The Space Reviewin association with SpaceNews

treaty signing
Richard Nixon and Leonid Brezhnev signing the Strategic Arms Limitation Talks Interim Agreement, or SALT I, and the Anti-Ballistic Missile (ABM) Treaty in May 1972. The satellites operated by the National Reconnaissance Office were used to monitor the treaties. They were euphemistically known as “national technical means.” (credit: Richard Nixon Library, White House Photo Office Collection)

Arms control and satellites: early issues concerning national technical means

Bookmark and Share

In 1972, the United States and Soviet Union signed the Anti-Ballistic Missile Treaty and the Interim Agreement, collectively known as SALT I. A phrase that appeared in the treaty is “national technical means of verification.” This was an agreement by the two parties that they would verify the treaty without on-site inspections, using their own assets. Both sides also agreed not to interfere with these “national technical means.”

It appears as if the high level of confidence expressed in intelligence reports on Soviet strategic weapons numbers and deployments by the late 1960s imbued government officials with confidence that the satellites were capable of detecting any new Soviet weapons that might violate treaty limits.

“National technical means” served as a euphemism for each country’s technical intelligence systems. Although these assets included ground, airborne, and other intelligence collection systems, the primary intelligence collectors for treaty verification were satellites, which both countries had been operating for over a decade, but neither country publicly discussed, certainly not with each other. The story of how American intelligence satellites became used for treaty verification has received relatively little attention. Declassified documents now make it possible to explore the issues surrounding their acceptance for this task in the late 1960s and early 1970s, although classification still presents limitations. Surprisingly, there appears to have been little initial skepticism on the American side about the ability to verify strategic arms control treaties using satellites. In fact, there are indications that by the early 1970s there was an overestimation of their capabilities, although the people who developed and operated them were concerned about their limitations, as well as the misperception about what they could do versus their actual capabilities.[1]

This article discusses how the National Reconnaissance Office (NRO), which managed the National Reconnaissance Program—a term encompassing the American satellite intelligence program—addressed this new task of treaty verification in the late 1960s and early 1970s, and the challenges it presented to the NRO.

Satellite reconnaissance

The first successful American photo-reconnaissance mission took place in August 1960 as part of the CORONA program. CORONA involved orbiting satellites equipped with cameras and film and recovering that film for processing. The early satellites orbited for approximately a day before their film was recovered, and it could take several days for that film to be transported and processed before it could be looked at by photo-interpreters in Washington, DC. Although the system was cumbersome, the intelligence data produced by each CORONA mission was substantial, revealing facilities and weapons systems throughout the vast landmass of the Soviet Union.

CORONA’s images were low resolution, capable of revealing large objects like buildings, submarines, aircraft, and tanks, but not providing technical details about many of them. In 1963, the National Reconnaissance Office launched the first GAMBIT satellite, which took photographs roughly equivalent to those taken by the U-2 spyplane that could not penetrate Soviet territory. Both CORONA and GAMBIT returned their film to Earth in reentry vehicles. By 1966, CORONA was equipped with two reentry vehicles, and GAMBIT was equipped with one, increased to two reentry vehicles by August 1969. The existence of multiple reentry vehicles on satellites and missiles was to become a source of concern for NRO officials as new arms control treaties were negotiated.

The two satellites complemented each other: CORONA covered large amounts of territory, locating the targets, and GAMBIT took detailed photographs of a small number of them, enabling analysts to make calculations about their capabilities such as the range of a missile or the carrying capability of a bomber. These photographic reconnaissance satellites provided a tremendous amount of data about the Soviet Union. That data was combined with other intelligence, such as interceptions of Soviet missile telemetry, to produce assessments of Soviet strategic capabilities. Signals and communications intelligence, collected by American ground stations around the world as well as satellites operated by the NRO, also contributed to the overall intelligence collection effort.

By the mid-to-late 1960s, these intelligence collection systems, particularly the photo-reconnaissance satellites, had dramatically improved American understanding of Soviet strategic forces and capabilities. A 1968 intelligence report definitively declared, “No new ICBM complexes have been established in the USSR during the past year.” As a CIA history noted, “This statement was made because of the confidence held by the analysts that if an ICBM was there, then CORONA photography would have disclosed them.” This kind of declared confidence in the ability of satellite reconnaissance to detect Soviet strategic weapons soon proved key to signing arms control treaties.[2]

At a speech in 1972 after the end of the CORONA program and at the establishment of a classified museum display in Washington of CORONA equipment, the Director of Central Intelligence Richard Helms sang CORONA’s praises:

It has been confidence in the intelligence estimates that has allowed President Nixon to enter into the Strategic Arms Limitation Talks and to sign the Arms Limitation Treaty this month. There can be no doubt that the photo reconnaissance satellite represents the primary means of verification for SALT, or that CORONA, the program which pioneered the way in satellite reconnaissance, deserves the place in history which we are preserving through this museum display.[3]

To date there are limited details about debates within the highest reaches of the US government about whether reconnaissance satellites could effectively monitor arms control treaties in the later 1960s, although that issue would reemerge periodically in later decades. It appears as if the high level of confidence expressed in intelligence reports on Soviet strategic weapons numbers and deployments by the late 1960s imbued government officials with confidence that the satellites were capable of detecting any new Soviet weapons that might violate treaty limits.

The ABM hunt

In addition to counting strategic weapons systems during the 1960s, there was another major area of concern for both President Lyndon Johnson and the US intelligence community, the development of Soviet anti-ballistic missile (ABM) systems. American intelligence sources, including satellites, had indicated in the early 1960s that the Soviet Union was undertaking efforts to develop an ABM capability. Intelligence assessments indicated that by the mid-1960s, the Soviets were making only limited progress on ABMs. However, US domestic politics increased pressure on LBJ who in turn sought better intelligence data.

