The Wizards of Langley (35 page)

*A review of one of the books on the
Glomar Explorer,
written by an official knowledgeable about technical intelligence projects, referred to “the JENNIFER (sic) project”—indicating that the term as employed in public accounts is misspelled. Logical alternatives include JENIFER and GENNIFER. (John Milligan, Review of “The Jennifer Project,”
Studies in Intelligence
23, 1 [Spring 1979]: 45.)

Carl Duckett’s departure would be a key event in the history of the Directorate of Science and Technology. Just as Bud Wheelon’s arrival allowed an empire to be established, Duckett’s departure marked the beginning of the empire’s decline—although it would not be a steep one. Within months, the directorate’s scientific and technical intelligence units would be transferred to the intelligence directorate. Several years later, the TACKSMAN sites would be gone. In addition, the CIA’s near-autonomy in signals intelligence operations would yield to congressional insistence that NSA manage all the government’s SIGINT activities.

Had Duckett remained, the transfer of the DS&T’s intelligence analysis components may have never been proposed, and if it had, he might have been able to block it. He might have convinced Congress to leave the CIA’s SIGINT operations alone. But his skill as a smooth talker would have been of no help in preventing the fall of the Shah of Iran and the loss of the TACKSMAN sites. Ayatollahs and fanatical Iranian students would, undoubtedly, have been impervious to his charms.

When Duckett departed on June 1, 1976, Leslie Dirks moved up from director of the development and engineering office to take Duckett’s place. Duckett’s deputy, Sayre Stevens, had left a few weeks earlier to become deputy director for intelligence. Replacing Stevens as the directorate’s number-two man was Ernest J. Zellmer, a former submariner, who had served in the Office of Scientific Intelligence under Wheelon and who would play a major role as the directorate’s voice in the National Underwater Reconnaissance Program.
¹

Dirks was different from Duckett both in looks and personality. One account described Dirks, halfway through his tenure as DS&T head, as “a tall, gaunt, balding physicist with the sterile aura of a pathology lab.”
²
Robert Phillips, who had helped establish the TACKSMAN II site and served in the directorate for three decades, recalled that of all the deputy
directors he served under, Dirks was the one for whom he had the least appreciation—considering him aloof, which was perhaps exacerbated by his preoccupation in 1976 with an ill wife. Phillips also found him too heavily focused on the satellite programs he had been managing before his promotion.* He recalled an incident in which he escorted the senior leadership of a foreign intelligence service, one Phillips had helped develop a facet of its technical collection capability, in to see Dirks, only to have him tell the visitors that he “didn’t know you had that capability.”
³

Philip Eckman, who headed the Office of Research and Development during most of Dirks’s term as deputy director, characterized his boss as a perfectionist who could be “tough to work for, pretty demanding”—a description some would probably have used to describe Bud Wheelon. At the same time, Dirks had “an excellent eye for identifying the right person for the right job.”
⁴

Although some aspects of Dirks’s personality may have irritated some, and although the directorate did not remain quite the empire that Wheelon and Duckett created, it did experience a number of triumphs during Dirks’s tenure—the most important being the success of the satellite program he had nurtured for more than a decade. It resulted in his receiving a medal from President Carter and in a nation with dramatically enhanced intelligence capabilities.

Shortly after becoming deputy director for intelligence, Sayre Stevens proposed that the Office of Scientific Intelligence and the Office of Weapons Intelligence be transferred to the intelligence directorate. Stevens felt that the separation of the two offices from the rest of the intelligence-production effort was artificial. Furthermore, he was concerned about the lack of interdisciplinary analysis on foreign military forces—particularly with regard to nuclear proliferation. Analysts from OSI and OWI did not communicate with those from the intelligence directorate. When OSI’s nuclear energy division suggested China might follow a certain path with regard to its nuclear weapons program, it did so
without any discussion with the political and economic analysts from the intelligence directorate. Bringing the technical people into his organization, Stevens felt at the time, was a solution.
⁵

