic implications. They were a key element in determining whether the Soviet strategic forces could strike certain types of targets, overwhelm missile defenses, or destroy U.S. ICBMs in a preemptive strike.

The information mined by FMSAC analysts, only some types of which were available for any given missile launch, included optical, radar, and telemetry data. To estimate the size and shape of a reentry vehicle, photographs were most helpful but were rarely obtained. Less directly indicative, but still useful, were radar cross-section data.
⁸⁶
From estimates of size, estimates of yield followed.

Extracting intelligence from telemetry data required not only a facility for technical analysis but also ingenuity. Soviet missile designers knew what aspect of the missile’s performance each channel of telemetry measured and how that performance was being measured, but FMSAC analysts did not. In addition, they usually were confronted with an incomplete set of telemetry, since during the 1960s the United States was rarely able to gather telemetry during the earliest launch stage because Tyuratam was over the radio horizon from U.S. eavesdropping antennae.

Despite such handicaps, FMSAC analysts could rely on the fact that certain basic measurements, such as acceleration and fuel pressure, were required during any test, and that the numbers associated with particular aspects of the missile’s performance would behave in a certain manner. Thus, when the system feeding propellant to an engine shut off, measured pressure would drop to zero in considerably less than a second, and the turbine would take four to eight seconds to coast to a stop.
⁸⁷

An analogy can be drawn to a car: A person riding in a car would expect to have a set of gauges on the dashboard measuring speed, oil pressure, and other aspects of the car’s performance. Even if those instruments were placed in unconventional positions in the car and no units of measurement were indicated, it would be possible to correlate specific instruments with the car’s behavior to determine both the function of an instrument as well as the units of measurement employed. Thus, “given a
fair sample of powered flight telemetry, the analyst can usually say whether the vehicle is liquid or solid-fueled, whether it has a single burning stage or multiple stages, and what ratio of payload to total weight it probably has.”
⁸⁸

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.

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