The Space Reviewin association with SpaceNews

Samos F-2
The Samos F-2 signals intelligence (ferret) spacecraft being prepared for launch. (credit: Paul Gatherer)

Ferrets of the high frontier: US Air Force ferret and heavy ferret satellites of the Cold War

Beginning in 1995 the National Reconnaissance Office began declassifying significant amounts of information about its photographic intelligence satellites. But the NRO has released very little information on its early signals intelligence satellites. Because of this, the study of early signals intelligence collection from space has been restricted largely to the use of the simple analytical tools that independent researchers developed during the Cold War to monitor classified space projects—examination of orbital data and satellite launch vehicles, and the occasional document.

There are signs that this may be changing, however. A handful of newly declassified documents have shed some light on 1960s signals intelligence satellites and may indicate that more information will soon be forthcoming. It is now possible to sketch a more coherent picture of the first decade of American sigint satellites and their role in some of the pressing intelligence debates of that period.

The origins of Air Force sigint satellites

The United States Navy launched the first successful signals intelligence satellite in 1960. The Navy started this project in the wake of Sputnik and rushed it to completion relatively quickly. But it was the US Air Force that had larger and more ambitious sigint satellite ambitions.

The US Air Force’s requirements for signals intelligence were formally established in a September 1958 Air Force statement that was itself a revision of a March 1955 document known as a General Operational Requirement, or GOR. The GOR defined the military service’s requirements for a weapon system. GOR-80 included several addenda defining different satellite intelligence systems. The addendum for an “electronic reconnaissance subsystem,” GOR-80-2, stated:

The electronic reconnaissance subsystem must provide the ability to intercept electromagnetic emissions from potential enemies, to return the intercepted information in a secure manner to an appropriate location, and to record this information in a form suitable for further processing.

The GOR also stated:

In order for the system to provide maximum intelligence, a capability for interception of communications (COMINT) signals is desired. Integration of COMINT and ELINT functions should be provided, if feasible. The combination of both functions will allow for a more timely and economical utilization of the weapon system.

It is now possible to sketch a more coherent picture of the first decade of American sigint satellites and their role in some of the pressing intelligence debates of that period.

The document also listed the required operational characteristics of the system: It had to provide electronic reconnaissance intercept equipment in the band of frequencies between 30 megahertz and 40 gigahertz. It should emphasize intercepting new or unusual signals. It should be able to intercept not only pulse signals, but continuous wave, AM and FM and unusual modulations, and attempt to preserve the original modulation to the greatest extent possible. It should have a direction finding capability with an accuracy of within five miles (eight kilometers), although this capability could not come at the expense of other capabilities. It should be able to be focused on specific areas of interest as well as broadened when the satellite was over areas with few signals. It should be capable of storing intercepted data from one orbit to facilitate readout during a later orbit when the satellite passed near a ground station. And finally, it should provide continuous calibration data in order to produce the most reliable intelligence information.

The Air Force began two projects as a result of this requirement. They were known as Samos F-1 and F-2. The F-1 was a secondary payload to be carried on Samos photo-reconnaissance satellites. The F-2 was to fly as the sole payload attached to the Agena upper stage that placed it in orbit and pointed it at the ground. Three F-1 payloads were launched in 1960-1961, but only one reached orbit. Thus, by 1961 both the US Navy and the Air Force had placed signals intelligence payloads in orbit. But the services’ accomplishment soon attracted attention from the National Security Agency.

Wars over the ether: the NSA vs. the NRO

According to popular movies like Enemy of the State, America’s National Security Agency (NSA) is all-seeing and all-knowing, operating giant satellites capable of lifting a cell phone conversation out of the ether and beaming it back to a glistening chrome and glass headquarters near Washington, DC. These movies, as well as breathless reporting in the European press about the NSA’s Echelon system, have created many false impressions in the minds of people around the globe. But as should be blatantly clear in this era following the terrorist attacks on New York and Washington in 2001 and the failure to locate weapons of mass destruction in Iraq, there is much that the National Security Agency is ignorant about. It is no all-seeing, all-knowing superagency. At its core, it is a bureaucracy with expensive tools.

One myth that seems to perpetuate, however, is that the NSA operates spy satellites. The reality is that although the NSA does operate satellite ground stations around the world for receiving satellite transmissions, it does not actually operate satellites. The signals intelligence satellites used by the NSA are designed, built, and operated by the National Reconnaissance Office (NRO). This division of responsibilities stemmed from an early dispute within the Intelligence Community.

