Blank Spots on the Map Read online

  The CIA’s Photographic Interpretation Division at the Steuert Building on Fifth and K Street N.W. grew from 13 people in 1953 to 150 people in 1956. In 1961, Eisenhower merged all of the intelligence community’s photo interpreters into the National Photographic Interpretation Center (NPIC). The NPIC grew like mushrooms in a dank basement. In 1996, NPIC merged with the Defense Mapping Agency and other parts of the intelligence community to become NIMA, the National Imagery and Mapping Agency. More than four thousand of its employees worked in the Washington metro area alone, headquartered in a collection of windowless brick buildings in Bethesda. In 2003, NIMA became the National Geospatial-Intelligence Agency, taking a place alongside the CIA, the NRO, and the NSA as a full-fledged intelligence agency. The NGA probably employs around nine thousand people in the Washington area and in St. Louis, Missouri. Essentially, the NGA is the Pentagon’s geography department. They sponsor professional geography meetings and conferences. They recruit at universities. The people who work for the NGA often have the same training I have.

  It wasn’t just the new armies of photo interpreters that grew to agency size. Throughout the U-2, CORONA, and OXCART projects, the Air Force and the CIA had an uneasy relationship. The Air Force, which provided much of the logistical work for the programs—from launching rockets at Vandenberg Air Force Base to maintaining the bases where the U-2s flew—felt like it was being asked to serve as the CIA’s water boys. The agency set the intelligence collection targets and maintained control over the photographic “product.” For its part, the CIA thought the Air Force was doing exactly what it was supposed to be doing: The CIA, not the Air Force, was the nation’s intelligence agency. With the number of reconnaissance flights increasing, the number of spy satellites on classified drawing boards and in night skies flourishing, and the venom between the Air Force and the CIA over control of it all getting ever more poisonous, President Eisenhower created a new secret agency to formalize the CIA/Air Force relationship in 1960. At first, this was the National Reconnaissance Program. In 1961 it became the National Reconnaissance Office. NASA was the public face of the “space race” against the Soviet Union, with its astronaut heroes like John Glenn and Alan Shepard, with Neil Armstrong walking on the moon, and with the Voyager expeditions to the outer reaches of the solar system. Meanwhile the National Reconnaissance Office, NASA’s black twin, would conduct the dark space race. By the time the NRO’s existence was made public in 1992, it had become, budget-wise, the largest agency in the intelligence community.

  After observing USA 129, a KEYHOLE-class satellite, from his threadbare balcony, Molczan and I sat back down at his living room computer to get an accurate prediction of where our next object, USA 144deb, would be in the sky. On one hand, the vigilant observations made by Molczan, Eberst, and the other satellite spotters had produced an astoundingly accurate map of the other night sky. But the Keplerian elements and TLEs they produced were only part of the story. The other night sky wasn’t a simple system guided through passive inertia. The observers found themselves in an elaborate epistemological ballet with the spacecraft gliding through the night skies above them. When Molczan looked up, the sky looked back; when Molczan made a move, the sky made a countermove. To observe the other night sky was to become a part of it, to change it. The other night sky could tell the truth, but it was just as capable of lying and deceit. A few minutes later, we’d spot a decoy in the sky created to fool both professional and amateur satellite trackers alike. For years, Molczan had been fooled by the USA 144deb object, seduced by looking into the sky and seeing what he wanted to see.

  8

  The Observer Effect

  3,000 km Altitude, 63.4 Degrees Inclination

  “So it should be about sixty percent down from this star to the next,” I say. Molczan and I are doing the final calculations for the USA 144deb overpass, which should be visible in a few minutes. Molczan’s just shown me how to hold the stopwatch—press the wrong button and all your data for the evening is useless. “Try to forget that number,” he says. “You don’t want to bias your observation.” He’s trying to teach me how to see. To get accurate data, he explains, you have to see what’s there, not what you want or expect to see. Getting good data means being as objective as possible; it means trying to see what’s actually there in spite of one’s own disposition. It requires a tremendous amount of self-discipline. Richard Feynman, the renowned physicist (and Manhattan Project scientist), once said that science’s first principle is that “you must not fool yourself, and you are the easiest person to fool.”

