After reading this introduction, go to the foreshock animation page. There you will find six images linked to animation sequences of seismicity. The view of each is a map view, looking down at the earth's surface from high above, with north at top. No reference locations are shown, so they do not bias your judgments. The scale is the same in every animation -- about 40 miles across (east to west) and about 50 miles tall (north to south). Each animation is 90 frames long, with each frame representing one day. Each frame shows the locations of the epicenters of earthquakes on each day in the 90-day sequence. Larger circles represent larger earthquakes, as shown by the colored magnitude scale on the left side of the images.
The idea behind this activity is to simulate the challenge of earthquake forecasting, using one particular method -- foreshock identification via a study of local seismicity over a period of several months. By studying the distribution of seismicity in an area for 90 days, you will attempt to forecast what will happen on day 91. For what should you be looking? Well, while there are no universally recognized patterns of seismicity seen preceeding mainshocks, one pattern that has been seen in several different parts of the world can be summed up generally by the following:
Seismicity increases over its normal rate in a particular area -- that is, you will see more earthquakes per day than usual. The seismicity in that area will then come to a stop, fairly quickly. A lull sets in, with about a month or two of relative seismic "quiet" (low rate of earthquakes). Then, just before the mainshock strikes (1 to 10 days, roughly), seismicity increases suddenly in the center of the area which showed increased seismicity before. A foreshock may precede that mainshock.
The cause of this pattern is thought to be related to a high level of stress along a large fault. This high stress causes smaller, weaker, and adjacent faults along the edge of the impending rupture zone to break, leading to an increase in seismicity in a particular area. Once all the minor faults have ruptured, the activity stops, and nothing but the primary fault remains unbroken. That fault, then, shoulders the stress entirely by itself, and after straining quietly for a short time, finally begins to give, eventually rupturing in a large earthquake.
Therefore, when watching the animations, try looking for an active area that eventually becomes "quiet" for a month or two. Then, immediately before the end of the animation, watch for renewed activity in the area which showed high activity early on. Or use your own method for deciding which quakes might be foreshocks. But beware! Not all of these animations lead to a large earthquake (mainshock) on day 91, and of those that do, not all have foreshocks that precede the mainshock.
Before you start, here's how to make your earthquake forecasts:
At the end of each animation you will have the chance to decide whether
or not you think the earthquakes that occurred on the last day (day 90)
of the animation are foreshocks of a large earthquake that will occur
on the following day. If you think that there will be a large earthquake
on day 91, then you'll want to issue an earthquake alert! Click "Quake
Alert!" to make that forecast. If you don't see anything unusual in the
seismicity prior to day 90, then don't issue an alert -- click the
"No Alert" button at the end of the animation. After making your choice,
you will be shown the actual seismicity for day 91. Don't feel bad if the
seismicity from day 91 doesn't conform to your predicition. It is not
uncommon for unusual seismicity patterns to occur without being followed by
a large earthquake. Conversely, it is also not uncommon for a large
earthquake to occur with no anomalous earthquake activity prior to it.
Go now to the foreshock animation page and see how you do as an earthquake forecaster!
After finishing with all the examples, come back to this page and consider the points below:
If you were uncertain about most of the examples, that's OK. This lesson was not designed to show how likely it is to guess correctly, but how difficult it is to be certain of your assessment before you are proven either correct or incorrect. For those of you who chose correctly on several examples, do you think this activity would have been much more difficult if the animations were of indefinite length, and you had to stop them on the frame before the mainshock? That's a fair analogy of an earthquake prediction.
to the Text