Inside the CIA (18 page)

The critical decisions are made by the NRO’s executive
committee, which consists of the deputy secretary of defense, who is the chairman; the science adviser to the president; and the director of Central Intelligence. It is the DCI, in his role as director of the intelligence community, who is in charge of requirements. This gives him the authority to control the satellites, assigning them to different areas of the world and making the decision to turn them on and off. The NRO receives proposals for new systems from each of the intelligence agencies. Larger projects must be approved by the president. Once the NRO approves a project, it is carried out either by the CIA’s Directorate of Science and Technology or by the Air Force Space Command.

Some 70 percent of the funding for the projects comes from the Defense Department, while the rest comes from the CIA funneled through Defense Department accounts.

Day-to-day tasking of the imagery satellites is done by the director of Central Intelligence through the Committee on Imagery Requirements and Exploitation (COMIREX), which is headed by a CIA officer. Usually, the committee meets at the office of the intelligence community staff at 1724 F Street NW in downtown Washington. The staff of 240 helps the director of Central Intelligence coordinate the dozen agencies that make up the intelligence community.

In effect, while the NRO operates the satellites, it is the director of Central Intelligence who clicks the shutter. It is also the CIA, along with the military, that analyzes the results through the National Photographic Interpretation Center.

B
UILDING 213 NEXT TO THE
W
ASHINGTON
N
AVY
Y
ARD IS
a seven-story, concrete complex that consists of three cubelike buildings covering two city blocks. Surrounded by a Cyclone fence topped with triple rows of barbed wire, the complex, at First and M Streets SE in Washington, is manned around the clock. What few windows the buildings have are tinted brown so outsiders cannot see in. Nearly two thousand people go in and out every twenty-four hours.

This is the National Photographic Interpretation Center (NPIC), a component of the CIA’s Directorate of Science and Technology that analyzes and interprets images. The center was started by Arthur C. Lundahl, a brilliant former Navy man and University of Chicago graduate who went to work for the CIA in 1953 when the agency was located in dusty temporary buildings at Twenty-third Street and Constitution Avenue NW. Lundahl became chief of the Photographic Interpretation Center. In 1961, the division became NPIC.
The CIA runs the center, but all the military intelligence services participate in it. For many years, it was located on the four top floors of the Steuart Motor Co., a huge, factory-like brick building at Fifth and K Streets NW in Washington.

At first, the photo interpreters used nothing more than magnifying glasses and measuring devices to help them analyze aerial photographs. Slowly, they built up a library so they could compare one frame with a similar frame taken a year or two earlier, allowing them to pinpoint changes.

“You look at a place and then what was it like last year or yesterday,” Lundahl said. “It’s like looking at a movie. The frames are further apart, but you can infer much more of the intentions by seeing the changes on the ground than by doing it one frame at a time.”
86

To bring out small differences in elevation, NPIC had CIA reconnaissance pilots take pictures separated by dozens of yards of the same area. By using stereo techniques, NPIC could use the exaggerated dimensionality of the pictures. For example, depressions in the grass from tires or people walking would leap out of a photograph when viewed in this way.

For the most part, NPIC concentrated on strategic intelligence and research that could be used to predict long-range trends. Tactical intelligence, which could be used to pinpoint targets for bombing missions, was left to the military.

Whether it happened on earth, in the air, or in the sea, NPIC learned to interpret and make sense of it. For example, photo interpreters looking at Virginia could tell that the people were meat eaters because they could see hog pens and cattle. They could pick out the industrial plants where rail lines enter. They could draw a floor plan of each house based on outside clues: the kitchen is under the exhaust vent on the roof. There is a sewer vent over the bathroom. The living room has a chimney. Everything else in the house is bedrooms. Based on the growth of the cemeteries, photo interpreters could tell how many people died each year. From that, they could estimate the area’s population.