  Minutes later, I’m kneeling on a foam pad out on the balcony with my right eye up against the binocular’s eyepiece. A point of light enters my field of view, glittering like a diamond against a black sky. I click the stopwatch as it bisects the guidepost stars Molczan has chosen for me. It looks like it’s about 60 percent down. I can’t tell whether it actually was there or if I’d only seen it there because I’d remembered the number. I continue tracking it for a while by guiding the binoculars along the object’s path. USA 144deb brightens and fades over long intervals as it glides through the sky. Then I lose it.

  I didn’t get any useful data from my observation of USA 144deb, but the attempt to accurately observe it reminded me of two things: first, the necessity of an objective observation; second, the impossibility of a truly objective observation. But there was something else underlying Molczan’s study of the USA 144deb object, another pitfall of seeing, something called the “observer effect.” It’s a principle that holds true for the natural sciences as much as the social sciences: When you observe something you tend to change it. In the sciences, the idea goes back to the early days of quantum physics: To “see” an electron, you have to make it interact with a photon. But the interaction with the photon changes the electron’s path. A more mundane example is the fact that if you want to measure the voltage in a circuit with a multimeter, you have to apply its tips to the circuit you want to measure. But applying the tips changes the circuit’s electrical properties. The observer effect in the human sciences is quite straightforward: People act differently when they know they’re being observed. The observer effect questions the easy distinction between observer and observed: To observe something is to become a part of the thing one is observing.

  There’s a corollary to the observer effect that comes from looking too closely at the secret world. The more you look at it, the more you learn to see. At some point you see it all around you and realize that you’ve somehow become a part of it. As the arch-spy Robert Baer once put it, “Involvement is the first step towards understanding.”

  With more than two decades of watching black spacecraft from his balcony under his belt, Molczan has accumulated a lot of knowledge he isn’t supposed to have. He’s developed sources and methods. He’s learned to see things that are meant to be hidden. Over the course of his hobby, Molczan has become involved.

  On May 22, 1999, the National Reconnaissance Office launched a classified satellite on board a Titan IVB rocket from Vandenberg Air Force Base. Two days before the launch, Molczan reported that he thought the spacecraft would be another member of the ONYX constellation, specifically ONYX 4. On the day of the launch, however, several facts put that theory in doubt. First, the rocket would use a fifty-foot fairing, where previous imaging satellites had used fifty-six-foot and sixty-foot fairings (the ONYX satellites typically used a sixty-six-foot fairing). Florida Today speculated that the payload contained either some kind of next-generation PARCAE constellation or a “ ‘super-secret-who-knows-what’ satellite,” pointing out, “The Air Force has never flown this Titan 4 configuration before and it is believed by military watchers that the NRO has a few new test platforms floating around.”

  When it started to become clear that USA 144 wasn’t going to be an ONYX-class spacecraft, Molczan turned his attention to another program rumored to be active; something code-named “8X.” Supposedly, the 8X concept came out of battlefield commanders’ complaints.
“One of the biggest criticisms [of the intelligence community] during the Gulf War,” an unnamed source told the Los Angeles Times, “was the lack of broad-area photographic coverage—the military wanted to be able to look at all of Iraq at the same time.” The KEYHOLES could take high-resolution photos of specific sites, but they had a narrow field of view. Looking through a KEYHOLE was too much like looking through, well, a keyhole, or a large-aperture astronomical telescope: good for a close examination of Jupiter’s great red spot, but bad for a broad view of the Milky Way. The 8X program was supposed to provide a big-picture look at an operational theater. To Molczan and other independent analysts like Allen Thompson, USA 144 was starting to look a lot like what 8X was supposed to look like.

