Foreign Relations of the United States, 1961–1963, Volume VII, Arms Control and Disarmament

102. Letter From the Chairman of the NSC Committee on Atmospheric Testing Policy (Seaborg) to President Kennedy⁰

Washington, November 29, 1961.

Dear Mr. President: As requested in NSC Action Memorandum No. 112 of November 13, 1961,¹ a careful study has been made of all of the nuclear weapon test shots proposed by the weapons laboratories and by DOD for inclusion in an atmospheric test program which might be conducted over a three-month period, beginning in the spring of 1962. The proposals were reviewed in the light of the continuing objectives of our nuclear weapons program, our position vis-à-vis the USSR—taking account of the marked advances revealed by their recent test series—and our current state of readiness as adversely affected by the test moratorium. Some 49 possible test shots were reviewed, of which a minimum of 27 are recommended for inclusion in the early program.

The program under consideration has both general and specific objectives. In common with all technologies, the development of nuclear weapons combines the techniques of theoretical and experimental science with the development and design of specific devices. At each stage, data based on past experiments and tests are combined with known theory to devise new and more revealing experiments and improved devices which are subjected to experiment and the process is then repeated. At appropriate points the results are used as the basis for the design of specific weapons. The trends of the developments and the objectives and characteristics of the weapons are, of course, subject to such external factors as delivery capabilities, expected enemy countermeasures, the availability of special materials, etc.

The normal cycle of a year or so was, of course, perturbed by the test moratorium. During the three-year period the information gained from the Hardtack series was exploited about as far as it is feasible to go without further experimental checks. New weapons introduced represent the greatest extrapolations from tested configurations that it is prudent to use without further test. Understanding of the underlying principles and of the probable behavior of extrapolated arrangements was extended by calculation and analysis to the limit in the absence of additional data. Except in a few limited areas, further substantial progress would not be possible without additional experiments. The resumption of underground testing has of course made it possible to progress in those aspects of the program for which experiments and tests are feasible in that environment. There remain, however, larger and more important areas that cannot be covered in an underground program. Hence, over a major portion of weapons technology, above-ground testing is essential to any substantial future progress and recovery of momentum by the Laboratories.

Urgency is added to the need for progress by the substantial advances made by the USSR. In certain areas of the underlying technology, they have surpassed any results that we have achieved or, indeed, know how to achieve with any degree of certainty; some of their devices, notably the larger ones, go beyond our own. Even more importantly, they have a new accumulation of experimental data on which to base still further advances. Especially if exploited through still further tests, these advances could well result in drastic, and possibly decisive, changes in the relative positions of the two countries unless we too make substantial progress.²

Certain specific areas are of special military interest. Each potential adversary now has, or soon will have, a stockpile of strategic nuclear weapons such that if delivered on target, virtual destruction of the enemy would result. Increasingly, therefore, much of the future effort of each side will be directed at measures designed to prevent the delivery of enemy weapons while, at the same time, preserving the deliverability of one’s own weapons in spite of the enemy’s preventive measures. This game of counter and counter-countermeasures can take many forms including (1) measures for destruction of enemy weapons on base—countered by hardened bases, and mobility (aircraft in flight, Polaris submarines, mobile Minutemen, etc.); (2) anti-missile systems including the use of defensive nuclear warheads—countered by hardening of offensive missiles, by clustered warheads and/or decoys and by offensive warheads effective at greater altitudes where defensive measures are more difficult; (3) increased defensive capabilities against aircraft—countered by low level penetration, requiring rugged “lay-down” types of weapons, or by long-range air-to-ground missiles; (4) interference with radar and communications through nuclear explosions as well as by conventional means.

In general, strategic effectiveness can be enhanced by increasing the diversity of methods of accomplishing a given mission, thus complicating the defenses, and by improvements lessening the requirements on the carrier. For example, large dividends can result from improvements permitting weapons of a given destructive potential to be carried by missiles as well as by aircraft, by Minuteman or Polaris as well as by Atlas, etc.

These considerations lead to urgent requirements for (1) increased knowledge of the effects of nuclear explosions on hardened bases and on missiles and on radar and communications—as bases for designing both improved defensive measures and effective counters against enemy defenses; (2) developments leading to decreased nuclear vulnerability of our own offensive warheads; (3) decreased weight-to-yield ratios of strategic missile warheads in order to permit increased diversity and greater pre-strike mobility and to increase penetration capabilities through the use of clusters and/or decoys, through hardening of the individual missiles or through use of greater yields at higher altitudes; (4) developments increasing the effectiveness of our own defensive weapons through enhanced effects and through decreased weight-to-yield ratios, permitting longer range and higher altitude delivery at given yields or greater kill range at given weights.

