EARTHQUAKE CLUSTERS: GLOBAL CONSPIRACY OR RANDOM CHANCE?

When several newsworthy earthquakes occur within a month or so, seismologists are faced with the inevitable questions: Why do earthquake clusters occur? Is there some global linkage between these events? The earthquakes that occurred in the last months of 1999 are one example of an earthquake cluster. Of course, we do acknowledge that large earthquakes trigger aftershocks on the ruptured fault zone, and that one earthquake may trigger another event on a nearby fault due to static stress changes, or possibly even at remote locations due to seismic wave dynamic stresses. But most of us would say that global-scale clusters of large events--for example, the events that occurred in November 1999 in Turkey, the Indian Ocean, New Britain, and the Vanuatu Islands--are just coincidence. Random occurrence of independent events can produce clusters and gaps. How can we illustrate this behavior in a classroom?

There are some simple experiments that students can do to discover the clusters and gaps in a random process. Coin tossing always stimulates the students, but die rolling is also a good seismicity simulation activity. You can begin by stating that an earthquake with a magnitude of 6.5 or larger occurs somewhere in the world about every six days or so, on average. Then ask the question: I wonder if such an earthquake will occur today? Choose one of the faces of the die to represent a "yes" answer to this question; the other five faces will be "no." Then simply roll the die. It is not too tedious to roll the die 60 times, thus producing 10 earthquakes, on average. The students should keep track of the entire sequence so that they can go back and analyze their seismicity for clusters and gaps. If there are several student groups that produce seismicity sequences, then you can tabulate various results, e.g., the number of "doublets" ("yes" answers for two consecutive days) and the maximum duration gaps. At this point, the students will be eager to see some discussion on how to estimate the expected number of "doublets" in a 60-day sequence, or how to predict the distribution of maximum gaps from many different trials.

Please send to me via e-mail your ideas on variations and useful calculations that students can do. Of course, this exercise does not fully answer the question about global linkage of earthquakes, but at least the students will understand that most earthquake clusters can be explained as random chance. By directly producing the clusters and gaps as they roll the dice, students are more likely to actually remember the features of a random process. Furthermore, it is a fun activity and the students may decide that statistics really are a happening subject!

SRL encourages guest columnists to contribute to "EduQuakes." Please contact Larry Ruff with your ideas. His e-mail address is eduquakes@seismosoc.org.

Posted: 19 May 2000