  About two weeks after the launch, something interesting showed up in the Space Track catalog. Although the military did not publish the requisite TLEs for USA 144, a search for its common name turned up ten international catalog numbers, beginning with 1999-208A and ending with 1999-208L. The line items implied that the launch had generated a cloud of debris that was now in orbit along, presumably, with the “A” object: the spacecraft itself. Watching USA 144 through a pair of binoculars from his balcony, Molczan suspected that something was wrong. The object was dim; he expected something far brighter. USA 144 also seemed to “flash” as if it were tumbling through space, its rotating hull sending flashes of reflected sunlight toward Earth below.

  By the end of June, Molczan and the other observers had found an unknown object in a coplanar orbit to where USA 144 was supposed to have been and decided that it was probably the “A” object, the missing spacecraft. In Molczan’s classified spacecraft catalog, he changed the designation of unknown object 99099A to 99028A, USA 144. The object had settled into a relatively circular 63.4 degree orbit at an altitude of 2,700 to 3,100 kilometers. At that altitude, a standard KEYHOLE satellite would provide a resolution of about 1.5 meters. It seemed to be exactly what the generals had ordered. “USA 144 is the first object in an 8X constellation,” thought Molczan. Mystery solved, but not without some reservations.

  It took almost three years for Molczan to conclude that something was terribly wrong with the object he’d designated USA 144. Listserv murmurings held that it was “something else” or “something really weird.” The problem was that the object didn’t appear to be part of any existing constellation—if it were an imaging satellite like 8X, one would expect it to operate in tandem with others in a similar orbit. There was no other object like it and nothing on the launch manifest pointing to a sister object.

  Something else wasn’t right about USA 144. “One of the things that bugged me,” Molczan confesses, was that “the object had a drag term that seemed too high, about ten times too big. It does have a nice small drag term, except it needed to be about ten times smaller.” It took Molczan years to come around to doing the calculations. He wanted to believe that the object he’d found was USA 144, in spite of visual observations indicating that something wasn’t quite right with it. Its orbit seemed far too affected by perturbations in the upper atmosphere and the surrounding space. “It took a while before my curiosity drove me to do the calculations.”

  One Sunday afternoon in the late summer of 2002, Molczan decided to start digging into USA 144’s subtle anomalies. Sitting at the desk in his apartment, he hypothesized that the strange drag term might be caused by solar radiation pressure, the effect he’d first learned to calculate from observing the PAGEOS satellite in the 1970s. “Once you set your mind to testing it, it’s not difficult. I exploited the fact that objects affected by SRP experience the effects in a predictable cycle: There are periods when there’s a net gain, there are neutral periods, and there are negative periods. One period follows another. You can even predict the dates when the transition occurs for the different modes,” Molczan says. With three years of observations now in hand, Molczan put the historical observations into an SRP equation. When he ran the numbers, they “matched like a glove”: USA 144 was being affected by SRP. Once he knew the effect of SRP on the mysterious object, Molczan could calculate the object’s other characteristics.

  “One neat thing that pops out of this,” Molczan explains. “A by-product of an orbit with measurable SRP is that you can figure out the area-to-mass ratio.” From the observational data, he said, “You have intrinsic brightness; you know other objects with similar intrinsic brightness, if reflective area is a proxy for cross-sectional area, which is true. So you can get mass.” When he ran these numbers, USA 144 got even weirder: “It’s less dense than a rocket body. It’s a light object as far as objects go. It’s about one tenth the mass of a payload. At the upper limit its mass is about a thousand kilograms, and at the lower limit it is a hundred kilograms. We know that the Titan IV could have put far greater mass at that altitude. This tells you conclusively that this can’t be the primary object.” Molczan realized that what the observers were actually seeing was something else. They were seeing what they were supposed to be seeing. The thing they were calling USA 144 wasn’t the real satellite at all: The National Reconnaissance Office had placed something, probably a decoy, in a plausible orbit. Whatever the real USA 144 was, it had vanished.

