International Handbook of Earthquake & Engineering Seismology, Part A

William H.K. Lee,Paul Jennings,Carl Kisslinger,Hiroo Kanamori

• Publisher: Academic Press
• Year: 2002
• ISBN:
  9780080489223
• Topic: Physical Sciences
• Subtopic: Seismology & Volcanism

Modern scientific investigations of earthquakes began in the 1880s, and the International Association of Seismology was organized in 1901 to promote collaboration of scientists and engineers in studying earthquakes. The International Handbook of Earthquake and Engineering Seismology, under the auspices of the International Association of Seismology and Physics of the Earth's Interior (IASPEI), was prepared by leading experts under a distinguished international advisory board and team of editors.The content is organized into 56 chapters and includes over 430 figures, 24 of which are in color.

This large-format, comprehensive reference summarizes well-established facts, reviews relevant theories, surveys useful methods and techniques, and documents and archives basic seismic data. It will be the authoritative reference for scientists and engineers and a quick and handy reference for seismologists.Also available is The International Handbook of Earthquake and Engineering Seismology, Part B.

1

History of Seismology

Duncan Carr Agnew, University of California at San Diego, La Jolla, California, USA

1 Introduction

At present seismology is the study of seismic sources (mostly earthquakes), the waves they produce, and the properties of the media through which these waves travel. In its modern form the subject is just over 100 years old, but attempts to understand earthquakes go back to the beginnings of science. The course of seismology, more than that of many other sciences, has been affected by its object of study: From Lisbon in 1755 through Kobe in 1995, destructive earthquakes have provoked scientific interest, and, quite often, social support for seismic studies. Table 1 lists some earthquakes (and one explosion) that have had an impact on seismology.

TABLE 1

Some Events of Significance to the History of Seismology

This article describes the history of seismology up to about 1960, with a brief sketch of major themes since then. To cover this history in the space available requires a fair amount of selection. Any reading of the older literature shows that a great many ideas were suggested long before they became generally accepted: For example, the ideas that shaking is a wave propagated from a source, that some earthquakes (at least) are caused by faulting, and that the Earth contains a liquid core. I have therefore focused less on the earliest occurrence of an idea than the time when it became something seriously considered within the seismological community. To make the narrative more readable, the sources for particular statements are given in a separate set of notes, referenced through footnote numbers; these notes, the reference list, and a bibliography of the history of seismology, are all included on the attached Handbook CD.¹

2 Early Ideas about Earthquakes

The most common explanation given for earthquakes in early cultures was the same as for any other natural disaster: they were a manifestation of divine wrath. In European thought this idea did not disappear from scientific discussion until well into the 18th century. But two premodern cultures, the Chinese and the Greek, also developed naturalistic explanations for seismic shaking. Greek natural philosophers suggested a variety of causes for earthquakes; the most influential (and extensive) treatment extant was by Aristotle (ca. 330 BCE), who attributed earthquakes to winds (driven by an “exhalation,” the pneuma) blowing in underground caverns. The classical authors also attempted to classify earthquakes by different types of shaking: something that remained a mainstay of seismology for a long time, and still survives in popular terminology.²

Chinese ideas about earthquakes were put forth by various thinkers at roughly the same time as the Greek ones, the dominant idea also being that shaking was caused by the blocking of a subtle essence (the qi). After the Han dynasty (200 BCE), Imperial state orthodoxy associated natural disasters with dynastic decline; this led to systematic preservation of accounts of earthquakes in the official annals. China’s technology also produced the first seismoscope, invented in AD 132 by Zhang Heng. This well-known device, which reportedly signaled the direction of an earthquake as well as the occurrence of shaking, is supposed on at least one occasion to have responded to unfelt shaking; the mechanism of its operation remains obscure.³

The Aristotelian view of earthquakes (as of many other aspects of the world) remained the primary theory during the medieval periods of both Islam and Europe. With the decline of Aristotelian thought in early modern Europe, other ideas were put forward, though many of the writers were from northern Europe and so (unlike the Greeks) had little direct experience of earthquakes. They did, however, know about gunpowder: This new technology of chemical explosives suggested that earthquakes might be explosions in the Earth (or in the air); of the various chemical theories put forward, the most popular involved the combustion of pyrites or a reaction of iron with sulfur. Such theories also explained volcanic action; that the most seismic part of Europe, namely Italy, was also volcanic helped to support this association. In the 18th century the development of theories of electricity, and especially their application to lightning, provoked several theories that related earthquakes to electrical discharges. In England, much of this theorizing was stimulated by the occurrence of several damaging earthquakes in the year 1750.

