Over The Horizon

Seismology and Nuclear Disarmament

In February, the Ad Hoc Group of Scientific Experts to Consider International Cooperative Measures to Detect and Identify Seismic Events meets in Geneva to advise the Conference on Disarmament on how a nuclear test ban might be monitored. The group meets twice yearly and, although New Zealand is not an official member of the conference, we participate in the group's scientific discussions.

If an underground test ban treaty were signed, how would compliance be verified? Adequate verification has always been a point of contention for moves towards a comprehensive test ban [Verification Technology Questioned, August 1992].

Since 1976, the Group of Scientific Experts has been working on the design of a seismological surveillance system which would monitor the world for clandestine nuclear tests. Underground nuclear tests behave quite a lot like earthquakes, hence the need for seismological surveillance.

There have been many changes in the political scene since the Group was formed, together with major developments in computers and global communications. It is now feasible to have continuous telemetry from a global network of seismographs to an international data centre, where all the data could be processed in a matter of hours. Rapid analysis is necessary, because in the event of a violation an on-site inspection would be mounted, and that needs to be done before the evidence can be removed, or modified in some way.

The progressive construction of such a network is now under way. It is scheduled to be operational by January 1995, with cooperation from as many countries as can be persuaded to take part. The network will then need to be tested, and to have a full evaluation of its capabilities done. The US is sponsoring a selection of new stations in South America and Africa, and also from Vanda in the Antarctic, with telemetry to a prototype data centre in Washington D. C., as a major contribution to better instrumentation of the Southern Hemisphere.

Why should New Zealand be involved in this? We do have an interest in monitoring the French test site at Mururoa, but we also have a political interest in international efforts to establish a comprehensive nuclear test ban. The task of verifying compliance with a nuclear test ban is threefold. The network must (i) detect and identify any clandestine explosions; (ii) be demonstrably sensitive enough to deter would-be violators of the treaty, for fear of being found out; and (iii) demonstrate that naturally occurring earthquakes are not explosions.

The last is the most difficult, especially given that a one-kiloton explosion is equivalent to an earthquake of about magnitude 4, of which there are very many throughout the world every day. So our local network will be needed to supplement the global net, to verify that events detected globally are indeed natural earthquakes.

Is it possible to discriminate between earthquakes and explosions? Usually. One distinguishing feature is the focal depth -- if it is greater than a few kilometres, clearly the event was not an explosion. But focal depth is difficult to determine accurately, especially on a global scale, so this approach does not always produce an unequivocal result.

Large explosions can be differentiated from earthquakes because of the relative paucity of surface waves they generate, but the reality is that violations of a treaty are not likely to be large events; they will be small, close to the detection threshold, and thus much more difficult to discriminate. Every available piece of seismological evidence will be necessary.

One development that will prove extremely useful is the deployment of seismological arrays. This type of instrumentation was pioneered in the 1950s by the UK, with installations in Scotland, Australia, Canada and India. Others have followed, and their processing capacity has improved with time. They enable sensitivity to be greatly enhanced, with the result that arrays in geologically favourable locations such as Scandinavia and northern Canada are able to detect small events throughout the world. To locate these events accurately, however, requires additional data from shorter distances.

Warwick D. Smith, Chief Seismologist, IGNS

Warwick D. Smith Chief Seismologist, IGNS