In September 1961, after several signals intelligence satellites had been launched, the Secretary of Defense Robert McNamara assigned processing responsibility for their intelligence take to the NSA. This order coincided with the formal creation of the National Reconnaissance Office, which was tentatively assigned responsibility for developing the satellites themselves. Although an Air Force office for managing satellite reconnaissance had existed since September 1960, the creation of the NRO the following year clarified relationships between various agencies with reconnaissance responsibilities, although much remained to be decided. But the increasing value of satellite intelligence began to raise the stakes for those involved.

The director of the NSA, Vice Admiral Laurence Hugh Frost, began arguing that his agency should develop signals intelligence payloads under authority he felt was granted to them by a National Security Council (NSC) document. The National Security Council is a White House advisory board that provides national security advice to the President.

In May 1962 top reconnaissance officials met at the posh Greenbrier resort in the hills of impoverished West Virginia to discuss an agreement between the Secretary of Defense and the Director of Central Intelligence on NRO responsibilities. During the discussion, Frost’s complaint was debated.

Dr. Herbert Scoville, who ran the Directorate of Research at the CIA, pointed out that under White House rules, the Secretary of Defense was the Executive Agent of the NSC and therefore was formally charged with implementing NSC decisions. As a result, it was up to the Secretary of Defense to assign development of signals intelligence payloads and he had already assigned this task to the NRO, not the NSA.

The reality is that although the NSA does operate satellite ground stations around the world for receiving satellite transmissions, it does not actually operate satellites.

Presumably the NSA’s attempt to gain control for developing signals intelligence satellites had to do with the agency’s concern about effectively meeting its intelligence requirements. But as Scoville pointed out, the NRO already received specific direction on signals intelligence requirements from the United States Intelligence Board, or USIB, which included representatives from various civilian and military agencies, including the NSA. In order to assuage NSA officials, Scoville suggested that “the NRO will look to NSA for guidance and assistance in interpreting the USIB requirements.” But Scoville felt that the NRO should be clearly in charge: “The NRO will make the decisions and will develop, operate, and turn the collected product over to NSA. Interaction will be handled on a case-by-case basis, and NRO will consider NSA views.” Scoville felt that a formal letter from the Secretary of Defense clarifying these responsibilities would be useful in calming the controversy.

Unlike the CIA, Air Force and Navy, the NSA did not have a component office within the National Reconnaissance Office. This inevitably meant that NSA’s requirements were filtered through a committee—the USIB—before reaching the NRO. Certainly the agency would have been first among equals on the USIB as far as signals intelligence requirements were concerned. But its interests were still diluted by other voices.

This was not the situation for some of the NRO component offices, which had greater ability to influence or even initiate the design of signals intelligence satellites without having to gain the explicit approval of the USIB. For instance, only two years after this dispute the CIA initiated a major signals intelligence satellite, ultimately named RHYOLITE, entirely on its own initiative, and without a stated requirement from the USIB. Such a situation must have been enormously frustrating to members of the NSA, which after all was the agency tasked with intercepting foreign communications and electronic emissions.

Also in early 1962, the NSA sought the creation of a separate security compartment for handling satellite signals intelligence. Dr. Joseph Charyk, the Director of the NRO, pointed out that the information was already classified at a high level, known as TALENT-KEYHOLE, and stated that “there does not appear to be any good reason for two separate security systems.”

In early August 1962 the NSA signed a “Memorandum of Agreement Concerning NSA Participation in the National Reconnaissance Office.” The agreement stated rather obliquely that “In the course of discussions between Dr. Charyk, Dr. Scoville, Admiral Frost and Dr. Tordella on 25 May 1962, it was made known that the National Reconnaissance Office in its SIGINT program will respond only to requirements levied by the United States Intelligence Board.” Left unsaid was that this was “made known” over the strenuous objections of Admiral Frost.

In fact, the dispute between the CIA and the NSA had burst into a full-fledged fight. Vice Admiral Frost appealed the decision over control of signals intelligence satellites to McNamara, who told Frost that the decision was final. At that point Frost apparently threatened to further appeal the decision to the President. McNamara then fired Frost and replaced him with Air Force Lieutenant General Gordon Blake.

The term “ferret” is derived from the aerial electronic intelligence community, where aircraft commonly would attempt to elicit electronic emissions from radars by faking attempts to penetrate their adversary’s territory, just like a ferret would burrow into the ground in pursuit of prey.

