CCD/351, 23 September 1971

Working Paper Containing Comments on the Canadian study of the Seismological Detection and Identification of Underground Nuclear Explosions (CCD/327) 23 and on its implication for the expanded seismic array system outlined in the United Kingdom Working Paper CCD/296.24

Reconciled a Canadian study of the capacity of the 1969 network with the theoretical study of CCD/296. Reaches similar conclusions. as to the limits of seismological techniques. Emphasized that useful technical contributions of this kind are open to all to make for themselves.

CCD/363, 25 April 1972 25

Working Paper on seismic yields of underground explosions-estimating yields of underground explosions from amplitude of seismic signals.

Demonstrated the uncertainties in reconciling a seismic yield with. true yields and the problems these pose for a test ban. Proposed using surface (R) waves in preference to P waves to reduce the scatter of data.

Working Paper on seismic data handling and analysis for a comprehensive test ban.

Described the United Kingdom proposals for developing a seismometer array station processor (SASP) to facilitate the handling, processing and analysis of data under test ban conditions.

CCD/401, 28 June 1973 27

Working Paper on a review of the United Kingdom seismological research and development programme.

Reviewed the United Kingdom programme and introduced the concepts of broad band detection systems and seismogram modelling as means to reduce uncertainties of identifying seismic disturbances, especially when it is suspected that deliberate attempts might be made to test in secret.

CCD/402, 28 June 1973

Working Paper on the estimation of depth of seismic events.

Described the value of and the methods for estimating depths to the focus of seismic disturbances. Introduced a method which effectively broadens the reception band by removing the distorting effect of the

seismograph as well as the transmission path in order to recover the original signal and its surface reflection.

CCD/440, 13 August 1974 28

Working Paper on a development in discriminating between seismic

sources.

Expanded the seismogram modelling technique and proposed its adoption to deal with 'anomalous' events.

CCD/459, July 1975

Working Paper on safeguards against the employment of multiple explosions to simulate earthquakes.

Demonstrated the value of broad band records for revealing attempts to simulate earthquakes by a series of explosions. Proposed an upper limit yield of 50 kilotons to attempts at evasion and advocated not specifying a threshold (for a CTB) in order to reduce this limit at least to half by causing the violator to introduce adequate safety factors in order to be certain of avoiding detection.

British Working Paper Submitted to the Conference of the Committee on Disarmament: The Processing and Communication of Seismic Data To Provide for National Means of Verifying a Test Ban, April 12, 19761

INTRODUCTION

- The need for processing recorded data to make the best use of seismometer arrays was referred to in United Kingdom Working Papers ENDC/155, September 19652 and CCD/296, July 1970.3 It was proposed that the main component of the processing system should operate at the array stations themselves. For this reason it was named a seismometer array station processor, or SASP for short. The advantage of processing at the station was to reduce by many times the quantity of data which would otherwise have to be transmitted to collating centres of which there may be several depending on the degree of national interest in making independent assessments. A processor of this kind however involves a small but very fast computer, and seismometer arrays are not sited in comfortable city squares where the housing of a complex piece of machinery would pose no environmental problems. It is therefore necessary to verify experimentally whether the United Kingdom proposal is realistic or not.

This working paper brings the CCD up to date on the progress of this project. At the same time it reports on the data communication

"See n. 14, above.

1 CCD/487, Apr. 12, 1976.

'Documents on Disarmament, 1965, pp. 408-411.

3 Ibid., 1970, pp. 342-349.

links which were established last year and on which the United Kingdom bases its national capacity to verify a test ban by seismological means alone.

THE PROTOTYPE SASP

To make the best use of the data from the arrays, the usual processing method is to sum each of the seismometers (normally up to 20 in all) after making corrections for the differences in the arrival time of a signal at each seismometer. This procedure is often called "beam forming" or "steering". The seismic waves from areas of interest traverse the arrays with apparent velocities of between 10 kms1 and 25 kms-1 (increasing as the distance to the source increases from 3,000 km to 10,000 km) so the differences in arrival time between one end of the array and the other could be as much as 2s.

Each seismometer in the arrays deployed in association with the United Kingdom is recorded on magnetic tape in analogue form. To delay and sum in a sufficient number of event search conditions to ensure optimum performance, the sequences must be carried out at a speed which only a digital computer can match. All twenty analogue records of a United Kingdom type array are therefore converted to digits before searches can commence.

The steering sequence is carried out off-line to take advantage of the great speed with which computers available in the last few years can operate. This enables the computer to be used by station staff for further analyses of detected signals during a normal 8-hour working day. Off-line processing also means that a breakdown does not affect the data recording systems. There is of course a delay of twenty-four hours built-in to the event assessing procedures, but this is acceptable under the terms of reference.

The method of converting the analogue recordings to digits is relatively new in that the data are divided into two streams, one of which is heavily filtered for signal detection, the other unfiltered for analysis. These two streams are now digitised simultaneously. The method of steering the array by forming a sufficient number of beams is similar to that employed by others, and the problem of false detections due to tape defects and spurious signals is also familiar.

At the time the prototype SASP was being designed, another team concerned itself with the solution of the threat to a satisfactory CTB imposed by the possibility of simulating earthquakes with a series of explosions. As the United Kingdom Working Paper demonstrated last year the safeguard against this device is a broad band recording system and the ideal method of so recording, in order to minimize the background noise by appropriate processing, is by digital techniques.* Bringing the two results together it was self-evident that the next generation of SASP must begin with digital recordings and thereby eliminate the analogue tape recording altogether. This proved to be a

*CCD/459, July 24, 1975.

convenient solution also to the principal disadvantage of analogue recording which was revealed by the prototype SASP. This is the large number of triggers which turned out to be spurious signals generated in the analogue replay system. To avoid some of these "false alarms" it was decided to reduce the speed of editing by half to X8 the recording time. This allows time to employ more elaborate detection programmes, but somewhat reduces the time available for further analysis by the station operators. Nonetheless the prototype system will be useful for research programmes for some years to come because of the library of existing analogue recordings.

A comparison of the SASP detection system was made using the results of the International Seismic Month.

COMMUNICATING THE DATA

After steering the array, and measuring the record so displayed, it is possible roughly to estimate the location and size of each seismic disturbance and to make a first assessment of what it might be. The process is outlined in the Annex to United Kingdom Working Paper CCD/386 of 22 August 1972.5 However, better accuracy in location and size and more certain assessments may be achieved if data from a number of well distributed arrays could be scrutinised together and in some cases be re-processed. It is therefore an advantage to establish rapid means of communicating data each day to other national centres and to have the facility for transcribing and despatching recordings on request. The figure illustrates the arrangements which have been developed so far in the United Kingdom for this purpose. There are two principal means of communication:

(a) by standard commercial services from arrays in Australia, Canada and India with which the United Kingdom has been associated;

(b) by ARPANET from the arrays associated with the United States. (ARPANET is sponsored by the Defence Advanced Research Projects Agency of the United States. In essence, it facilitates the access to the principal computing centres of the United States and Europe with the help of the communication on-line satellite. (see figure))

Between them they test the variety and reliability of methods which are available for national verification of a test ban.

Where the commercial telex is available for only part of the way, the gaps can be bridged by air services or, as demonstrated by India, by a purpose-built radio link, GBA to Trombay. The data from Canada (YKA) is to be obtained in digital form through the post office dialled line services. The United Kingdom stations EKA and Wolverton are near enough to the data centre to afford direct private lines.

"Documents on Disarmament, 1972, pp. 557–583.