Starting in 1966, at the direction of higher-level authorities, the NRO increased its intelligence collection efforts focused on Soviet ABM systems. This included the launching of new satellites and the modification of ones already in production. By the late 1960s, the US intelligence community had sufficient data to determine that the Soviet Union was not about to deploy a substantial and effective ABM system. Thus, existing intelligence data made it possible to enter into negotiations on an ABM treaty with confidence that the Soviets were unlikely to substantially upgrade their ABM capabilities, and that any efforts to do so would be detected. That same confidence presumably led to a belief that the treaty could be effectively verified.

treaty signing
The HEXAGON reconnaissance satellite entered service in 1971. The satellite’s four film-reentry vehicles posed a problem for arms control negotiators, because there was no easy way to distinguish between a satellite or a missile equipped with multiple independent reentry vehicles. (credit: NRO)

Satellites’ impact, and the impact on satellites

In August 1963, the United States, Soviet Union, and the United Kingdom signed the Treaty Banning Nuclear Weapon Tests in the Atmosphere, in Outer Space, and Under Water, more commonly known as the Partial Test Ban Treaty. The United States monitored Soviet adherence to the treaty through a variety of means, including US Air Force Vela satellites, which could detect the flash of a nuclear explosion in the atmosphere or space. Vela satellites were, in effect, the first satellites used for treaty verification.

The problem, according to the NASA paper, was that relying on these highly classified reconnaissance satellites for arms control monitoring placed them at risk for exposure. NASA was proposing “the development and utilization of an open satellite system designed for the single purpose of verifying US and USSR adherence to treaty conditions.”

New arms control discussions had started during Lyndon Johnson’s administration, and by spring 1969, the Nixon Administration indicated an interest in pursuing arms control negotiations with the Soviet Union. In mid-May 1969, NASA administrator Thomas Paine wrote a top-secret letter addressed to the Secretaries of State, Defense, the Director of the Central Intelligence Agency, and the Administrator of the Arms Control and Disarmament Agency. He indicated that he was sending them a paper summarizing “a possible new Administration initiative in the field of strategic arms limitations.” The paper was also labeled top secret and was to be handled by the special Talent-Keyhole security control system that covered satellite reconnaissance.

The NASA paper stated that a major problem for developing strategic arms limitation agreements was that of creating a “credible means for verification.” At the time, there was discussion of relying upon “national means of verification,” meaning existing US reconnaissance satellites.

The problem, according to the NASA paper, was that relying on these highly classified reconnaissance satellites for arms control monitoring placed them at risk for exposure. NASA was proposing “the development and utilization of an open satellite system designed for the single purpose of verifying US and USSR adherence to treaty conditions.” This approach “would not require the disclosure by either the US or the USSR of the existence, scope, utility, or sophistication of the present overhead reconnaissance programs of both nations. The concern over international confrontation on this issue could thereby be minimized, while both parties could rightly claim to be acting in the spirit of using space for peaceful purposes.”The NASA paper added that “the existence of an overt system would provide a reasonable basis for the possible challenges that might become necessary in the event treaty violations were discerned through any cover means.”

Paine’s proposal was not accepted, but it highlighted a fundamental dilemma concerning using reconnaissance satellites for arms control verification: the intelligence community sought to keep the very existence of the satellites classified, and yet knowledge of their existence was necessary to reassure members of Congress and the public that treaties could be effectively verified. This dilemma never went away but was glossed over with euphemistic language that both the United States and Soviet Union found acceptable.

In the summer of 1969, NRO officials witnessed firsthand how important the issue of treaty verification was to become for their work, and how it could directly impact the procurement of new satellite systems. The NRO then had two major photo-reconnaissance satellite programs in development: the HEXAGON search system and the Manned Orbiting Laboratory (MOL) and its DORIAN optical system. HEXAGON was designed to provide GAMBIT-quality resolution and CORONA-equivalent area coverage. MOL/DORIAN would produce very-high-resolution imagery of specific targets. Both satellites were large and required expensive rockets to place them in orbit.

That spring, the Nixon administration canceled HEXAGON in order to reduce intelligence community budgets. Director of Central Intelligence Richard Helms objected. At that time, Roland Inlow was leading a study on the impact of near-real-time reconnaissance—which did not yet exist—on intelligence production. According to a CIA history, “Inlow, who had been deeply involved in planning for the Strategic Arms Limitation Talks (SALT) in his former position as deputy director of the Office of Strategic Research, had to work quickly. Inlow’s efforts resulted in a series of papers pointing out that the SALT negotiations and the analytical basis for the SALT proposals, then being made by the United States, were all predicated on the availability of large-scale search photography to be obtained by HEXAGON satellites.” In addition, agency studies had identified problems with the MOL/DORIAN vehicle that would blur its high-resolution photographs. Together, this information was used to appeal to President Richard Nixon, and as a result the HEXAGON cancellation was reversed and the MOL was canceled instead.[4]

According to the CIA history: “In November 1969, the Budget Bureau made another attempt to cancel HEXAGON, but there was general agreement in the Intelligence Community that it was needed more than ever, now that the SALT talks were underway.”[5]

What that experience demonstrated to NRO officials was that treaty verification was now a powerful argument in favor of maintaining, and perhaps increasing, NRO capabilities.

Early SALT negotiations

In November 1969, the United States and the Soviet Union began negotiations in Helsinki on new arms control treaties. Because the NRO would have to develop and operate systems to monitor the treaties, and would also be affected by those treaties, NRO officials sought to keep tabs on treaty negotiations.