The NSA would provide advice and consultation to the NRO on how best to meet the requirements levied by the USIB and would provide one of its personnel to become a full-time member of the NRO. The NRO would “assign primary responsibility for development of certain aspects of the SIGINT collection program to NSA.” This probably meant that the NSA would be responsible for developing the ground stations for receiving the satellite data. NSA would also be responsible for advising the NRO on the desired format of the data to be collected and would be responsible for supervising analysis and reporting of the intercepted signals.

The fact that the NSA was in many ways locked out of determining requirements for signals intelligence satellites soured relations between that agency and the NRO for many years. It was only during the 1970s that the agencies forged closer ties.

Samos F-2
The Samos F-2 signals intelligence satellite launching from Vandenberg Air Force Base. (credit: Paul Gatherer)

The Agena ferrets

Throughout the 1960s, signals intelligence satellites were designed, developed, and operated by the US Navy and the US Air Force, working within the framework of the National Reconnaissance Office. The Navy’s efforts were partially revealed with the declassification of the GRAB program in 1998, but the Air Force efforts are still cloaked in greater secrecy.

Air Force sigint satellites developed during the early 1960s fell into two categories, the “heavy ferrets” and the ferret subsatellites. The heavy ferrets were the direct descendants of the Samos program and the Air Force requirements established by GOR-80-2 in 1958. The ferret subsatellites, however, appear to have been started in 1962, after the end of the Samos F-1 program. The term “ferret” is derived from the aerial electronic intelligence community, where aircraft commonly would attempt to elicit electronic emissions from radars by faking attempts to penetrate their adversary’s territory, just like a ferret would burrow into the ground in pursuit of prey.

The heavy ferrets are better designated as Agena ferrets, for they flew as sole payloads aboard Agena spacecraft, which also served as the upper stage for their Thor-Agena launch vehicles. The term “heavy ferret” is slightly misleading, as these payloads were certainly lighter than most other reconnaissance payloads flown during the 1960s, and they clearly fit inside relatively small launch shrouds aboard Thor rockets, which by the latter 1960s were in the small-lift category for American military launch vehicles, behind the Atlas and the Titan 3.

These Agena ferrets had several official designations, although it is unclear if some of these designations were applied concurrently. Some satellite projects had both classified and unclassified designations. In particular, formal “program” numbers served as an unclassified means of identifying a project without indicating its payload.

The most common classified designation for intelligence satellites was a code name as part of the BYEMAN security control system. Names like “CORONA,” “GAMBIT,” and “LANYARD” (which almost always appeared in capital letters) would never appear in documents that were not strictly controlled under the rules of the BYEMAN security system. They could not be used in an unclassified setting.

In addition, satellite programs could also have many other designations. For instance, the CORONA photo-reconnaissance satellites had a classified mission number such as 1052, a classified spacecraft designation such as KH-4A, a classified manufacturer-supplied camera type designation such as J-1, and an unclassified Air Force program number such as Program 162.

The last designation, the program number, was often used when referring to the program in an unclassified setting. For example, if an Air Force officer assigned to CORONA had to travel on official business, his travel papers would only state that he was working on Program 162 business and would not include any of the classified designations.

American signals intelligence satellites during the majority of the 1960s were apparently not given BYEMAN code names, but were instead assigned program numbers. This is not unprecedented, however, as other highly classified satellite projects, like the early military weather satellites and some of the latter Samos photo-reconnaissance satellites, never had BYEMAN code names either.

Official program designations remain somewhat of an enigma. Although they were not classified, project managers still became concerned when they were publicly reported in the press and identified with a particular satellite effort. For instance, the Defense Support Program (DSP) missile warning satellite had several different program numbers during the late 1960s; a new one was applied after the old one was exposed in the media. Despite their unclassified nature, security officers were justifiably concerned that the numbers could be used by an adversary or the press to track the progress of a specific satellite project—just as they can be today.

The first designation applied to the Agena ferrets was Samos F-2. The F was simply a sequential designation in the alphabet and did not stand for “ferret.” The first Agena ferret launch is the only confirmed Samos F-2, but before the first flight a total of four F-2 satellites were planned. Further details on the Samos F-1 and the origin of the F-2, including information about the shift from the Atlas to the less powerful Thor rocket, can be found in the article, “Tinker, Tailor, Radar, Spy: Early American Ferret and Radar Satellites,” in the July 2001 issue of Spaceflight.

Since the publication of that article, further information has been released by the NRO confirming speculation that there was indeed a Samos F-3 program and that the Samos F-4 was also planned but canceled. A February 11, 1960 Lockheed report discussed the conversion of Samos spacecraft from an Atlas-Agena launch vehicle to an upgraded Thor-Agena then designated the Thorad. The purpose of shifting these payloads to the Thorad was to reduce costs and to lower pressure on the Atlas launch facilities, which were then scheduled to launch a large number of Samos spacecraft.

