The Space Reviewin association with SpaceNews

Titan rocket
KH-9 HEXAGON satellite sitting on its launch pad in 1971. (credit: USAF)

A ray of sunshine into a dark world: the future declassification of satellite reconnaissance information

Soon after the declassification of the CIA’s CORONA reconnaissance satellite system, I started traveling around the country—primarily to Massachusetts and California—to interview people who were involved in the program. One of the people that I interviewed on the west coast was Frank Buzard, who had been in charge of the first Corona launches. Buzard had a mind as sharp as a razor and was the source of a great deal of information, some of which corrected the official and documentary records of CORONA’s early days. For example, official Air Force records indicate that Discoverer 1, launched in February 1958, became the first satellite placed in polar orbit (Wikipedia says the same thing, but Wikipedia has been known to contain the occasional error). Buzard said that this was untrue. The real story was that the Air Force enthusiastically issued a press release stating that Discoverer 1 was in orbit before they had any proof. When they failed to get a signal, Air Force leadership told Buzard and another officer to prove that the satellite was indeed in orbit. And that’s what he did, even though the “proof” was not credible, and everyone involved in the program believed that Discoverer 1 fell near Antarctica. Similarly, when Discoverer 2 was launched six weeks later, there were reports that its capsule reentered and landed on Spitzbergen Island in the arctic, where it was probably recovered by the Soviet Union. Buzard threw cold water on that claim as well, essentially saying “do you think that with all that open ocean it just happened to fall on the one tiny bit of dry land?”

Before I spoke to him, somebody had told me that after CORONA, Buzard had gone on to be a program manager for CORONA’s replacement, the massive KH-9 HEXAGON satellite commonly referred to as “the big bird.” HEXAGON is said by those involved to be the most complex machine ever launched into space, an impressive engineering accomplishment for its day. As I was leaving, I mentioned that when the KH-9 was declassified, I’d like to come back and talk to him about it. “That was quite a program,” he replied. “There’s a lot to talk about.” My hope was that I would see him again in a few years.

Maybe, just maybe, in the next few years we might actually learn more about the amazing machines like the KH-9 that collected intelligence and helped win the Cold War.

That was over twelve years ago. I don’t know if Buzard’s memory is still as sharp, but I never had the chance to talk to him about the KH-9, because it still has not been declassified. It was going to be declassified. The National Reconnaissance Office (NRO) declared in 1997 that the post-CORONA systems such as the KH-7 and KH-8 GAMBIT and the KH-9 HEXAGON were obsolete technology and no longer had to remain secret (See “Ike’s gambit: The development and operations of the KH-7 and KH-8 spy satellites” and “Ike’s gambit: The KH-8 reconnaissance satellite”, The Space Review, January 5 and 12, 2009). The then-Director of the National Reconnaissance Office even discussed placing the remaining KH-9 engineering test article—as big as a locomotive engine—on display in the Smithsonian (see “The invisible Big Bird: why there is no KH-9 spy satellite in the Smithsonian”, The Space Review, November 8, 2004). For an agency that had fought the declassification of CORONA a little over two years earlier, it was a pretty amazing turnaround. It was also short-lived. In 1998 something happened and senior intelligence officials put the GAMBIT and HEXAGON declassifications on hold.

I bring all this up now because there is reason to hope that the situation will change and in the next few years some of the reconnaissance systems that the United States built during the Cold War may be declassified. It is possible, although not necessarily probable, but that is a significant difference from only a week ago. Maybe, just maybe, in the next few years we might actually learn more about the amazing machines like the KH-9 that collected intelligence and helped win the Cold War.

A change in secrecy policy

On January 21, President Barack Obama issued several new policy statements and an executive order that collectively called for greater government transparency, changes in policy for the Freedom of Information Act, and “a presumption in favor of disclosure” of information. The fact that these statements and the executive order were issued on his first full day as president indicates the importance that Obama places upon the subject. They stand in stark contrast to the Bush administration, which took until October of its first year in office to address the subject of Freedom of Information Act policy, and over the next eight years imposed far greater secrecy requirements on all levels of government. Clearly the new administration seeks to adopt greater openness and accountability in government than its predecessor.

But as Steven Aftergood of the Federation of American Scientists noted, right now those statements and the executive order only establish a tone, they do not actually change anything… yet. Obama’s actions so far only concern records that are subject to “discretionary release.” Many records like those on intelligence collection are excluded—government officials do not have discretion to release them, the records must be formally approved for release. And senior officials—many of them civil servants who remain in their jobs despite the change in administrations—still oppose their release. In order for more information on GAMBIT and HEXAGON to be released, Obama will probably have to sign an executive order, just as President Bill Clinton signed executive order 12951 in early 1995 declassifying CORONA.

The chances of such an executive order are certainly much greater now than they were only a week ago. But it is too early to hold one’s breath and wait for it. For when it comes to the declassification of intelligence information, logic rarely wins over entrenched bureaucracy and what can best be described as gut-level beliefs that discussing past classified operations, even ones long ago, is wrong.