One source of opposition was Dirks, who pointed to the close relationship that existed among the analysts, collectors, and collection system developers in his directorate. In addition, Dirks warned Stevens that “you’ll leave or get thrown out, and they’ll put some political scientist in there” who would ruin everything. Stevens himself was concerned that the funding for OSI and OWI contracts with outside consultants would be jeopardized. Not only was the money involved many times the entire budget for the intelligence directorate, but the DI did not favor contracting.
⁶

But Stevens was able to convince DCI George Bush, and the offices were transferred. The wisdom of that transfer is, even today, a subject of disagreement. Evan Hineman, who served as OWI deputy director from 1974 to 1976 and as its director from 1976 to 1979, believes the transfer was “the right thing to do” because all analytical efforts should have been brought together. According to Hineman, the result was an improved product, with some of the softer sciences rubbing off on the engineering types and some of the harder sciences rubbing off on the social scientists. When he became science and technology chief in 1982, it never entered his mind to attempt to reclaim OWI.
⁷

That view was shared by Rae Huffstutler, who served as director of the Office of Strategic Research (1979–1982) and then as head of the Office of Soviet Analysis (1982–1984). In his view, if one was going to do Soviet analysis, it was necessary to have all the analytical threads; political and military analysts had to work together to capture the threat. In 1996, Stevens, however, characterized the transfer as a “foolish move on my part,” and said that “to some extent” Dirks’s warning about the consequences of his departure proved accurate.
⁸

Dirks’s directorate did receive something in return—the Foreign Broadcast Information Service (FBIS), whose open source collection activities provided a significant portion of the information used by agency analysts. The loss “caused a good deal of heartburn in the DI,” according to Stevens. But it was felt that FBIS was more of a collection activity than an analytical unit and thus belonged in the science and technology directorate.* More important, FBIS was badly in need of modernization, par-
ticularly automation, and the DS&T was much better at such tasks than the intelligence directorate. The “DI culture had them just struggling along,” according to Huffstutler, because the intelligence directorate “didn’t know how to run programs” or justify modernization budgets to Congress.
⁹

The transfer would eventually pave the way for a 1982 modernization program, designated MIDAS, to update FBIS—to move it out of the “green-eyeshade” era and allow it to catch up with industry. CIA representatives visited the
New York Times
as well as the
Philadelphia Inquirer
to learn how a modern news organization worked.
¹⁰

In addition to losses and gains, Dirks’s tenure was also marked by continued efforts to “reform” the Office of Research and Development, which still had a reputation as a “rogue” group, according to Dirks’s choice to lead the reform effort, Philip Eckman. In October 1976, Eckman, who held a doctorate from Carnegie Tech in electrical engineering, was working at Cal Tech’s Jet Propulsion Laboratory in Pasadena. That month he received a call from Dirks, who told him he was coming to Los Angeles and would like to have dinner. Eckman, who was recommended to Dirks by Bud Wheelon’s intended successor, Frank Lehan, almost didn’t go, but finally agreed to meet the CIA’s science and technology chief.
¹¹

At dinner, Dirks briefed Eckman, who knew little about the agency, on the CIA’s science and technology effort and offered him the ORD directorship. According to a brief memoir written by Eckman, Dirks “pitched the wonders of ORD [but] gave absolutely no hint that he had recently proposed to the CIA’s executive council that ORD be abolished!” In late February 1977, Eckman arrived for his first day at work. He expected to stay a few years but remained through March 1989.
¹²

When he arrived, Eckman found an office “which carried considerable baggage, with some questioning its relevance to the rest of the Agency.” In addition, two acting directors followed the tenures of Stevens and Hirsch, giving the office four directors in less than five years. Eckman’s charter from Dirks was to make the office relevant, “make it grow . . . and make it central.” Before attempting to carry out that mandate, ORD’s senior officials
spent time identifying the roles of the CIA and ORD and settled on technology development in support of collection and analysis as ORD’s mission. A reorganization followed that involved creation of a Collection Technology Group and a Processing and Analysis Group. And “for the free spirits who shunned structure, regular management, and orthodoxy, we formed an Advanced Concepts Staff that was intended to be counter-cultural and light on its feet.”
¹³

