Foreign Relations of the United States, 1969–1976, Volume E–14, Part 2, Documents on Arms Control and Nonproliferation, 1973–1976

35. Memorandum From Jan Lodal of the National Security Council Staff and the Counselor of the Department of State (Sonnefeldt) to Secretary of State Kissinger¹

Washington, March 21, 1974.

SUBJECT

• Threshold Test Ban and Your Moscow Trip

Our previous paper on the nuclear test ban (Tab A) concluded that, of all further limitations on testing, a threshold test ban (TTB) made most sense. At our meeting yesterday you indicated agreement with this and raised specific questions regarding the impact of a TTB on the Trident warhead. Accordingly, this memorandum goes a bit further than the Tab A paper in examining the implications of a TTB and addresses your questions on Trident.

I. OPTIONS FOR AGREEMENT

There are a number of variations on types of TTB.

1. A seismic magnitude TTB. This is the standard TTB, with nuclear explosions being prohibited if they produce a seismic signal whose magnitude crosses the “threshold.” The range of sub-variations is illustrated by specific threshold levels:

a. A threshold set at a high seismic magnitude, e.g., 5.75. This would allow explosions of nuclear yield equivalent to about 140 KT in hard rock (such as at Semipalatinsk. There are insufficient depths of soft rock to prevent explosions of such high yields from venting in violation of the Limited Test Ban.) Such a TTB would be an international laughing matter unless coupled with a firm commitment to decreasing threshold levels, a coupling which has been proposed by the Japanese at the CCD. [2 lines not declassified]

b. A threshold set at an intermediate magnitude, e.g., 4.5. Such a TTB would allow explosions in the range of 5 KT (in hard rock) to 50 KT (in soft rock). [less than 1 line not declassified]

c. A threshold set at a low magnitude, e.g., 4.0. Such a TTB would allow explosions in the range of 2 KT to 20 KT. It would place a premium on “decoupling” (e.g., setting off the explosion in an underground cavity) to increase the explosive yield for a given magnitude. [less than 1 line not declassified] It would be desirable to have collateral constraints prohibiting such decoupling but these would be difficult to negotiate and monitor.

2. TTB plus quota test ban. This could involve any of the above TTBs plus an annual quota, fixed or decreasing, of explosions or of cumulative seismic yields below the threshold.

3. A moratorium on all testing above a given seismic yield. Such a moratorium could be useful if there was the clear will and means on both sides to proceed quickly to a TTB at or below the moratorium’s threshold.

II. KEY ISSUES

It is difficult to generalize about the implications of a TTB, as these vary significantly depending on the threshold and on other collateral constraints. Thus, for simplicity of discussion, the following will address a high-TTB (i.e., threshold at around 5.75), a mid-TTB (4.5) and a low-TTB (4.0).

Political

Like a CTB, a TTB of any level is not as important a measure as it was fifteen years ago. (The U.S. has not tabled or pushed a TTB since 1961.) The SALT process, the NPT and the LTBT have preempted much of the value originally foreseen for a TTB.

The Soviets have favored a TTB only if coupled with a moratorium of undefined length on explosions below the threshold; this is essentially a CTB verified by national means. The Soviets might find a TTB—without the moratorium—acceptable for a variety of reasons:

—It would be a significant reaffirmation of détente;

—It would increase pressure on the PRC in an area where the PRC could not be expected to go along in the near term and would thus have an adverse impact on U.S.–PRC relations;

—It would allow the Soviets to continue some testing and thus avoid any psychological disadvantages vis-à-vis the PRC.

The UK would probably go along with any formulation of a TTB although it would undoubtedly prefer that a TTB not enter into force until full testing is completed on the Super Antelope warhead and re-entry vehicle, which will require about two more years. This could probably be compressed to less than a year on a crash basis. [1 line not declassified]

The French would likely oppose any type of TTB in the near term.

A mid-TTB or a high-TTB would, by itself, not significantly impede proliferation and might well be signed by key non-nuclears who have avoided the NPT, Israel and India. Such signature would lend them respectability and protect their nuclear options. Both are parties to the LTBT. A low-TTB would severely inhibit proliferation by its parties.

Domestically, the initial reaction to any TTB would be favorable. Whether this persisted would depend on the campaign that would be mounted against it, especially for a low-TTB. The Senate would undoubtedly approve a high-TTB or mid-TTB.

Effects on U.S. Weapons Programs

A TTB would have the following major effects on U.S. weapons programs:

—[6 lines not declassified]

—A TTB would inhibit, but not prevent, improvements in weapons of yield greater than the threshold. Some improvements could be made and tested out at reduced yields below the threshold.

—Under any of the illustrative TTBs, the U.S. would be able to develop small “clean weapons.”