Declassifying satellite reconnaissance

What exactly happened in 1998 to halt the declassification of the GAMBIT and HEXAGON is unknown outside of the intelligence community. The rumors are that intelligence analysts found definitive proof that India concealed its preparations for nuclear weapons tests from American spy satellites. But how exactly this required the continued classification of obsolete reconnaissance systems that had been designed in the 1960s, and last flew in the mid-1980s, is unclear. After all, in 1998 then-current reconnaissance systems were deceived, apparently through relatively unsophisticated methods of monitoring satellite orbits, something that even an amateur with a pair of binoculars can do. The most likely explanation is that intelligence officials decided that any discussion of satellite reconnaissance capabilities attracted too much attention and should stop, and this put a damper on discussion of releasing obsolete information.

What exactly happened in 1998 to halt the declassification of the GAMBIT and HEXAGON is unknown outside of the intelligence community.

That is not too difficult to accept if you understand bureaucracy, and the often-contradictory logic of classification decisions. When it comes to declassifying information, especially entire categories of information, often a dozen people have to say yes in order for a classified piece of information to be declassified, but only one person has to say no to prevent it from happening. And those saying no do not necessarily have to justify their decision, as secrecy is the status quo, even for information that is decades old.

In addition, keeping historical information classified is relatively inexpensive as it simply sits in some giant secure warehouse, alongside the Ark of the Covenant and the alien remains from Roswell. In contrast, declassifying information costs money because people have to be paid to review it for any remaining secrets. Intelligence officials claim that they lack the money for declassification. Of course, they lack the money because they do not make it a budget priority.

There was never anything approaching unanimous agreement inside the intelligence community about releasing any information on satellite intelligence collection, and so even a small change in personnel or perception, or the argument that there was insufficient money for declassification, could have easily blocked the declassification of the GAMBIT and HEXAGON programs.

Although opposition to declassifying GAMBIT and HEXAGON sprouted during the Clinton administration, it found fertile ground within the Bush administration. The Bush administration, driven largely by Vice President Dick Cheney, adopted policies of secrecy far greater than its predecessor. Cheney believed both that increased secrecy was necessary for national security, and that it would enhance the power of the Executive Branch, which he believed had been eroded over time, beginning with the creation of the Freedom of Information Act in the 1960s. An administration characterized by its short-sightedness (see: Iraq occupation planning) apparently did not pause to consider that a strengthened presidency might eventually be taken over by Democrats. Now they can only hope that the Obama administration does not use its newly-acquired executive authority against Republicans or their interests.

The Bush-Cheney secrecy philosophy had a more subtle effect upon the declassification of historical intelligence information than one might expect, and the National Reconnaissance Office did not retreat to being an entirely black organization like it had been in the past. Whereas the GAMBIT and HEXAGON declassification decision gained traction and then stalled under the Clinton administration, in fall 2002 the US government declassified imagery from the KH-7 GAMBIT and the HEXAGON’s low-resolution mapping camera. Why the imagery was declassified whereas information about the hardware and its history remained secret is still unknown outside of intelligence circles. In fact, the imagery declassification decision apparently caught some within the intelligence community by surprise. The decision was almost certainly influenced by commercial interests and no imagery was released that was any better than that which was commercially available out of a concern that it would prompt foreign complaints about current American commercial remote sensing satellites.

Some other declassification decisions that were also apparently in the works were canceled. One former intelligence official informed me around 2000 or so that the government would soon declassify information on the Glomar Explorer’s recovery of part of a Soviet submarine during the 1960s. That declassification decision, assuming it was actually planned, was halted.

The Bush-Cheney policies had a measurable effect on the declassification of historical intelligence systems. Early in this decade, the CIA released substantial amounts of information on intelligence systems at the National Archives. The material was contained in the database of the CIA REcords Search Tool, or CREST. As an example, early in this decade a search using the keyword “NRO,” which included not only records that mentioned the NRO but also ones that had been reviewed for declassification by the NRO, initially turned up over 2,000 records. Problems with optical character recognition and the inclusion of foreign language documents meant that perhaps a third of these records had nothing to do with the NRO. The vast majority of the rest were CORONA-related, but included the occasional unexpected revelation, like mention of spy balloons over the Soviet Union in the 1960s, or joint aircraft-satellite operations.

But throughout the Bush administration each year the total number dropped—to approximately 1,400 documents, then to 1,100, then 900, 700, 400, until a point where the latest annual document release contains perhaps a hundred relevant documents. Much of the drop is probably due to the lack of any more CORONA material in CIA files. The fact that GAMBIT and HEXAGON remain classified means that no new records are eligible for release. But decisions made during the Bush-Cheney years also helped dry up other eligible records. For instance, in 2002 the NRO obtained permission to exempt “operational files” from release under the FOIA as part of the intelligence authorization act. However, it was left to the NRO to define “operational files,” which the agency defined broadly to include historical records, including official histories of its activities.

When it comes to declassifying information, especially entire categories of information, often a dozen people have to say yes in order for a classified piece of information to be declassified, but only one person has to say no to prevent it from happening.