Official program designations remain somewhat of an enigma. Although they were not classified, project managers still became concerned when they were publicly reported in the press and identified with a particular satellite effort.

The Lockheed document stated that both the Samos F-2 and Samos F-3 could be shifted from the Atlas to this new vehicle. The F-2 was described as “An electromagnetic reconnaissance system using digital techniques, with solar power, and launched on a polar orbit at 320 nm [593 kilometer] altitude.” The F-3 was described as “An electromagnetic reconnaissance system using video techniques, powered by solar cells, and launched on a polar orbit at 320 nm altitude.”

The document indicated that the weight of the F-2 payload was 412 pounds (187 kilograms) and the F-3 payload was 350 pounds (159 kilograms). This did not include the weight of related equipment such as communications equipment that was considerably heavier for the F-3 than the F-2.

According to the report, at the time four F-2 and one F-3 payloads were scheduled for launch aboard Atlas-Agena rockets, with the first F-3 scheduled for launch in October or November of 1962. Lockheed indicated that switching these payloads to the Thorad would not affect their launch schedule.

Based upon this information and assuming that the F-3 program, unlike the F-4, was not canceled, it now seems likely that the fifth Agena ferret launch, in February 1964, was the first Samos F-3 satellite. That assumption is further supported by photographic evidence of an increase in the size of the payload shroud starting with the fifth launch. The number of F-3 satellites planned or built remains unknown, but it seems likely that three satellites of this type were launched before Lockheed switched to an updated Agena, which probably coincided with a payload change.

The first Agena ferret launched on February 21, 1962, under the designation Program 102. The second launch, on June 18, 1962, had a new designation, Program 698BK, although it was also described in a declassified launch report as “the second [launch] of a Program 102 (698BK) vehicle.” These launch reports were apparently prepared at Vandenberg Air Force Base to record basic details of military space launches. Although a complete collection of them does not exist, they do contain limited information on what kinds of payloads were launched from what vehicles and have provided some useful pointers enabling the evolution of these satellites to be tracked.

This Program 698BK designation was applied to the next five launches. However, this designation was also applied to various other launches, including a January 1964 launch of several Naval Research Laboratory payloads and the December 1964 launch of the QUILL radar reconnaissance satellite. Program 698BK therefore appears to have been a catch-all designation applied to non-photographic intelligence satellites launched from Vandenberg. The Air Force also used other similar designations for satellites during this time period, such as Program 694BK, for the early classified weather satellites, and Program 698BJ, for the June 17, 1962 launch of the second Samos E-6 reconnaissance satellite, which had previously been designated Program 201.

No explanation has yet been found in historical files for the designations in the six hundred/dual letter series. It is apparent from contemporary documents that Air Force record-keepers occasionally found these designations confusing, and they were dropped after approximately two years.

Beginning with the eighth launch in July 1965, the Agena ferrets switched to an upgraded version of the Agena vehicle, as indicated by a change in the Agena serial numbers. The first seven launches all had Agenas with serial numbers in the 2300 series. But the eighth launch had Agenas with serial numbers starting with 2702, indicating a vehicle block change.

It is at this point that the Air Force redesignated the spacecraft, for the February 1966 launch of the ninth Agena ferret was designated Program 770. A document referring to this launch indicates that it was the third satellite launched with that designation and two Navy POPPY signals intelligence satellite launches in 1965, under scientific cover, shared the Program 770 designation.

The next two Agena ferrets, launched December 29, 1966 and July 24, 1967, also had this designation. But confusingly, a May 31, 1967 launch of several Navy satellites and an Air Force ferret subsatellite was also designated as Program 770. Thus, Program 770, like Program 698BK, also appears to have been a catch-all designation that covered non-photographic intelligence satellites launched from Vandenberg Air Force Base, including not only Navy POPPY satellites, but Air Force Agena ferrets and ferret subsatellites.

The eighth Agena ferret launch, which included a switch to a new block type of the Agena upper stage/spacecraft, also coincided with a switch to a lower inclination orbit. All previous Agena ferrets had orbital inclinations of around 82 degrees. The eighth launch had an inclination of 70.2 degrees. The ninth launch returned to an 82.1-degree inclination, but the tenth and all subsequent launches were all around 75 degrees. No reason for this is known, although it is possible that increased payload weight prevented the rocket from reaching a higher inclination orbit.

page 2: shroudology >>