In the area of tactical weapons, two considerations predominate: (1) developments permitting greater diversity of delivery methods for given missions such as developments permitting the use of simpler and more mobile launching mechanisms; enhanced effects and decreased weight and size are important here; (2) improvements and economies making possible wider deployment of tactical weapons in advance of knowledge as to places of potential need; reductions in requirements for special nuclear material per weapon are of importance in view of the large numbers that are involved. (Clearly, in this area, the number of weapons deployed is far larger than would ever actually be used.)³

In view of the above considerations and our present situation, our test program can be considered under three major headings: (1) measurements of weapons effects; (2) tests of advanced concepts leading to better understanding and to markedly improved and less vulnerable weapons; and (3) tests of actual weapons incorporating substantial extrapolations from previously tested configurations so that both developmental and verification objectives are achieved.

It is clear that measurements of weapons effects must be conducted in the environment to which they will pertain and, hence, require experiments in and above the atmosphere. It is also important that many of the tests required, for example, those relating to anti-missile systems and to improvements in strategic penetration, perforce involve large yields and, hence, either cannot be conducted underground at all or only slowly, with great difficulty.

It has been our experience underground to date that such things as the mechanical difficulties of digging very deep (up to 2,000 ft.), large diameter holes (36 inches), bring into question the feasibility of using the vertical hole approach for testing yields much above 20 KT. We have proposed and did plan that yields up to 100 KT be conducted underground in tunnels; however, our experience with post-shot contamination in the tunnels at the Nevada Test Site up to this time gives us some concern that our planning in that aspect was not realistic. It is now estimated that, if feasible at all, tests of this magnitude could be conducted at a rate of not more than a very few per year, even under favorable conditions, in a given tunnel complex. In contrast, large yield devices can be tested in rapid succession in the atmosphere with relatively little advance preparation of the means of testing.⁴

A second important factor in relative capability is that of diagnostic instrumentation. For test shots which can be satisfied by relatively simple instrumentation plus good radiochemical analysis, tests in the atmosphere can be much more effective in furnishing early information than can underground experiments, at least with current techniques. This is especially important when the results of one test determine parameters to be used on a succeeding test in the same series. (In one, so far extreme, case it required almost two months to do the digging necessary to obtain an adequate radiochemical sample from a shot in the current series.) Even relatively complex instrumentation is easier for atmospheric tests if barge shots can be used.

Both time and place will influence the choice of test shots that can be included in an atmospheric test program. In accordance with your memorandum we are planning on a target date of April 1, but recognize that it may be advisable to postpone the date by a month or so, as more specific plans may require.

The choice of test site will dictate how the tests can actually be conducted. Technically the Eniwetok Proving Ground is the most desirable, extending as it does over a substantial area, with a lagoon suitable for barge shots. All of the contemplated tests could probably also be conducted, though not so well, at Christmas Island. Since Eniwetok has political difficulties and the availability of Christmas Island is at best uncertain, prudence dictates that we be prepared to test elsewhere if necessary. Fortunately, many of the proposed tests could be conducted without a highly-developed island site, although they would benefit from such a site. Some could be done by air drops probably staged from Hawaii with limited instrumentation on some small island, such as Johnston, not suitable for more extensive development; with some degradation of diagnostic information, others could be carried out by air drop over the open ocean using such instrumentation as could be carried in accompanying aircraft or on ships. However, some of the most complex experiments are of questionable feasibility except over an extended land base such as Christmas Island; in the absence of such a base, serious consideration should be given to conducting some of these above ground in Nevada. In any case, for geophysical reasons, one important test to determine the effects of a nuclear explosion on hardened missile bases could be conducted only in Nevada. A special requirement also exists in connection with rocket lifted test shots for which Johnston Island is the logical launch point because of available facilities.

It has been proposed that tests be conducted in space in order to avoid fallout. Preliminary analysis of space vehicle capability and reliability and difficulty of instrumentation indicates that it would cost approximately 100 million dollars over a 2-year period to develop such a capability and about 20 million dollars per shot thereafter. Consequently, it was not considered available for this program.

Another important consideration in planning the test series is the restraint to be placed on yield because of fallout and its political implications. The series now proposed would have a total yield of approximately [less than 1 line of source text not declassified] this compares to the recent Soviet tests totalling approximately [less than 1 line of source text not declassified]. By appropriate test techniques, it should be possible, within limits, to restrict the worldwide fallout, at the cost of producing local fallout. The balance struck would be dependent on the choice of site.

In studying the proposed series of test shots, it is apparent that no great advance in weapon technology can be attributed in advance to any one proposed experiment. Indeed, it is rare that a single test shot can, by itself, lead to a major advance in nuclear weapon technology. Progress in this area is much more likely to come from a broadly based, interrelated series of experiments—some empirical in nature and some the result of careful calculations—from which one can then advance to bolder, more advanced concepts. The group of 27 tests which appears to be a minimum program must therefore be considered in the context of the gains that could be made from this group of experiments as a program, recognizing that they fall logically into three categories.