In addition to the steady decline in documents released in the CREST archive, I also experienced the change in policy more directly. At the beginning of the decade I submitted several FOIA requests for documents on the PYRAMIDER communications satellite system that Christopher Boyce revealed to the Soviet Union in the early 1970s, the Satellite Data System program, and NRO policies concerning the Space Shuttle dating from the 1970s. All of these FOIA requests were initially accepted and I was led to believe that at least some documents would be declassified. But after several years they were rejected. My suspicion was that in at least some cases officials delayed them until new policies were in place that allowed the FOIA requests to be rejected. In the case of the Shuttle, the NRO eventually replied using the “Glomar response”, refusing to confirm or deny that any records existed, even though the NRO had already confirmed that its satellites had flown aboard the Space Shuttle. Clearly the policy was to deny FOIA requests at the outermost layer of legal defense, even if that policy contradicted previous disclosures.

An odder shade of black

While it refused to release detailed information about its past programs, the National Reconnaissance Office apparently adopted a new policy of acknowledging more past activities in very general terms. The CORONA declassification allowed the NRO to finally begin issuing public awards to “pioneers of national reconnaissance.” During annual awards announcements, the NRO revealed basic details about the awardees’ accomplishments, such as their involvement in signals intelligence collection during a specific period, or the development of new imaging technology and techniques. The descriptions were always very general, but they did represent new information.

Whereas the 2002 release of GAMBIT and HEXAGON imagery had been initiated in 1998, in 2006 the NRO finally acknowledged another program first proposed for declassification in 1998: the POPPY signals intelligence satellite program that operated during the 1960s and early 1970s. However, as with the 2002 release, the POPPY announcement contained few details. Clearly the approach adopted under the Bush administration was to only acknowledge the existence and basic details of obsolete systems, but to not release any detailed information as done during the CORONA declassification. Notably, when the government did declassify information on GAMBIT, HEXAGON, and POPPY, it was the result of efforts initiated during the Clinton administration; no declassification initiatives were started and finished during the Bush administration.

There is one minor exception to this overall trend: in July 2008, the NRO suddenly acknowledged something that had been reported for decades, that the agency operated radar reconnaissance satellites. Unlike other declassification decisions, this was done because of a current requirement: the extreme secrecy surrounding such programs prevented the NRO from cooperating with non-NRO personnel to develop future systems.

Quo vadis?

The world has shifted in other ways that have nothing to do with a new presidential administration. The Bush administration already established a precedent for partial releases of information (for instance, acknowledging the KH-7 but not the KH-8), so internal arguments that programs had to be declassified in all-or-nothing fashion no longer have credence. A declassification decision could release some information but withhold other information, such as the best quality imagery (and, of course, pictures of Israel).

Similarly, commercial satellite imagery continues to catch up to past American reconnaissance capabilities, and satellite reconnaissance technology has proliferated around the globe to a point where over a dozen nations have some form of their own reconnaissance satellites. Now that half-meter resolution imagery is readily available, and 0.4-meter resolution imagery is on the verge of being available, there is less reason to refuse to discuss past American systems that used obsolete technology.

There are other arguments in favor of undertaking the hassle of declassifying obsolete reconnaissance information, such as to improve the image of the intelligence community, particularly the NRO, after nearly a decade of negative publicity and some embarrassing recent missteps like the Future Imagery Architecture.

The new Obama statements indicate a desire to develop a presumption of release of information—in other words, rather than requiring those seeking the release of information to prove why it is necessary to release it, government officials must prove why it is necessary to keep it secret. It is good that this will be the basic assumption of the new administration, because it is more consistent with democratic principles. But intelligence officials start with an assumption that their actions and programs (or sources and methods) are inherently secret, so it is doubtful that they will be burdened with the task of proving why their history must remain secret.

Claims that history is important are unlikely to sway government agencies that are concerned with current events and their insufficient budgets for dealing with them. Most government agencies, even the civilian ones that conduct no secret operations, generally make little effort to preserve their history. The claim that openness about past actions is necessary for accountability in a democratic society is also not a winning argument for bureaucrats who view accountability in narrow terms of responding to elected officials at budget time.

But there are other arguments in favor of undertaking the hassle of declassifying obsolete reconnaissance information. One reason is to improve the image of the intelligence community, particularly the NRO, after nearly a decade of negative publicity and some embarrassing recent missteps like the Future Imagery Architecture. In fact, this was the reason why the NRO went from opposing the CORONA declassification to advocating the GAMBIT and HEXAGON declassifications. The NRO leadership—and many in the rank and file—saw that it improved their image and demonstrated that the NRO had done great things.

Today, when military space acquisition is universally viewed as a mess, and some people blame the situation on a loss of engineering talent, it is worth asking if improving the NRO’s image by releasing information on past engineering successes could help attract new talent. Maybe the NRO can use its bragging rights to appeal to a new group of employees, and appeal to the kinds of people who will become future pioneers of reconnaissance.

There’s no way to know. But if the Obama administration takes the next step and declassifies some of these systems, I do know one thing: I’ll quickly be on an airplane to California to start knocking on some doors.