Earthquakes - An Essay by Geoffrey King, Ph.D.

n our Special Topics analysis on earthquake research over the past decade, the scientist who ranks at #3 is Geoffrey King, with 17 papers cited a total of 565 times. Dr. King is a co-author on the most-cited paper in our analysis, "Static stress changes and the triggering of earthquakes," (Bull. Seismol. Soc. Amer. 84[3]: 935-53, June 1994). His record in the broader ISI Essential Science Indicators Web product includes 24 papers cited a total of 757 times to date in the field of Geosciences. Dr. King is the Director of the Laboratoire Tectonique at the Centre Nationale de la Recherche Scientifique’s Institut de Physique du Globe de Paris. In the essay below, he discusses his highly cited work.

To be a major contribution, scientific papers must score well in both "believability" and "correctness." The history of science is littered with papers that were correct, and not believed at the time, and the reverse. Thus some modesty is appropriate when one finds one’s papers are highly cited. The papers I have written with various co-authors on stress interactions between earthquakes have met with approval because most people wanted to believe them. The concept that stress changes due to one earthquake must influence future events has seemed seductive for many decades, but attempts to clearly demonstrate an effect failed. Three factors allowed myself and co-workers Ross Stein and Jian Lin to start to change this. The Landers earthquake in Southern California in 1992 was geometrically simple, and data on the mechanism of the main event and aftershocks were readily available. We recognised small, though important, errors in the way that previous workers had approached stress interaction in the past—in particular, the relation between the prevailing stress field in a region and stress changes due to each earthquake. We also had a computer code that could be immediately modified to examine the Landers earthquake data. Finally we produced colour figures that everybody liked. From this first paper a series of others followed and have continued to attract attention. The early successes were to show how the stress distribution due to a large earthquake controlled the distribution of the subsequent aftershock sequence. This was then extended to show that big events interact in a similar way. More recently, with David Bowman, I have shown that small earthquakes, prior to a major event, occur in the regions where stresses can be shown to have been high. This increase in activity occurs at a distance from the future earthquake fault—a feature that mystified many workers. However, we have shown that this effect is a simple consequence of the traditional elastic rebound model of earthquakes.

“Although the concepts of earthquake interactions are now recognised as correct, it is also clear that much remains to be understood.”

Although the concepts of earthquake interactions are now recognised as correct, it is also clear that much remains to be understood. Very small stress changes seem to affect the location and timing of future earthquakes. Some researchers have found it hard to believe that such small changes can have such a major effect and seek reasons why these small effects may be amplified. The delay between a stress change and the timing of a future event also remains poorly understood. The role of stresses associated with propagating waves has also become a central topic of research into stress interactions. Numerous authors, including myself and co-workers, have attempted to address these problems.

While my work has excited basic research, it has important social implications. A study prior to the destructive 1999 Izmit earthquake in Turkey showed that the region was at risk; a sadly satisfying triumph for stress modelling. Subsequent work indicates Istanbul to be in great danger and a catastrophic event is likely to end the lives of a significant fraction of its current inhabitants. If we now consider the location of mega-cities in earthquake-prone zones around the world, it seems that a million or more people will be killed by earthquakes in the next 50 years. Stress interaction techniques can allow us to define danger areas more closely.

Although my most-cited work is computer modelling, most of my effort is actually expended in data collection or encouraging data collection. Mathematical and computer models of earthquake processes are only as good as the data available to justify them. In response to a paper describing years of careful fault mapping, a reviewer remarked that "…data rarely if ever distinguishes between models." A Ph.D. from a top university, he was unwilling to accept that our data rendered his models improbable. Data is in short supply and is badly needed. In recent years, I have been fortunate to work with Paul Tapponnier and Rolando Armijo mapping active faults and determining their short- and long-term slip rates. The aim is to understand the basic mechanical processes of the deformation of continents. At the same time, this work, which occupies the greater part of my time, also provides the basic information for stress-interaction studies and establishing earthquake risk.

It is my hope that the next 10 years will see a growth in the number of young researchers working in this field. Furthermore that the necessary funding is available for field research and to sustain careers that necessarily publish less frequently and are cited less often. Such an effort would tell us much about the Earth and could help to save many more lives than are at risk from terrorism. Sadly, I doubt if it will happen.

Geoffrey King, Ph.D.
Institut de Physique du Globe de Paris
CNRS
Paris, France

Read an interview with Jian Lin, discussing the Special Topic of Earthquakes in ESI Special Topics.

Read an interview with Ross Stein, discussing the Special Topic of Earthquakes in ESI Special Topics.

ESI Special Topics, January 2004
Citing URL - http://www.esi-topics.com/earthquakes/interviews/GeoffreyKing.html