—Any TTB would allow us to continue enough testing to maintain the reliability of existing weapons.

—Any TTB would allow us to continue “effects tests,” used for hardening weapons systems and command, control and communication systems to the effects of nuclear explosions.

—If we wished to develop a major hard target counterforce option, a TTB would be inhibiting. [6 lines not declassified]

Effects on Soviet Weapons Programs

The principal security advantage to a TTB which we had foreseen over the past few years was that it would severely limit Soviet development of relatively small, high-beta, hard-target MIRVs. Now, with the development of relatively small MIRVs for the SS–X–17 and 19, which could also be retrofitted on the SS–X–18, and with the SS–X–18 MIRV itself, the Soviets probably have a small enough warhead already tested which, in increased numbers and together with improved accuracy, would give them a silo killer, even if not an optimal one.

[5 lines not declassified] However, as stated above, it is likely that they have already tested warheads for their new ICBM MIRV programs. Timing considerations, such as the time required to perfect a design and then start up a production line, would support this.

Verification

A major attraction of a TTB is that its threshold could be set consistent with the capacity for verification by national means (and consistent with the level of testing we wish to continue).

[1 paragraph (5 lines) not declassified]

While having verification advantages over a CTB, a TTB does have major disadvantages resulting from the threshold itself:

—It is difficult to monitor, as seismic signals are unpredictable and vary depending on the individual path traveled through the earth.

—Unintended violations may result from the unpredictability of seismic signals, increasing international tension. Similarly, it may be difficult to prove an intended violation.

—We would expect the Soviets to test up to the threshold level, while staying conservatively away ourselves. This would create an asymmetry in obligations.

—The threshold level in seismic signal does not correlate directly with explosive yield. A TTB would create a premium on “decoupling” techniques which reduce the seismic magnitude for a given yield, thus tending to circumvent the intent of the TTB. Collateral constraints preventing such decoupling (e.g., geographical constraints, no testing in cavities) would be difficult to negotiate and verify.

Peaceful Nuclear Explosive Devices (PNEs)

Without safeguards, it would be possible to conduct clandestine nuclear weapons related testing in the course of PNE detonations above the threshold. Such possible safeguards are discussed in the paper at Tab A.

If PNEs are covered just as nuclear weapons under a TTB—the most straightforward procedure—a low-TTB and mid-TTB would prevent almost all PNE applications of interest to the U.S. and USSR (e.g., gas stimulation). A high-TTB would allow these (including oil shale recovery) but prevent excavation applications (e.g., canal building) of interest to the USSR.

Bureaucratic Situation

NSSM 195, which will update the NSSM 128 study on nuclear test bans, is examining TTBs intensively. It should be finished by the end of April.

AEC, OSD and JCS oppose a TTB as detrimental to our nuclear posture and because they view it as a slippery slope to a CTB. However, as the TTB would permit some testing to continue, their opposition to it is much less strong than for a CTB. This would be especially true for a high-TTB.

III. CONCLUSIONS AND RECOMMENDATIONS

A TTB, like other further test limitations, does not have overriding political or security advantages or disadvantages.

As we recommended in the Tab A paper, if for overriding political reasons you deem it important that the U.S. take some initiative on the test ban issue, the TTB would be the safest course. The TTB has less verification problems and less domestic opposition than other alternatives and would cause us less problems internationally. Nonetheless, as we believe the Tab A and above analysis clearly demonstrate, the TTB would be relatively meaningless in its contribution to national security. Nor would it receive significant support from domestic political elements pushing for a CTB. It must be understood for what it is—essentially a cosmetic agreement with little real significance.

If you do decide to broach the TTB with Brezhnev, we recommend either suggesting no threshold level or a very high level (such as 5.75) pending completion of the NSSM 195 study and the building of some consensus in the bureaucracy. You might also wish to refer detailed discussion to the technical level (Lodal could pair off with their relevant experts). Your Talking Points are written along the above lines.

Bill Hyland concurs.

1. Summary: Lodal and Sonnenfeldt examined the implications of a threshold test ban and addressed Kissinger’s questions about Trident missiles. Lodal and Sonnenfeldt also outlined various types of threshold test bans, summarized key issues, and recommended that Kissinger raise the prospect of a test ban with Soviet officials if he felt it important that the “U.S. take some initiative on the test ban issue.”

   Source: Ford Library, National Security Adviser, NSC Program Analysis Staff Files, Convenience Files, Box 46, Test Ban Messages, 1974 (1). Secret; Urgent; Eyes Only; Completely Outside the System. Sent for information. All brackets are in the original except those indicating text that remains classified. Tab A is Document 32. For the March 20 meeting, see Document 34.