Air Force Captain John R. Meceda, who was assigned to the NRO, regularly reported on the discussions of the “Verification Group” that was part of the American treaty negotiating team. In October 1969, he reported that the verification group had raised the issue of a ban on multiple objects in space. Such a ban would disallow testing of multiple independent reentry vehicles (MIRVs), but still permit operations of satellite intelligence collection systems.[6] Meceda also reported that the group had discussed the question of what constituted “suspicious new activity” in terms of Soviet weapons development. The group also identified the need for a better definition of a new or improved missile in treaty language.[7]

The United States was at that time developing MIRVs to place atop Minuteman ICBMs. American intelligence experts assumed that the Soviet Union was also developing MIRVs. Meceda explained that a MIRV test ban would stop the United States but not the Soviet Union. Congress would not appropriate the funds for any clandestine testing while the Soviet Union had no such restrictions.[8]

The problem was that a MIRV missile would release multiple objects in space, and then-current and planned NRO vehicles also released multiple objects in space.

One of the methods that the United States used to gather intelligence on Soviet missiles was to intercept the signals they sent to the ground during test flights. Much of this so-called telemetry intelligence, or TELINT, was conducted by ground stations on the periphery of the Soviet Union as well as ships and aircraft that operated under the paths of Soviet reentry vehicles as they returned to Earth. At the time, the NRO had developed rudimentary capabilities to collect TELINT from space, but had in the works a potentially far more capable TELINT collector, a satellite named RHYOLITE, overseen by the CIA. Because TELINT provided important data on missile performance, the United States negotiators were considering a negotiating position that prevented the Soviets from altering their missile telemetry to make it harder to understand. Meceda reported that the verification group also identified a need for a new definition of a ban on suppression of telemetry.[9]

Meceda also indicated that the verification group’s summary report contained a requirement for the control of “throw weight” capability—how much a Soviet or American missile could carry. But Meceda noted that there was no indication of how that could be verified.[10]

The MIRV problem

Issues surrounding any MIRV test ban could have had a potential impact on NRO satellites and were therefore of interest to the leadership. A few days after Meceda’s report to him, Colonel Lew Allen sent a memo to Lieutenant General Royal Allison, who was then assistant to the chairman, Joint Chiefs of Staff, for Strategic Arms Negotiations, explaining the NRO’s view on how a ban on MIRV testing could affect NRO satellites. Allen explained that the NRO had worked on developing a better definition on the release of multiple objects in space to assist the MIRV verification task without seriously impacting intelligence collection systems. “We have not been able to do so,” Allen reported.[11]

The problem was that a MIRV missile would release multiple objects in space, and then-current and planned NRO vehicles also released multiple objects in space. The GAMBIT and CORONA satellites then had two reentry vehicles. The HEXAGON satellite, planned to replace CORONA in the near future, would have four. They returned about one RV per week on a mission. The time could be significantly shortened but not lengthened.

Colonel Allen noted that the reentry vehicles used by NRO satellites had what was known as low betas, which was a description of how much drag they produced in the atmosphere. Vehicles with low betas produced a lot of drag at high altitude and slowed down quickly. Missile reentry vehicles tended to have high betas and did not slow down much in the atmosphere. This was a potential way to differentiate between tests of missile reentry vehicles and satellite reentry vehicles. But Allen added that reentry vehicles could be tested with a wide array of betas, making it harder to identify their mission. The weight of reconnaissance reentry vehicles also varied a lot, but a MIRV could be tested using almost any weight.[12]

Colonel Allen also explained that reconnaissance reentry vehicles contained deboost propulsion packages. MIRV systems would also have individual propulsion packages. Very detailed knowledge of a reentry vehicle would still tell very little about the dispensing system. “In general, therefore,” Allen explained, “none of the observables (time of release, place of release, beta, weight, propulsion) are sufficiently unique for reconnaissance RVs versus weapons system RV’s to permit a precise definition of a ban which would be verifiable.”[13]

“A ban on maneuvering in space has a less profound, albeit significant, impact on current and programmed systems,” Allen added. “Both the CORONA and GAMBIT vehicles use orbit adjust propulsion systems to reorient latitude of perigee and to provide for drag make-up.”[14]

A day after Colonel Allen explained how a MIRV ban would affect satellite intelligence, Captain Meceda sent Allen a memo about the implications of mandatory bans on US satellite intelligence collection activity. Meceda explained that the United States knew the current status of Soviet reconnaissance activity and could challenge their activities if there was a change. But if change in status—such as an increase in the number of reentry vehicles on a satellite—was the basis for a treaty challenge, the first HEXAGON satellite launched would result in a Soviet challenge.[15]

ABM Treaty concerns: the need for more satellites

As preliminary discussions continued within the US government prior to the start of negotiations with the Soviet Union, the NRO evaluated what it would require to monitor different treaty options. The NRO estimated the costs of monitoring several treaty options, calculating the average costs over a five-year period. Because of the requirement that all launches be successful, the planning included backup launches, which increased the overall program cost.[16]

As the negotiations continued, NRO officials realized that those involved in the negotiations might have too-high expectations of NRO capabilities to monitor treaty compliance.

At that time, the average number of planned HEXAGON and GAMBIT missions was four each per year, although the “worst case” to satisfy SALT monitoring would increase that to six each per year. That would cost an additional $96 million for the HEXAGONs and $48 million for the GAMBITs. The additional HEXAGON and GAMBIT satellites were driven by two factors. In the case of the HEXAGONs, the requirement was “driven by double coverage every 6 months of built-up areas to search for land-mobile ABM’s.” The two additional GAMBIT missions in a worst-case situation was “driven by requirement for bi-monthly coverage of naval weapon system targets.”[17]

The NRO was about to introduce two new major signals intelligence satellites operating in much higher orbits than the current STRAWMAN signals intelligence (SIGINT) satellites. The numbers and costs of adding those new SIGINT systems remain classified. The NRO also operated small SIGINT satellites in low Earth orbit which were used to monitor the technical characteristics of Soviet ABM systems. These satellites, part of Program 989, were relatively inexpensive. The plan at the time was to launch two satellites per year, but that would have to increase to three to six per year for SALT monitoring, at an additional cost of $8–28 million. There would also be some cost increase for additional photo processing, as well as upgrades required to the Satellite Control Facility to deal with the occasional periods during which both HEXAGON and GAMBIT spacecraft would be in orbit simultaneously.[18]