  Using the SRP analysis to learn about USA 144 was a triumph for Molczan, but it came with a dose of self-criticism. For more than two years, he’d ignored evidence of SRP affecting the object. He could have learned about the deception far earlier if he’d been able to be a more objective observer. “So strong was my belief that the object was USA 144, that I was not open to contrary evidence,” he said. “That made it easy to ignore or explain away inconvenient facts, like the high rate of decay and the slow rotation.” Molczan had allowed his observations to become biased by what he wanted to see.

  But he was now confident that what the observers had been tracking could not be the primary payload and that whatever was going on was some sort of deception, Molczan saw the present rhyming with the past. In a Listserv post, he wrote:

  This mission reminds me of Misty, aka AFP-731, aka 90019B. It was shuttle-deployed into a low 62 deg orbit. A week later, Russia reported that it had vanished, leaving behind only debris. Speculation was that it had exploded. Seven months later, Russell Eberst, Daniel Karcher and Pierre Neirinck found it in a 65 deg, 800 km orbit. Soon after, in early Nov 1990, it disappeared again. . . .

  USA 144 remains more mysterious than ever.

  And so, Molczan started to strongly suspect something that he’d probably never be able to prove: the object he assumed was USA 144 was actually a decoy; its designation should be USA 144deb instead of USA 144. The real USA 144 was AFP-731’s younger, possibly more advanced, sister. Both spacecraft were part of the same stealth satellite program going by the code name MISTY. AFP-731 was MISTY-1; USA 144, wherever it was, was MISTY-2. Once again, the other night sky wasn’t just hiding. It was hiding from him.

  As he thought about the 8X rumors and the MISTY program, something else dawned on Molczan: Perhaps the whole “8X” thing was an elaborate cover story for the MISTY program, a cultural analogue to MISTY’s propensity to hide under debris clouds, to exhibit an exceedingly small signal-to-noise ratio, to “blind them with bullshit,” as the saying goes. “Now, this is pure speculation,” he says, “but given the widespread rumors and almost open acknowledgment of the 8X program in the late 1990s, I wonder whether the Russians might have been fed some disinformation along the lines that the 8X would orbit at a nearly three-thousand kilometer altitude, sixty-three point four degrees.” The point of the ruse would be to “[condition] them to accept the eventual MISTY 2 decoy orbit. Makes me wonder whether or not 8X was purely deception to provide a cover for MISTY 2,” he explains, but admits that he can’t go nearly that far with the available evidence: “I make no great claim for this hunch.”

  The secrecy surrounding the spacecraft is difficult enough to peel away. When coupled to the possibility of disinformation, studying the other night sky becomes an epistemological hall of mirrors.
/>
  That was the crux of it: The black architecture of spy satellites—machines designed to be all-seeing, to serve as the ultimate instruments of rationality, to count Cold War missiles and tanks just as Galileo had counted Jupiter’s moons—those orbiting telescopes, spyglasses, had helped to create a world where one’s own eyes and own ears could not be trusted. Eisenhower had lied about the true purpose of the U-2 in the aftermath of Gary Powers’s fateful flight in 1960. The Discoverer program had been a massive exercise in disinformation. The lies had persisted, grown, become institutionalized. And along with them, a tremendous bureaucracy arose to keep the secrets safe, to keep the lies plausible. It was as if the Pentagon and National Reconnaissance Office had created a world in which two plus two could equal five after all.

  The endless equations from Newtonian mechanics Ted Molczan diligently solved to bring a measure of light to the other night sky were rife with the unknown variables of a new kind of physics. The only constant was the fact that the world was not quite what the Pentagon said it was. Beyond that, Molczan couldn’t really say.

  Retired CIA analyst Allen Thompson puts his training to use trying to understand some of the more secretive episodes in American history. He’s spent years compiling and updating a document called “A Stealth Satellite Sourcebook,” a 174-page compendium of declassified documents, patent applications, news articles, and e-mail exchanges he made available through the Federation of American Scientists (the same organization that Niels Bohr, Leo Szilard, and the other architects of the Manhattan Project had helped form after World War II).