The first category covers the general area of the effects of nuclear weapons on such things as hardened missile bases, on missiles in flight, on radar and communications and on naval vessels and their equipment. It is axiomatic that weapons effects in a given environment must be understood if we are to be able to plan for the proper design of weapons to fit a particular military requirement. While there have been many nuclear tests which sought to develop information about weapons effects, it is apparent that there are still significant areas of unknowns of importance to the military use of nuclear weapons. Four tests that fall solely in the effects category appear to be feasible for inclusion in an early series, but other recommended tests will have effects implications. The area of greatest interest lies in effects at very high altitudes as they apply to AICBM defenses and to the kill capability of our own AICBM warheads. A series of five such tests has been projected, varying in yields from a [less than 1 line of source text not declassified] to about [less than 1 line of source text not declassified] and in altitudes from 20 Km to above 400 Km. However, practical considerations with respect to development of instrumentation techniques and related preparations make it unrealistic to plan for more than two of these shots in the short time available. It is important that preparations go forward to conduct the others, and perhaps additional ones in the next test series. The effects of antisubmarine warheads are also of interest, as are the effects of the electromagnetic pulse from a surface burst as they relate to the operational effectiveness of hardened missile bases. One each of such tests is included.

The second major category into which test proposals fall is that of advanced concepts for improving weapon effectiveness and decreasing warhead vulnerability. Advances range all the way from exploring changes in [less than 1 line of source text not declassified] design of “standard” devices through the exploring of very advanced concepts for greatly improving the efficiency of thermonuclear “burn”. New ideas in thermonuclear technology promise great enhancement of the fusion-to-fission ratio [1 line of source text not declassified]. Direct practical objectives include significant reductions in the weight-to-yield ratio of all classes of offensive and defensive weapons; decreased vulnerability of our own offensive weapons; enhanced weapons effects, including progress [less than 1 line of source text not declassified] and reductions in special nuclear materials, particularly for tactical and defensive warheads.

[2 paragraphs (37 lines of source text) not declassified]

Although it would certainly be desirable to include all of the proposed tests of advanced concepts, practical political considerations, and the short time available for design and fabrication of devices dictate that some of these proposals might be postponed to a later test series.

[1 paragraph (22 lines of source text) not declassified]

While the program recommended here represents the current best thinking, it must be pointed out that the underground test program particularly in the area of advanced concepts, is almost certain to produce surprises which will dictate changes in any atmospheric series which might be planned now. Accordingly, one must recognize this and retain flexibility to add, substitute, or otherwise modify the program. On the assumption that such flexibility will be possible, for example, the recommended program has not included tests proposed by the laboratories that repeat completed or scheduled underground tests, or other tests that can presumably be carried out underground; in the event of failures in the underground test program, appropriate revisions to the atmospheric series will be required. On the other hand, knowledge gained from successful underground tests may call for elimination or modification of some of the presently proposed atmospheric tests.

In addition to provisions for flexibility, current planning should provide for preparations for a second test series, about a year later, looking to more dramatic advances than are possible in an early time frame. In fact, it is essential to plan for a second series in order to accomplish the very important effects tests that could not be included in this early time scale. In looking at what test devices can be made available in the April to July period, it is apparent that we are suffering to some extent from the three-year test moratorium. During that period, while the United States was negotiating in good faith, the bulk of our nuclear weapon design effort was oriented towards devices that could be stockpiled with adequate assurance without tests. Thus, the climate was not conducive to bold, new concepts requiring experimental checks. The possibility of being able to test seemed very remote. In contrast, the available evidence indicates that from the very outset of the moratorium, the Soviets have anticipated atmospheric testing and have oriented their efforts toward significant advances requiring such testing.

In summary, it is clear that a rate of progress adequate to maintain our relative military posture can be attained only through resort to atmospheric testing; indeed, much vital information on effects and many possible technical advances, would not be realized at all through underground tests alone.

Accordingly, it is recommended that a program to consist of a minimum of 27 test shots as indicated on the attached Enclosures “A” and “B”⁵ be approved for atmospheric testing to be conducted over a period of approximately three months beginning in the spring of 1962. A description of each of the additional tests proposed for consideration by the laboratories and by the DOD, but not now recommended for inclusion in this first series, is also attached as Enclosure “C”. For your information Enclosure “D” lists all known nuclear explosions conducted to date by the various countries.

Sincerely,

Glenn T. Seaborg⁶

0. Source: Kennedy Library, National Security Files, Subjects Series, Nuclear Weapons Testing, 11/29-12/6/61. Secret; Restricted Data.
1. See footnote 2, Document 94.
2. Next to this sentence Bundy wrote a marginal note: “Spell this out.”
3. Bundy’s marginal note next to this paragraph reads: “not impressive.”
4. Bundy’s marginal note next to the upper part of this paragraph reads: “Of course it can be done.”
5. The attachments to this letter are not printed.
6. Printed from a copy that bears this typed signature.