Treaty negotiations that started in late 1969 continued in 1970. A third round occurred in late 1970, and discussions in Helsinki ended on December 18, 1970. Toward the end of the discussions the Soviets presented a “ABM only” treaty proposal that included blanks for radius permitted and numbers of ABM systems to be positioned around Moscow and Washington. The Americans interpreted this as the Soviets’ beginning position and not their final position. There were also informal discussions related to verification by “national means.” The Soviet position indicated their “desire to protect their own satellite capability.” “There was an implied ‘understanding’ of non-interference with national means as well as an ‘understanding’ of no deliberate concealment which deviates from established practices of construction and building.”[19]

As the negotiations continued, NRO officials realized that those involved in the negotiations might have too-high expectations of NRO capabilities to monitor treaty compliance. An NRO officer noted that Major Harold Coyle, who represented the NRO within the verification group, needed to “temper the optimism which all parties seem to evidence when it comes to the capabilities of NRO systems.”[20]

In November 1970, Electromagnetic Systems Laboratories (ESL), which was a major contractor providing signals intelligence payloads for NRO satellites and performed other SIGINT-related analysis, produced a report on the STRAWMAN satellite’s value for arms control. Three STRAWMAN satellites had already been launched and a fourth was scheduled for launch in summer 1971 before the program would be retired in the next few years. STRAWMAN was therefore not going to be used for treaty verification. ESL’s report used STRAWMAN as an example of the value and limitations of SIGINT satellites for treaty verification.

The ESL report stated that quantitative improvement in Soviet ABM systems would first be detected by photo-reconnaissance satellites, which would spot new construction of launch sites and support facilities. But an increase in ABM sites would also likely include the deployment of many new radars, which would be detected by SIGINT satellites. SIGINT satellites would be more useful for detecting qualitative improvements by identifying new radars and radar operating modes for the ABM systems, as well as by gathering telemetry from tests, noting that Soviet test facilities were heavily monitored by US intelligence.[21]

As it became increasingly likely that the United States and Soviet Union would be able to sign a treaty on limiting anti-ballistic missiles, the NRO had to determine just how many more satellites it would require to monitor such a treaty. An NRO official suggested that the United States Intelligence Board should begin determining the requirements for verification systems so that NRO could begin to determine its mix of resources.[22]

NRO was concerned “that there may be continuing highly concentrated requirements, such as monitoring a large geographic area for verifying deployment of mobile ICBMs, which would be additive to the normally tasked required coverage.” NRO wanted the Committee on Imagery Requirements and Exploitation (COMIREX), which directed NRO on what targets to image and how much overall photographic coverage was required, to “define the total photographic requirements expected with the advent of a Strategic Arms Limitation agreement with the Soviet Union.”[23]

Revealing satellites

In October 1970, the Soviet Union conducted a series of anti-satellite tests using the Cosmos 373, 374, and 375 satellites. Information on those tests had leaked to the press.[24] In November 1970, the US Air Force launched the first Defense Support Program (DSP) missile warning satellite. But the rocket suffered a malfunction and the satellite was placed into the wrong orbit. Although DSP was not an NRO satellite and had very limited utility for treaty monitoring, it was at the time a mostly-classified system. Some information leaked about the satellite’s improper orbit and had an effect on the negotiations.[25]

Both the Cosmos ASAT tests and the problems with the first DSP satellite appear to have contributed to reports in the French press that a United States satellite had been interfered with. The NRO had no indication that any of its satellites had been affected.[26]

The French reports led to concern at NATO that the SALT agreement might not contain “a specific non-interference with national means clause.” The NRO became concerned that aspects of the US intelligence space program were being discussed “in an international forum.” NRO officials wanted it made clear to NATO members that the US government did not openly discuss the “fact of” satellite reconnaissance. More specifically, the NRO wanted the CIA to “reappraise representatives of the State Department associated with NATO and SALT of the current policy regarding discussion of satellite reconnaissance.”[27]

The NRO was paying close attention to the negotiations at this time. One NRO official wrote an assessment of National Security Decision Memorandum 102, which outlined the latest US negotiating positions.[28] By February 1971, NRO sought to have formal representation as a consultant to the Verification Panel Working Group. This was to ensure that the working group had access to the latest information on the capabilities of the National Reconnaissance Program and its “dynamic nature.”[29]

In their work with the Verification Panel, the NRO sought to temper expectations: “Our main concern has been to moderate the patent optimism regarding verification which has tended to be incorporated in drafts of the studies under review,” an NRO officer wrote to NRO Deputy Director Robert Naka. The problem was that “other agencies have been speaking for the NRO in terms of verification capabilities. We do not feel that we have been in the communication loop sufficiently to preclude this.” An official consultation role would hopefully allow NRO officials to explain the capabilities of the National Reconnaissance Program before negotiators got carried away.[30]

Carl Duckett, the director of CIA Reconnaissance Programs, responded to NRO concerns that people involved in SALT negotiations were talking about reconnaissance satellites.[31] NRO determined that Duckett’s response indicated that the “situation is well in control.”[32]

The Helsinki SALT negotiations were on hiatus in July 1971 as the White House and State Department deliberated. There were some differences of opinion within the US delegation over several issues, including: definition of an ABM system, mutual agreement vs. consultation on non-ABM radars, and inclusion of futuristic ABM systems (such as space-based systems) in the negotiations. There were also some disagreements on offensive systems, including operational status vs. “external completion” of ICBM silos.[33]


A key issue for the United States, and for the NRO, emerged during the treaty negotiations: assuring that the Soviet Union would not interfere with the systems the United States used for treaty monitoring—particularly satellites. By March 1972, paragraph one of Article XII of the Joint Draft Treaty stated, “For the purpose of providing assurance of compliance with the provisions of this Treaty, each Party shall use national technical means of verification at its disposal in a manner consistent with generally recognized principles of international law.” “National technical means” had become the accepted euphemism for intelligence satellites and other intelligence collecting systems. What the NRO specifically wanted was no statement by the US government that it operated intelligence satellites. This was known as a “fact of” statement.