However, the greatest stimulus to seismological thinking was undoubtedly the Lisbon earthquake of 1755, partly for its destructiveness, but even more for providing evidence of motion at large distances: It caused seiches over much of Europe. At least two writers, J. Michell (1761) and (much more obscurely) J. Drijhout (1765), proposed that this distant motion was caused by a wave propagating from a specific location, thus more clearly than before separating the earthquake source from the effects it produced. The type of wave envisaged was like a traveling wrinkle in a carpet; Michell also suggested that the vibrations close to the source were related to waves propagated through the elasticity of the rocks, as sound waves were known to propagate through the elasticity of the air. The elasticity and pressure of gases, more specifically of high-temperature steam, also provided Michell’s driving force for the earthquake itself, which he took to be caused by water vaporized by sudden contact with underground fires. (This steam also supported the propagation of waves to great distances.) Michell attempted to locate this “place of origin” by comparing the times of the seicheinducing wave and the observed sea wave, and also hazarded a guess at the depth. While these ideas were not forgotten, they did not lead to any additional research, and certainly did not replace older theories of earthquakes.⁴

3 The Nineteenth Century to 1880

The expansion and professionalization of science in the 19th century meant that earthquake studies, like many other parts of science, could become a specialization, at least for a few scientists for part of their careers. One type of research that began in this period was the accumulation of large volumes of data, with the aim of finding underlying patterns—a style that has been called “Humboldtean” when applied to the Earth, but was in fact much more general. For earthquake studies, this meant the first systematic catalogs of shocks (as opposed to lists of catastrophes); leaders in this were K.A. Von Hoff and A. Perrey, the latter being a disciple of A. Quetelet, one of the founders of statistics. Many of these compilations were used to look for possible correlations between earthquake occurrence and astronomical cycles or meteorological events; this was one of the main topics of 19th-century seismology. Along with these catalogs came studies of individual shocks: Most European earthquakes after about 1820 stimulated some sort of special study, by individuals (often local professors) or by commissions set up by governments or local scientific societies. The first such commission was established after the Calabrian earthquake of 1783; a century later, one earthquake (in Andalusia, 25 December 1884) would bring forth three commissions, one each from Spain, France, and Italy.

These special studies developed many of the tools and vocabulary still used to describe the felt effects of a large earthquake. One such tool, quite in keeping with the overall trend in science toward quantification, was scales of intensity of shaking: the first by P. Egen in 1828, followed by many others, notably those of M. de Rossi, F. Forel, these two working together, and G. Mercalli. The first cartographic application of an intensity scale, creating the isoseismal map, was by J. Nöggerath in 1847; in turn, the accumulation of maps stimulated questions about why the distribution of shaking was as observed, and to what extent it could be explained by waves radiating from a central source.

This period also saw the first efforts to relate earthquakes to other geological processes. Von Hoff was explicitly interested in this relationship and in the English-speaking world it was promoted most assiduously by C. Lyell, whose program of reducing all past geological change to current causes was aided by showing that earthquakes could cause vertical motions over large areas. Prominent examples of such motion in Lyell’s treatment were the 1819 Rann of Cutch (India) and 1822 Chilean earthquakes, and later the 1835 Chilean and 1855 Wairarapa (New Zealand) earthquakes. The 1819 and 1855 earthquakes produced some of the earliest known examples of a break at the surface, though the first scientific observations of fault rupture did not take place until much later, by A. McKay in 1888 (North Canterbury) and by B. Koto in 1891 (Nobi).⁵

In retrospect, the basis for a different approach to earthquake study can be seen to have begun in 1829 to 1831, with investigations by S.D. Poisson into the behavior of elastic materials. He found that in such materials wave motions could occur, and were propagated at two speeds; the slower wave had particle motion perpendicular to the direction of propagation, and the faster one included dilatation of the material. A wave motion with transverse vibrations was of great interest because it offered a model for the recently discovered polarization of light, and many of the 19th-century investigations into elastic wave propagation were in fact made with optical observations in mind, attempting to explain light as transverse waves in an elastic “luminiferous ether.” Notable examples were the study of G. Green (1838) into wave transmission across a boundary, and of G.G. Stokes (1850) on radiation from a limited source.⁶

