241. Memorandum From the President’s Special Assistant for Science and Technology (Wiesner) to President Kennedy0

The claim continues to be advanced in a number of quarters that the use of unmanned seismic stations in conjunction with the proposed system would so improve our ability to detect and identify underground nuclear tests that the requirement for on-site inspections could be eliminated from our nuclear test ban treaty proposals. While a properly designed system of such unmanned stations could substantially improve the capabilities of our presently proposed monitoring system, and thereby reduce the number of on-site inspections required to accomplish a given level of surveillance, such a system would not eliminate the need for on-site inspection. In fact, such a system, if designed in a technically acceptable fashion, might require greater overall access to the Soviet Union than our present proposals.

It is certainly possible to design an unmanned seismic station that will record or transmit information on seismic events. Although such stations would be much less sensitive than the elaborate manned stations that are proposed as part of the present monitoring system, the close proximity of these stations to events in a seismic area would permit the collection of sufficient close-in data to permit the identification of many events that would only have been detected by the large distant stations. For example, in a recent review of the problem, it was concluded that a properly located network of 100 unmanned stations would reduce the number of unidentified events above magnitude 4.0 (2 kilotons in Nevada Tuff) in the Soviet Union from 75 to perhaps 20 per year. The estimated remaining 20 unidentified events should be recognized as simply a best guess as to the number of natural events whose characteristics would be essentially the same as nuclear explosions. This number of events would remain unidentified more or less independent of the number of additional unmanned stations added to the system. There is, however, inadequate information on the close-in characteristics of earthquakes in the Soviet Union (or elsewhere) so that this number could be considerably larger or smaller. It should be noted that a much smaller number of unmanned stations, possibly only 6 or 10 properly located, could probably accomplish about half the improvement estimated for the larger system.

In addition to its inability to eliminate entirely the need for inspections, the unmanned stations themselves would probably need to be [Page 595] inspected to prevent them from becoming a source of suspicion as to the validity of the entire system. While tamper-proof devices can almost certainly be devised as well as reporting techniques, it would appear that the installation and maintenance of those devices would have to be subject to inspection. If this were not done, there appear to be a number of obvious methods that the Soviet Union could employ in installing the “tamper-proof” equipment provided by this country so that signals from a nuclear explosion would be distorted in such a way as to appear to be from a natural event. This possibility would certainly be pointed out by opponents to a treaty and would at a minimum be a confidence-destroying factor.

Full inspection of the installation and maintenance of a system of 100 unmanned stations would appear to involve a greater degree of inspection of the Soviet Union than is involved in our presently discussed inspection quotas. It is true, however, that these inspections would be at specific locations picked by the Soviets and not over a 200 to 500 square kilometer region whose location could not be predicted in advance. In addition, it would probably be possible to achieve adequate confidence by a periodic random inspection of a small fraction of the unmanned stations. For example, if the Soviets were to install improperly only one out of ten of the unmanned stations to cover areas where clandestine testing might occur, a total of ten annual random inspections would have about two chances out of three, the first year alone, to determine that the equipment was installed with clandestine testing in mind. Nevertheless, even this degree of station inspection when coupled with the continued need for some on-site inspections would on balance appear to rule out this approach as an answer to the inspection problem. We have previously indicated that a small number of inspections, i.e. 10 or possibly fewer, would be acceptable to the United States. I cannot imagine a number less than three being reasonable whatever the seismic detection system.

I would recommend, therefore, that we should not include this proposal in any modifications of our position on the nuclear test ban treaty. At the same time, I would recommend that we be prepared to discuss this proposal with the Soviets if they advance it and to include it in a treaty proposal if they are willing to allow the necessary inspection of the equipment.

  1. Source: Kennedy Library, National Security Files, Subjects Series, Nuclear Weapons, Testing, 8/62-12/62. Secret.