By the late 1960s, American intelligence analysts determined that the SA-5 was not an anti-ballistic missile. But there were some within the US military who continued to argue that it could eventually be upgraded to that capability.

The US had planned to introduce an “interpretive statement” in the record of the agreement, and the Soviet delegation had agreed to remain silent about the agreement. The wording of the statement was: “It is agreed that each side shall use national technical means of verification at its disposal, in a manner consistent with generally recognized principles of international law. This obligation will not require changes from current operating practices and procedures with respect to employment of national technical means of verification.”[34]

But on February 2, the Soviets changed their position and indicated that if the Americans introduced the statement, they would state that Article XII was adequate, and no interpretive statement was required.[35] This resulted in differing opinions among the relevant American agencies. The Joint Chiefs wanted an agreed interpretive statement of legality of current practices, whereas the Office of the Secretary of Defense wanted a formal statement for the record that current practices were legal. The Arms Control and Disarmament Agency, State Department, and CIA wanted no formal statement on the legality of current practices.[36]

Major Harold Coyle was the NRO officer assigned to keep track of the SALT negotiations, and in March 1972 he wrote a memorandum for the record about the latest dispute: “What bearing this has on ‘fact of’ and ‘national means’ insofar as it is associated with satellites remains to be seen. It is speculated that if no interpretive statement is introduced, the Soviets could be in a position to challenge satellites or third country ground stations at some future date.”[37]

In April 1972, the NRO delivered to COMIREX the results of a detailed study of the ability of HEXAGON and GAMBIT satellites to provide required coverage to monitor provisions of a Strategic Arms Limitation Treaty and a Mutual Balanced Force Reduction treaty. The study was the result of computer simulation of multiple satellite missions through 1975, using different assumptions. The fundamental assumption was that the launch rates for the satellites would not increase, but coverage of other targets (not related to the treaties), might be affected in some circumstances.[38]

The SAM upgrade blues

Concern within the US government during the mid-1960s about Soviet ABM developments generally fell into two categories. The first was ABM systems detected around Moscow and Leningrad. These ABM sites were obvious and monitored closely by American reconnaissance systems. The other issue concerned the SA-5 missile system, which initially appeared to have some characteristics consistent with an ABM.

By the late 1960s, American intelligence analysts determined that the SA-5 was not an anti-ballistic missile. But there were some within the US military who continued to argue that it could eventually be upgraded to that capability. CIA official Sayre Stevens, who throughout the 1960s had been involved in many assessments of Soviet space and missile systems, later wrote a classified article titled “The SAM Upgrade Blues,” about the debate that raged in the late 1960s concerning the Soviet Union’s ability to upgrade its surface-to-air missiles to engage ballistic missiles. As Stevens explained it, the primary proponents of the argument that Soviet SAMs could be turned into ABMs were “defense technologists,” not members of the intelligence community. He further claimed that the intelligence community had always been convinced that the SA-5 sites were for shooting down aircraft, not missiles.

Stevens also wrote that members of the intelligence community believed that the idea of upgrading SAMs to shoot down missiles made no sense, and that any plausible missile upgrades would be extremely limited in effectiveness. Nevertheless, a considerable amount of effort was spent examining possible upgrades to both the SA-5 and the much more well-known SA-2 “Guideline” missile that had shot down Gary Powers’ U-2 spyplane over Siberia in 1960 and was regularly shooting American aircraft out of the sky in Vietnam. The CIA concluded that the SA-2 would be almost useless in an ABM role, but the SA-5 still had sufficient unknowns about it that some people in the US government argued that it was at least a potential threat. This controversy lingered as the United States began negotiating with the Soviet Union over an ABM treaty.

In June 1970, Stevens was at the American embassy in Austria during the SALT negotiations when the subject of SAM upgrades came up. Up to this point, the Americans had been pressing the Soviets about their ability to upgrade the SA-5 to make it an ABM. The Soviets had been deliberately evasive about the matter. Eventually, a Soviet official confirmed what the CIA had been telling American officials for years—that in order to “upgrade” the SA-5 to be an ABM system, virtually all of the system’s components would have to be replaced. According to Stevens, the upgrade issue did not completely go away at that time, but he thought that the Soviet admission substantially put it to rest.

Soviet ABM
A Soviet Galosh anti-ballistic missile in its launcher. The Galosh was one of the missiles covered by the ABM Treaty of 1972. (credit: Wikipedia)

The ABM Treaty

In May 1972, the United States and Soviet Union signed the Anti-Ballistic Missile Treaty, which limited the number of sites in each country that could be devoted to ABM systems to two—one for the capital and one for ICBM fields. It also limited the total number of interceptors, and also limited how interceptors could be based—notably banning basing missile interceptors in space.

Article XII of the ABM Treaty contained language of great importance to the NRO:

  1. For the purpose of providing assurance or compliance with the provisions of this Treaty, each Party shall use national technical means of verification at its disposal in a manner consistent with generally recognized principles of international law.
  2. Each Party undertakes not to interfere with the national technical means of verification of the other Party operating in accordance with paragraph 1 of this Article.
  3. Each Party undertakes not to use deliberate concealment measures which impede verification by national technical means of compliance with the provisions of this Treaty. This obligation shall not require changes in current construction, assembly, conversion, or overhaul practices.
Both “non-interference” and “deliberate concealment” would prove to be issues the NRO and other intelligence community agencies would have to deal with in the following decades. Was Soviet development of further ASAT weapons “interference” with national technical means?[39] And what constituted “deliberate concealment” of activities?