In addition to organizational changes and streamlining the project approval process, ORD soon sought to change its relationship with its customers. Eckman credited his first deputy, Ed Cates, with getting ORD to reach out to its customers through annual briefings to customer office directors and promoting awareness of the office’s work through an annual accomplishments book.
¹⁴

On the morning of December 19, 1976, a Titan 3D rocket blasted off from Vandenberg Air Force Base. Its mission was to propel yet another KEYHOLE satellite into orbit. Through December 18, there had been 296 days of satellite coverage that year. Although there had been no crises equivalent to the Middle East wars, the Soviet invasion of Czechoslovakia, or the Cuban missile crisis to monitor, U.S. spy satellites certainly had not lacked targets.

The launch went smoothly, with the Titan 3D sending its payload into a polar orbit. Seemingly, yet another KH-9/HEXAGON had been lofted into space. But close observers noted at least one difference. The new satellite had a markedly higher perigee and apogee than previous KH-9 satellites, coming no closer than 165 miles to the earth’s surface—as compared with the normal KH-9 perigee of a little over 100 miles. In an article published in early 1978, space expert Anthony Kenden noted: “A Big Bird was launched on 19 December 1976 into an unusually high orbit, from [153 to 330 miles]. . . . This new type of orbit may indicate that it was the first test of a Program 1010 vehicle.”
¹⁵

Indeed, the payload that was placed in orbit on December 19 was the first launch of the KENNAN program (designated 1010 by Lockheed, the primary contractor), which had successfully produced a spacecraft with an electro-optical system—known as the KH-11. The specific spacecraft bore the designation 5501 to specify the particular satellite and mission number.
¹⁶

The KH-11 was launched within two months of the target date, and despite reports to the contrary, it came in substantially under budget. Before Dirks’s tenure would end, another three KH-11s would be successfully launched—on June 14, 1978, February 7, 1980, and September 3, 1981. In addition, the code name for the program would be changed from KENNAN to CRYSTAL in 1982.
¹⁷

In general, no two versions of any imaging satellite are necessarily identical, because modifications are often made to sensors and other equipment. The basic dimensions of the KH-11 remained the same from the initial launch—the cylindrical spacecraft measured about 64 feet long and 10 feet in diameter and weighed about 30,000 pounds.
¹⁸

The optical system, however, underwent a major change between the first and subsequent satellites. The first satellite relied on light-sensing diodes to collect the light reflected from the target. By the time the second spacecraft was constructed, Lockheed was able to turn to charge-coupled devices, or CCDs. The CCD originated at Bell Telephone Laboratories in the late 1960s when two researchers, William S. Boyle and George E. Smith, invented a new type of memory circuit—a development that had been in the works for most of the decade. The researchers quickly realized that the tiny chip of semiconducting silicon they first demonstrated in 1970 had a variety of other applications, including signal processing and imaging (the latter because silicon responds to visible light). By 1975, scientists from the California Institute of Technology’s Jet Propulsion Laboratory and the University of Arizona were using a CCD in conjunction with a 61-inch telescope to produce a picture of Uranus, about 1.7 billion miles from earth.
¹⁹

The optical system of the KH-11 (and its successors) scanned its target in long, narrow strips and focused the light onto an array of CCDs with several thousand elements. The light falling on each CCD during a short, fixed period of time was then transformed into a proportional amount of electrical charge. In turn, the electrical charge was read and fed into an amplifier, which converted the current into a whole number, between 0 and 256, representing a shade of color ranging from pure black to pure white. Thus each picture was transmitted as a string of numbers—one from each element.
²⁰

More specifically, the CCD captured particles of energy, visible light, in an array of picture elements known as pixels. The pixels automatically measured the intensity of the particles and then would “send them on their way in orderly rows until they are electronically stacked up to form
a kind of mosaic.” The standard CCD used in the Hubble Space Telescope has a total of 640,000 pixels arranged in an 800-by-800 format and occupies less than half a square inch.
²¹