Computers added another tool to the photo interpreters’ bag of tricks. As William E. Burrows described them in his
book
Deep Black,
the computers inside Building 213 were “routinely being used to correct for distortions made by the satellites’ imaging sensors and by atmospheric effects, sharpen out-of-focus images, build multicolored single images out of several pictures taken in different spectral bands to make certain patterns more obvious, change the amount of contrast between the objects under scrutiny and their backgrounds, extract particular features while diminishing or eliminating their backgrounds altogether, enhance shadows, suppress glint from reflections of the sun, and a great deal more.”
87

The computers could even analyze the smoke coming from smokestacks and determine, through spectral analysis, what was being burned. With infrared, which senses heat, analysts could not only see inside buildings, they could tell that a plane, for example, had been on a runway hours after it had left.

The most critical need when NPIC was started in 1961 was to determine what missiles the Soviets had. It was NPIC that counted Soviet missiles during the missile gap debate, demonstrating that there was no gap at all. NPIC went on to use the same techniques to uncover the Soviet missiles being sent to Cuba. By then, NPIC knew what the erectors and transporters for the missiles looked like and what kinds of crates they were shipped in. Even though the missiles were covered with canvas before being deployed, their length, shape, and width gave them away. NPIC analysts developed expertise in interpreting missile packaging.
88

“There was a science of crateology,” Lundahl said. “Also they had shelters for steamrollers and other heavy equipment, so we had shelterology. Cuban troops slept in one kind of tent, Soviet troops in another kind.”
89

“The Soviets used a standard deployment pattern. The cable lines ran in a certain way. There had to be a place for fuel storage. The minute we saw that we knew what it was,” R. Jack Smith, who was assigned to the Directorate of Intelligence at the time, said.
90

Based on its previous experience in analyzing missile deployment in the Soviet Union, NPIC was able to tell President Kennedy how long it would take for the missiles in Cuba to
become operational. This was supplemented with information from the manuals to the missiles provided by Col. Oleg Penkovskiy, the Soviet intelligence officer who began spying for the British in early 1962. Finally, on October 16, 1962, Lundahl presented Kennedy with photographs that convinced him that the missiles really were there. On that basis, Kennedy confronted Khrushchev and by threatening retaliatory strikes, got him to remove them.

“Politically, the Cuban missile crisis had demonstrated that overhead reconnaissance, and satellite reconnaissance in particular, was a stabilizing factor because it greatly reduced the element of surprise and, in the process, lessened the chance of a dangerous, all-out preemptive attack for fear that the enemy was getting ready to do the same thing,” Burrows wrote in
Deep Black.
“The satellites substituted imagery for imagination and provided a realistic look at what the opposition had and did not have.”
91

“We gave our leaders answers, gave real substance to our national estimate, gave enlightenment when there had been darkness,” Lundahl said. “I think we avoided nuclear war a couple of times, particularly in Cuba when people knew exactly what the facts were. We provided a basis for the strategic arms limitations treaty, which for years they referred to as confirmation by National Technical Means of Collection. Generally, the whole litany of our national intelligence was moved steadily into a technical arena, where the scope and speed and detail of information kept track of world events. People from the president on down became accustomed to this kind of service.”
92

NPIC helped to predict that the Soviets would launch an earth satellite before the U.S. and that the Chinese Communists would detonate an atomic bomb. With its help, the CIA also predicted the 1967 Arab-Israeli war, the India-Pakistan war, and the Soviet invasion of Afghanistan. NPIC ensured that the Soviets were complying with arms limitation agreements. In 1990, it was NPIC that first pinpointed the movement of Iraqi troops toward Kuwait three weeks before the invasion. The center’s photo interpreters pointed out that
Iraqi troops had months of supplies of fuel—far more than would be taken on a training exercise. Once the war started, NPIC pinpointed biological and nuclear facilities.

NPIC routinely forecasts harvests, predicts natural disasters, pinpoints marijuana fields, and estimates the size of oil spills and forest fires. Some of the information is provided to other government agencies, such as the Commerce Department, to help them plan relief programs. By tracking the size of wheat and other critical harvests, the Agriculture Department can better estimate what prices will be like on the world market.