The two countries also signed the Interim Agreement, which limited the number of ICBMs and submarine-launched ballistic missiles. The complicated issue of limiting MIRVs was put off for later. And now that a treaty had been signed, the NRO had to focus on the particulars of verifying it.

Aiding the NRO in the verification task were several newly operational satellite systems. The first RHYOLITE telemetry intelligence satellite had been launched in June 1970, providing telemetry intelligence on Soviet missiles. In March 1971, the NRO launched the first JUMPSEAT signals intelligence satellite. Although JUMPSEAT’s primary mission was taking over the tasks of the STRAWMAN satellites monitoring Soviet anti-aircraft radars, JUMPSEAT was reportedly equipped with a staring infrared telescope for monitoring the short-burn rocket engines of ABMs. The first HEXAGON photo-reconnaissance satellite was launched in June 1971, dramatically improving area coverage of the Soviet Union. The degree to which these systems improved confidence within the US government that Soviet treaty compliance could be verified is unknown, but all three were substantial improvements over previous collection systems.

The ABM Treaty limited the number of ABM sites in the Soviet Union to two. Monitoring that would be relatively simple: if the Soviets started building an entirely new ABM site, HEXAGON would detect it. The more complicated verification challenge was detecting new missile and radar development. That would require not only imagery, but signals intelligence, and possibly newer and more complicated methods such as measures and signature intelligence (MASINT) to determine the characteristics of Soviet ABM missiles in flight.

Difficult intelligence requirements such as detecting Soviet MIRVs or mobile ICBMs were going to drive US intelligence collection whether or not they were banned by a treaty. The same would be true for other Soviet weapons developments.

In May 1972, as the ABM Treaty was being signed, COMIREX reviewed the status and coverage of the current imagery launch schedule “in the context of assuring coverage related to possible SALT monitoring needs over the next several months.” They particularly looked at developing a back-up capability, “on standby status,” in case of significant delay or unsuccessful operation of the third HEXAGON mission, then scheduled for launch in June 1972. They determined that if necessary, the best solution would be to move up a GAMBIT-3 mission from August to July.[40]

By June 1972, the Joint Chiefs of Staff issued a paper titled “Requirements to Implement the Measure Proposed by the Joint Chiefs of Staff in regard to the Strategic Arms Limitation Agreement.” The paper called for upgrading and expanding national technical means to meet the needs of SALT verification. The NRO’s view was that when it came to satellites, it was limited by the intelligence budget and was not dependent on SALT. However, as Colonel Bradburn noted, “it appears as if satellite upgrading is receiving the attention of a great many more people who are not normally spokesmen. The result of this attention could be that the interested congressional committees might become confused as to who speaks for the reconnaissance satellite budget.” Bradburn stated that the National Reconnaissance Program was already justified in the Congress independent of other factors.[41]

By this time, Lew Allen was now a Major General and in charge of the NRO’s West Coast office, known as the Secretary of the Air Force Special Projects Office—SAFSP or “Special Projects” for short. SAFSP had oversight of several major operational NRO programs, including JUMPSEAT and HEXAGON. In August 1972, Allen reiterated a concern he had for some time that the NRO had received little guidance on how their “planning should be influenced by the specific verification requirements of SALT.” At least in the near term, current requirements and priority statements for collection would remain valid. Tasking and scheduling of photographic systems would not change. But the formal acceptance of SALT I would require a careful review of the specific provisions in the treaty and how they would set verification requirements.[42]

Allen was most interested in appraising the importance of a specific system for verification of the ABM treaty. Although the name of the system is deleted, he was most probably referring to JUMPSEAT. He noted that changes to the system were already under consideration (notably, JUMPSEAT’s predecessor, STRAWMAN, had been upgraded to include ABM collection). The most relevant part of the treaty was “Statement E” about US concern about upgrades to surface-to-air missile systems “or surreptitious development of a new interceptor with ABM capability which could be deployed under cover of SAM.”[43]

“The identification of an ABM masquerading as a SAM poses difficult problems,” Allen wrote to NRO Director John McLucas. “By definition the program would be concealed and TLM [telemetry] could be denied with no opportunity to protest since operational SAM tests often do not use TLM. Photographic identification would be very difficult unless one believes nuclear warhead facilities are uniquely identifiable. The provisions of Unilateral Statement E define three test conditions which the U.S. would interpret as SAM components ‘tested in ABM mode.’” These conditions were:

  • A SAM launcher is used to launch an ABM;
  • A SAM interceptor is tested in conjunction with an RV, an ABM missile, an ABM radar, or to an altitude inconsistent with aerodynamic vehicles;
  • A SAM radar is tested in conjunction with an RV, an ABM missile, or an ABM radar.[44]

Allen added that the statement “serves to notify the Soviets that we intend to base verification on detection of a SAM ‘tested in the ABM mode.’ This must be considered a statement of national policy and implies a firm obligation to provide the necessary intelligence.” That intelligence would require a system with specific capabilities that are deleted from Allen’s memo, although they included “precision location.”[45]

Allen stated that “the capability to detect surreptitious development of defensive missile systems is the prime justification for the program, and the need for that capability has not seemed time critical. To date, the Soviet activities in ABM development have not proved very hard to identify, and there has been little immediate concern about a surreptitious development. But that is now changed, and the national need for verification that such development is not occurring has been clearly enunciated by high authority. Soviet capability and philosophy is indicated by the increasing use of techniques to avoid satellite detection and the success of EMCON [emissions control] for the Square Pair radar.”[46]

In September 1972, Major General Allen informed the head of the NRO, John McLucas, of the value of a signals intelligence satellite for arms control—most probably JUMPSEAT—indicating that it had “certain capabilities which are unique and crucial to the verification process.” He also believed that any design changes to the satellite made in the future should be prioritized based on the verification mission. Aerospace Corporation had produced scenarios “in which specific violations of the treaty are postulated, followed by explanations of the ways that” the satellite “would be used to detect these violations.”[47] A member of the NRO Staff proposed that General Allen be allowed to engage in direct dialogue with “appropriate members of the Verification Panel,” and that Aerospace Corporation be involved in future work on the subject.