These results were applied to earthquake studies by W. Hopkins (1847), and by R. Mallet (1848 onwards). Hopkins showed how timed observations of wave arrivals could be used to locate an earthquake, but went no further than this purely theoretical exercise. Mallet, a polymathic engineer, not only coined the term seismology but tried to develop it systematically as a science of earthquakes, observed through the waves they generate. Mallet constructed one of the most complete earthquake catalogs to date, which he summarized in a map (Color Plate 1) that clearly delineates the seismic and aseismic regions of the world (1858). But Mallet aimed to do more than describe: Whenever possible he argued for the application of quantitative mechanical principles to determine how much, and in what direction, the ground moved in an earthquake. This quantitative emphasis is perhaps most notable in his 1862 study of the 1857 Basilicata (Neapolitan) earthquake, in which he attempted to estimate the direction of arrival of the shaking at many points, and so infer the location (and depth) of the source. It is also apparent in his earlier attempts (1851) to measure actual wave velocities from explosions and compare these with known elastic constants; he obtained much lower values than expected, which he attributed to inhomogeneity but which were more likely caused by insensitive instruments. Like Michell, Mallet believed that earthquakes were caused by the sudden expansion of steam as water met hot rock; because of the explosive nature of such a source, he believed that the earthquake waves would be almost entirely compressional.⁷

What was lacking in Mallet’s otherwise comprehensive program was an adequate method of recording earthquake motion; though he and others proposed possible ways to do this, few of these schemes were actually built and even fewer were used by more than the inventor—though it was recognized early (for example, by Hopkins) that a network of instruments was really what was needed. The first instrument to automatically record the time and some aspects of the shaking was the seismoscope of L. Palmieri (1856), used in Italy and Japan. However, the first network of instruments, set up in Italy starting in 1873, was not intended to record earthquake shaking. Rather, these “tromometers,” developed by T. Bertelli and M. de Rossi, were used to measure ongoing unfelt small motions, looking for changes in the amplitude or period of these related either to weather or to earthquakes, a study called “endogenous meteorology.”⁸

4 The Birth of the “New Seismology”: 1880–1920

With its emphasis on background vibrations, the substantial Italian effort turned out to be less fruitful than what happened in Japan as a consequence of the Meiji restoration of 1868: the establishment of modern science in a very seismic region. This was begun by the Japanese government bringing foreign experts (yatoi) to Japan, including professors well trained in the latest methods of physics and engineering. Of these, the most important for seismology was John Milne, who arrived in Japan in 1876 to be a professor (aged 26) at the Imperial College of Engineering in Tokyo. He made seismology his main interest after the earthquake of 22 February 1880, which also led to the foundation of the Seismological Society of Japan, with Milne as its effective leader. Such organization among the foreign experts was soon paralleled by similar initiatives from the Japanese: the Meteorological Agency established (or rather, took over from Milne) a regular reporting system in 1883, and S. Sekiya became the world’s first professor of seismology in 1886. The routine reporting of earthquakes allowed Sekiya’s successor, F. Omori, to develop his law for the decay of aftershocks from data for the 1891 earthquake.

Even before the 1880 earthquake, attempts had been made by foreign scientists in Japan to record the time history of felt shaking. This developed into a rivalry between J.A. Ewing at the University of Tokyo, and Milne (with his colleague T. Gray). Ewing applied the horizontal pendulum to get the first good records of ground shaking (at what would now be regarded as the lower limit of strong motion) in 1880–1881. These records showed the motion to be much smaller than had been assumed, and also much more complicated: nothing like a few simple pulses and not the purely longitudinal motion envisaged by Mallet. Thus, as soon as seismologists had records of ground motion, they faced the problem that has been central to the science ever since: Explaining observed ground motion and deciding how much of the observed complication comes from the earthquake and how much from the complexities of wave propagation in the Earth.

While Milne may not have been the first to record earthquake shaking, he soon became a leading seismologist, not so much from any new ideas he brought to the subject as from his energy and flair for organization. Like Mallet, Milne aimed to study all aspects of earthquakes and elastic waves, but he added to Mallet’s quantitative emphasis a regular use of instrumental meas...