“A camera can take a picture of anything, and ipso facto you are involved in everything, whether it be economics or crops or biological warfare or missiles or submarines or missile testing,” Lundahl said. “Anything at all that man does on the face of the earth that is exposed to the sky, you get images that, if interpreted correctly, can tell you a tremendous amount of information.”

Today, nearly all the CIA’s satellites provide data in real time, meaning that a battlefield commander can see enemy movements while they are happening. By combining images from magnetic, heat, radar, and visual-based images, NPIC can zero in on particular questions. This is particularly helpful when the Soviet Union or other countries try to camouflage their facilities.

“It’s a game between hiders and seekers,” Lundahl said. “The art of camouflage, decoys, deception, is very well developed.”

Yet for all the wonders of science, no technology can divine an adversary’s intentions.

“Photo interpretation when well done can give you an excellent view of an enemy’s capability,” Lundahl said. “You can’t get his intentions. You can get some intentions. If you see a furnace is being moved or a railroad track is being torn up, you can tell it is being moved. Or cement is being poured.”

These developments, in turn, may mean the enemy is planning to build a missile silo, for example. But the facility could initially be made to look like a factory. Nor can photos convey long-range plans and intentions.

“Nothing will replace a good human spy,” Roland Inlow, a former chairman of COMIREX, said.

In the area of human spying, too, the Directorate of Science and Technology plays a major role through its supersecret Office of Technical Service.

I
N WINDOWLESS OFFICES WITHIN
CIA
HEADQUARTERS, THE
Office of Technical Service (OTS) within the Directorate of Science and Technology devises state-of-the-art devices for bugging rooms, tapping phones, sending covert messages, and photographing documents. Several hundred CIA employees—including engineers, cabinetmakers, woodworkers, leatherworkers, and physicists—plant bugs or create secret compartments in everything from kitchen cutting boards to felt-tip pens. Over the years, to conceal its devices, the office has used car oil filters, videotape cassettes, false bottoms of toolboxes, toy trains, batteries, cigarette lighters, basket covers, teddy bears, chess sets, paintings, wallets, statues, hot plates, and toilet kits.

For an agent of the CIA in such areas as Cuba or the Soviet Union, these tools of the spy trade are critical for maintaining security and carrying out his mission. An agent must photograph documents, deliver messages to his case officer, receive
cash, and arrange meetings. Of greatest importance, he must have a way to signal in an emergency that he needs to leave—or “exfiltrate”—the country. The CIA usually provides agents with several alternative escape plans. One plan requires the use of a signaling device that sends a coded message to an embassy or other listening post.

Sometimes the simplest methods are best. In Moscow or Havana, a window left open at night or a shade drawn halfway may signal a meeting or a drop at a prearranged site. In such cases, no high-tech methods are necessary. Instead of a custom-fitted wig, an operations officer may slick down his hair and wear glasses.

“I might walk out of the embassy at the end of the day with a suit and tie and long hair and my briefcase,” a former operations officer said. “I might go into the bathroom of a hotel and take off my coat and tie, put them in a briefcase, put on a sport shirt, wet my hair and slick it back, and walk out with maybe my raincoat draped over my briefcase. If I am arrested, I don’t have any spy gear. If you use spy gear, you are in trouble. If you are arrested and they peel your face off, you have a lot of explaining to do.”

Bugging and wiretaps also have their limits.

“You can get too much data,” a former officer assigned to the Office of Technical Service said. “The case officer thinks, ‘I’ll just put a bug in the room, and then I’ll know everything.’ It’s not that easy. Number one, you’ve got to be able to make sure one can do that. Number two, you’ve got to make an installation, and you’ve got to worry about the other side finding it. Then you have to have a listening post where it comes in. You have to have someone changing the tapes. Then you have to translate and transcribe the tapes. Once a month, something worthwhile comes out. But often you are inundated with a lot of marginal data that requires a lot of sifting. Is it worth it? The answer is frequently no. One guy in a key location can often tell you more in five minutes than all of the taping.”