Although continued classification makes it difficult to fully understand how disputes over verification and ongoing negotiations were affecting the NRO, particularly after the signing of the treaty, there are indications that NRO officials knew that they would still be asked about the ability of their systems to detect upgrades to the SA-5, as well as other ABM developments. In October 1972, Major Coyle contacted Colonel Raymond Bradburn about how qualitative missile limitations might relate to a SIGINT issue raised by Colonel Lew Allen. Coyle believed that the CIA had done a good job with meeting those concerns (presumably with the CIA’s RHYOLITE satellite). “One gets the distinct impression that the other collectors are not very useful in SALT verification and that there is little effort being made to make the other collectors more useful,” Coyle wrote.

But Coyle also reflected the view that treaty verification requirements would not substantially alter American intelligence collection in any event. “Collection would be mainly oriented toward the threats to the U.S., whether covered by treaty or not,” he wrote. “Monitoring test range activity is a principal problem. Some of the potential qualitative improvement bans such as a MIRV ban are very difficult to handle from the negotiating standpoint and are difficult to verify today. These bans may never get into a treaty, but they are important intelligence factors. Needed improvements in collection capability can be specified, I feel, with proper coordination.”[48] Difficult intelligence requirements such as detecting Soviet MIRVs or mobile ICBMs were going to drive US intelligence collection whether or not they were banned by a treaty. The same would be true for other Soviet weapons developments.

Another question that emerged was how analysts involved in treaty verification should interact with the NRO. Should that be at the level of NRO Headquarters in the Pentagon, or with the NRO’s West Coast office in Los Angeles, which was actively developing satellite systems and could possibly adjust them to improve their collection of data relevant to treaty verification?[49] There is no unclassified record of how that issue was resolved.


As negotiations on strategic arms limitations commenced in 1969, the National Reconnaissance Office sought to monitor those discussions, contribute where it could, and also protect its own interests, primarily the ability to continue operating its reconnaissance systems. As the negotiations proceeded through 1970 and 1971, NRO officials were less concerned that their satellite operations would be restricted, but still concerned that they could be publicly exposed. They were also interested in any potential increase in requirements or budgets imposed by the need to verify a treaty.

NRO leaders probably viewed arms treaty verification as mostly more work for them, but not a major additional burden compared to their existing mission of monitoring the Soviet Union. This additional work, however, came without additional money.

By the time the ABM Treaty was signed in 1972, the existing evidence indicates that NRO officials did not anticipate any major change in their requirements or budgets—existing NRO satellites could perform the treaty verification task in addition to other demanding tasks such as monitoring Soviet strategic weapons. In 1969, the plan had been to launch four HEXAGON and four GAMBIT missions per year, with a “worst case” scenario—driven by treaty verification requirements—of six missions each per year. But in 1972 and 1973, the NRO launched only three HEXAGONs per year, and that dropped to two per year in 1974 and 1975, and one per year after that. Similarly, whereas the NRO launched four GAMBIT missions in both 1971 and 1972, this dropped to three per year in 1973 and 1974, and only two by 1975 and 1976. Although the ABM Treaty was not highly demanding in terms of verification requirements, obviously the NRO determined that its existing systems were so capable, and improving continuously, that their baseline launch rates could be reduced while still accomplishing treaty verification.

The NRO had to face other issues related to the interpretation of the ABM Treaty language. For example, an American low altitude signals intelligence satellite was at one point illuminated by a Soviet ABM radar near Moscow, an event that could be interpreted as interfering with National Technical Means of verification. The State Department wanted to issue a protest, but the NRO did not want to reveal to the Soviet Union that its satellite had collected data from the radar. There were other incidents where satellite operators suspected attempts to interfere with American satellites. The Soviet Union also encrypted data from some of its missile tests, a possible violation of the “non-concealment” clause. Decisions on whether to issue formal protests were made at far higher levels than the NRO, but NRO leaders sought to protect the existence of their satellites and their capabilities at virtually any cost.

NRO leaders probably viewed arms treaty verification as mostly more work for them, but not a major additional burden compared to their existing mission of monitoring the Soviet Union. This additional work, however, came without additional money. The increase in budgets apparently came later. Hans Mark, who served as director of the NRO in the late 1970s, stated in an interview that the NRO budget was essentially flat in the first half of the decade. The budget did not increase substantially until late in the decade when he sought the approval of large new programs, which he justified as essential for SALT verification.[50] Treaty monitoring finally had a major impact on the National Reconnaissance Program.


