Value of historical seismicity
by
Nicholas N. Ambraseys
Department of Civil & Enviromental Engineering, Imperial College of Science, Technology & Medicine, London SW7 2BU

The purpose of this note is to describe how historical evidence can be used to address some fundamental questions: when and where have earthquakes happened in the past? How can accounts of ancient events contribute to our scientific understanding of earthquake activity? The aim here is to present a constructive statement of our views on a range of general points, rather than to engage in a detailed critique of individual cases.

As we cannot know what will happen in the future, to estimate likely earthquake hazards we have to find out what happened in the past and extrapolate from there. Previous research has uncovered evidence of destructive earthquakes in areas where only small events have been experienced recently . This is not surprising: the timescale of geology is vastly different from that of human history, so some areas will suffer a short period of violent earthquakes only once in a few hundred years. It follows that if we took account only of information about the last century, in which earthquakes have been recorded by instruments (and even then not uniformly throughout the globe), we would have no way of knowing whether an apparently "quiet" area is in fact at risk from a damaging earthquake. The use of the historical record is invaluable, not only in the study of earthquakes but also of the climate and weather, and can guide the engineer to design structures to resist the forces of nature without being taken by surprise by unanticipated events.

The reappaisal of the seismicity of the Eastern Mediterranean and the Middle East, east of the Adriatic and the Ionian Seas all the way to India, shows that although the historical record is incomplete, careful reading of the available data can provide valuable insights into the long-term seismicity of the region. The pattern of seismic activity of many areas is seen to have changed little over the past 2,500 years, while other areas which are at present quiescent can be shown to be capable of generating earthquakes of significant size.

Some of the lessons we believe we have learned from these studies are noted here.

· Archaeological evidence for an earthquake is not always unambiguous and can seldom be used to provide a precise date for the damage caused. Nevertheless, archaeological evidence can provide confirmation of long-term seismicity in a given region and with greater collaboration between disciplines it is likely that many refinements of the existing database will be possible.

· For the earth scientist and earthquake engineer the main objectives of historical research into primary sources are to refine and extend the information contained in secondary studies and catalogues, and to provide an objective measure of the reliability and completeness of the data retrieved.

· It is important to establish unambiguously the simultaneity of damage to different localities in an historical earthquake. Often one finds cases in which two separate events have been transformed into a large earthquake. This is understandable in view of the tendency of both contemporary and later writers to amalgamate seismic events, whether for lack of sufficiently precise information, from ignorance of the true nature of earthquakes, or from simple convenience. Such an amalgamation of effects will over-estimate the size of the damage area, and hence of the size of the event.

· While some of the problems of dating can be resolved, it is often more difficult to determine a sufficiently accurate location for historical earthquakes. The epicentral area of an historical event is not always certain and judgement has to be exercised to ascertain its location. The primary aim should be to avoid both the amalgamation and duplication of events.

· The size of an historical earthquake can be assessed in terms of its magnitude: such an assessment for historical events can be made only approximately and depends on the reliability of information regarding their effects at large epicentral distances or from the dimensions of their epicentral area. For events in which this information could be estimated, the magnitude of the event should be estimated using a calibration formula derived from 20th-century earthquakes for the region.

· In estimating intensities we find that at large distances an earthquake may cause the collapse of a few important but vulnerable constructions, for which there may be archaeological or historical evidence. This information alone should not always be taken to mean that all the other man-made structures at these sites have been destroyed. The observed effects can be the result of the high vulnerability of long-period structures to sustained ground motions, rather than of the severity of the shock.

· For many historical events, the data are wholly insufficient to permit assessment of intensity in terms of any of the scales currently in use, let alone to reckon the magnitude of an event, except in very general terms. We find that precise local or epicentral intensities assigned by modern cataloguers to many historical events, particularly in Greece and the Holy Land, are hypothetical and often grossly inflated.

· Earthquake catalogues are often used by earth scientists and engineers to assess earthquake hazard. A more critical attitude is needed to rely only on those that combine the interpretation of primary sources with estimates of the reliability and completeness of the data provided.

· The location and size of historical earthquakes should not be used for scientific purposes without proper scrutiny of the associated historical material.

· The historical record confirms that some regions that are active today (e.g. the north Anatolian fault zone) were also active 2,500 years ago, demonstrating the long-term nature of their seismicity. It also shows that some regions that are at present quiescent (such as the Jordan Rift Valley), are capable of generating relatively large earthquakes. For some of these events this is consistent with their known active tectonic environment.

· Too many modern catalogues of historical seismicity are not sufficiently rigorous to be treated with confidence. This has often been due to the inter-disciplinary nature of this field of study, which requires scientists to examine literary texts and historians to glean scientific information from their sources. The result has been the production of a large number of false earthquakes, or of seismic events of a size beyond the limits of the possible, often with a sensationalist tinge. This is of no technical consequence, provided the earth-scientist and engineer is aware of it.

Date received: June 24, 2002

Copyright © 2002 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caji-35.