  1. There is a small but growing body of scholarship on the history of arms control treaties. See, for instance, James Cameron, “Soviet-American Strategic Arms Limitation and the Limits of Co-operative Competition,” Diplomacy & Statecraft, March 2022; James Cameron, The Double Game: The Demise of America's First Missile Defense System and the Rise of Strategic Arms Limitation, Oxford University Press, 2017; John D. Maurer, Competitive Arms Control - Nixon, Kissinger, and SALT, 1969-1972, Yale University Press, 2022.
  2. Directorate of Science and Technology, Central Intelligence Agency, “CORONA Program History, Volume I: Program Overview,” 1973, p. 9-1.
  3. Ibid., p. 9-5.
  4. Donald E. Welzenbach, “Project HEXAGON, Saved by SALT,” December 18, 1985, p. XI-27. “HEXAGON (KH-9) Mapping Camera Program and Evolution,” NRO, 1982.
  5. Ibid., p. XI-28.
  6. Captain John R. Meceda, USAF, Memorandum for Colonel Allen, “Meeting of the Verification Group Principals,” October 1, 1969.
  7. Ibid.
  8. Ibid.
  9. Ibid.
  10. Ibid.
  11. Colonel Lew Allen, Jr., Memorandum for General Allison, “Impact of Collateral MIRV Test Bans on Satellite Intelligence Collection Systems,” October 6, 1969.
  12. Ibid.
  13. Ibid.
  14. Ibid.
  15. Captain John R. Meceda, USAF, Memorandum for Colonel Allen, “Verification Working Group Meeting, October 7,” October 7, 1969.
  16. Colonel Lew Allen, Jr., Memorandum for Mr. R.L.Hewitt, CIA/OSR, “Additive NRP Costs Associated with Monitoring SALT Option III, IV, or V,” October 30, 1969.
  17. Ibid.
  18. Ibid.
  19. Major Harold S. Coyle, Jr., Memorandum for the Record, “Debriefing on the SALT Negotiations,” January 21, 1971.
  20. Major Frederik L. Hofmann, USAF, Note for Colonel Sweeney, “SALT,” January 19, 1971.
  21. Electromagnetic Systems Laboratories, “Application of a Collection System to Arms Control Verification,” U.S. Arms Control and Disarmament Agency, Task I, November 1970, p. 40.
  22. Major Harold S. Coyle, Jr., Note for Dr. Naka, “SALT Debriefing,” January 21, 1971
  23. Major Harold S. Coyle, Jr., Memorandum for Captain Draim – SS-7, “COMIREX Guidelines,” February 11, 1971.
  24. Major Harold S. Coyle, Jr., Memorandum for Colonel Sweeney and Dr. Naka, “Question Raised at NATO Meeting on Soviet Anti-Satellite Capability,” February 24, 1971.
  25. History of Space and Missile Systems Organization, 1 July 1971-30 June 1972, Robert F. Piper p. 54. Contained in NASA History Division, Reference Collection files: "Air Force Satellites: DSP."
  26. Major Harold S. Coyle, Jr., Memorandum for Colonel Sweeney and Dr. Naka, “Question Raised at NATO Meeting on Soviet Anti-Satellite Capability,” February 24, 1971.
  27. Ibid.; also: Deputy Director F. Robert Naka, National Reconnaissance Office, Memorandum for Mr. Helms, “Discussion of ‘Spy Satellite’ Programs,” March 10, 1971.
  28. Major Harold S. Coyle, Jr., Memorandum for the Record, “NSDM 102 – SALT IV Instructions,” March 16, 1971
  29. F. Robert Naka, Deputy Director, NRO, Memorandum for Dr. K. Wayne Smith, National Security Council, “Consultation to Verification Panel Working Group,” February 24, 1971.
  30. Major Harold S. Coyle, Jr., Memorandum for Dr. Naka, “Consultation to Verification Panel Working Group,” February 22, 1971.
  31. Carl E. Duckett, Director, CIA Reconnaissance Programs, Memorandum for Dr. Robert Naka, Deputy Under Secretary of the Air Force, May 7, 1971.
  32. Major Harold S. Coyle, Jr., Memorandum for Colonel Sweeney and Dr. Naka, “SALT IV,” May 11, 1971.
  33. Major Harold S. Coyle, Jr., Memorandum for Colonel Bradburn and Dr. Naka, “Status of SALT,” July 21, 1971.
  34. Major Harold S. Coyle, Jr., Memorandum for the Record, “A Relevant Issue for SALT VII, Helsinki: Current Verification Practices,” March 16, 1972.
  35. Ibid.
  36. Ibid.
  37. Ibid.
  38. Charles L. Murphy, Deputy Director for Satellite Operations, National Reconnaissance Office, Memorandum for Chairman, Committee on Imagery Requirements and Exploitation, “GAMBIT and HEXAGON MBFR/SALT Simulations,” April 18, 1972.
  39. See: Aaron Bateman, “Mutually assured surveillance at risk: Anti-satellite weapons and cold war arms control,” Journal of Strategic Studies, January 2022, pp. 119-142.
  40. George F. Smith, Jr., Memorandum for Dr. Naka, “Imagery Satellite Scheduling for SALT-Related Coverage,” May 10, 1972, with attached: Henry C. Howard, Acting Chairman, COMIREX, “Imagery Satellite Scheduling for SALT-Related Coverage,” May 8, 1972. At the time, GAMBIT-3 Mission 4335 and CORONA Mission 1117 were both scheduled to be in orbit at the same time. This was considered not optimal, but the NRO’s interest in ending the CORONA program in order to save money was deemed to be a legitimate overriding factor.
  41. Colonel David D. Bradburn, Note for Dr. McLucas, “JCS Paper,” June 6, 1972, with attached: Colonel William L. Shields, Jr., Memorandum for Colonel Bradburn, June 6, 1972.
  42. Major General Lew Allen, Jr., Memorandum for Dr. McLucas, “SALT Verification,” August 28, 1972.
  43. Ibid.
  44. Ibid.
  45. Ibid.
  46. Ibid.
  47. Brigadier General David D. Bradburn, Memorandum for Dr. McLucas, “[Deleted] Capabilities for SALT Verification,” September 8, 1972.
  48. Brigadier General David D. Bradburn, Note for General Allen, “SALT and NRP Systems,” October 16, 1972, with attached: Major Harold S. Coyle, Jr., Memorandum for General Bradburn, “SALT and System Improvements,” October 11, 1972, and: Brigadier General David D. Bradburn, Note for SS-5, Major Coyle, “SALT and System Improvements,” October 16, 1972.
  49. Ibid.
  50. Hans Mark interviewed by Gerald Haines, March 12, 1997, p. 10; Hans Mark, interviewed by Rebecca Wright, Austin, TX – 8 July 2015, International Space Station Program, Oral History Project.

Note: we are using a new commenting system, which may require you to create a new account.