Are There More or Fewer Earthquakes in the Future?

Recent reports on television, radio, and in newspapers in Los Angeles have dealt with a so-called "earthquake deficit" in southern California. This issue has its origins in a 1995 technical report published by the Southern California Earthquake Center and authored by a Working Group on California Earthquake Probabilities. We would like to explain the issue and where things stand at the present time.

First of all, what does the term "earthquake deficit" mean? Earthquakes are the result of plate tectonic motions, which have been going on at relatively constant rates for millions of years. Earthquakes occur at the margins of plates, where one plate rubs against another. In California, the North American plate is rubbing against the Pacific plate along the San Andreas system of faults (the system includes all active faults in California). The plates are constantly on the go, but they "hang up" along their margins, so stresses (forces) and strains (ground deformation) continue to increase at plate boundaries until they finally "snap" in an earthquake, much like what happens when you bend, and finally break a stick. The earthquake relieves the pent-up stresses and strains, and the cycle begins again.

Now stress and strain build-up is like charging your credit card, and your once-a-month-payment is like the earthquake. During the month your credit card account shows a deficit until you pay it off at the end of the month and then the cycle starts again the next month. So even for those who pay off their credit card charges every month, the account always has a deficit, but the deficit is dealt with on a regular basis and never gets out of hand. However, there are those who don't pay off their entire charges every month, and then the deficit gets larger and larger until they must make one very large payment or a series of smaller payments in more rapid succession.

Similarly for the earthquake process-- we are always living with an earthquake deficit. But just as in the credit card case, the problem arises when there is not a regular payment, or when the earth falls behind in the number of earthquakes necessary to relieve the plate tectonic strain on a regular basis. Then at some point in the future we must "pay the piper" with a really large quake or a series of moderate-sized events within a relatively short period of time. Incidentally, many scientists believe that earthquakes do not occur with great regularity, but can cluster in both time and space. Thus, a series of potentially destructive earthquakes can occur in a seismically prone region within a relatively short period of time-- for example, within several years to a few decades.

Based on a variety of data and assumptions about earthquakes in southern California, the Working Group on Southern California Probabilities developed, in 1995, a complex model that forecast future earthquake probabilities (over the next few decades) and ground motions from these earthquakes for all of southern California. The model also allowed us to estimate (as a byproduct) what the annual rate of occurrence of earthquakes of various magnitudes ought to be. And interestingly it was discovered that the model predicted an annual rate of moderate-sized quakes (M6 to M7 (M=Magnitude)) twice as great as had been observedover the last 150 years. Thus, it appeared as if earthquakes had not been occurring often enough over the last 150 years to release the strain being stored by plate tectonics, and an earthquake deficit was declared. This meant a payback sometime in the future -- or more earthquakes. Incidentally, we now prefer to refer to the difference between the model's prediction and what has been observed as a "discrepancy" rather than a "deficit", since as noted above, we are really always in a deficit state.

Of course, this discrepancy was a very provocative suggestion that encouraged a number of researchers to challenge it. A paper published by Mark Stirling and Steven Wesnousky in the Bulletin of the Seismological Society of America used the same data and a procedure similar to the Working Group on Earthquake Probabilities to forecast the total earthquake rate for southern California, with specific attention to the uncertainty bounds on the size and rate distribution of earthquakes expected from interpreting a wide range of data relating to past earthquakes. In contrast to the 1995 Working Group Report, this study concluded that the observed rate of historical earthquakes is statistically consistent with the rate predicted by the 1995 model (if uncertainties are accounted for).

Tom Hanks and Ross Stein of the U.S. Geological Survey posited at the 1998 annual meeting of the Seismological Society of America that we probably didn't have a complete record of earthquakes back to 1850, and it was not fair to compare the model's prediction with an incomplete record. They also questioned another aspect of the model, which had to do with the distribution of earthquake sizes. By considering only the historical record back to 1903 (which they believe is complete) plus the relatively well-known rate of occurrence of "big ones" on the San Andreas fault, they argue that the annual earthquake rate in southern California is keeping up with plate tectonics within the limits of uncertainty in the data. This means that seismicity in the future should not look much different from that of the past 100 years, or about one M6+ event every 3 years somewhere in southern California, one M7+ event about every 30 years, and a M7.5 to M7.9 event on the southern San Andreas about once every century or two (the last one was in 1857).

An issue that remained unresolved was how to incorporate geological fault slip information into a "fully integrated" model without producing the rate discrepancy. A report by Ned Field, James Dolan, and David Jackson, also at the 1998 annual meeting of the Seismological Society of America, provided a solution to this remaining problem. Specifically, they presented a model that is consistent with historical seismicity rates, does not require earthquakes larger than magnitude 8 or any other undocumented phenomenon, and is not dependent upon factors with high uncertainties. They also outlined ~6 equally influential factors that produced the apparent earthquake deficit in the previous study, one of which was the inadequate earthquake record identified by Stein and Hanks.

So given the new understanding, the "earthquake deficit" in southern California may not exist, and we will continue to have earthquakes at about the same rate we have seen in the last 100 years. But even though these recent refinements have concluded that the predicted rate of seismic activity is comparable to the historic rate in the southern half of the state, one local area where we seem to have an earthquake discrepancy is the greater Los Angeles Metropolitan area. The historic seismicity does not seem to account for the rate of strain build-up in the Los Angeles region. This local discrepancy may be explained by some combination of creep, large magnitude events (greater than M7), a future higher rate of seismic activity, and/or some other form of strain accommodation that we still do not understand. Therefore, although the explanation for this discrepancy is uncertain, it would be prudent to base public policy on the assumption that earthquakes considerably larger than Northridge can occur in the immediate vicinity of the metropolitan area.

The foregoing summary should be read as a scientific statement and not a statement of fact. Science works by hypothesis testing, and models are hypotheses that need to be tested. Newer models, that hopefully are better predictors of the future, take the place of older ones. And furthermore, the scientific method must not be misinterpreted to mean that scientists cannot get their act together. Rather it is this "give and take" and willingness of one scientist to challenge another's ideas that bring us closer and closer to the truth. And as scientists provide better knowledge about earthquakes in California, they contribute to better policy decisions on important issues, such as earthquake resistant design, emergency response, and earthquake insurance. What we can say with a high degree of certainty is that earthquakes will continue to be a way of life in southern California. Whether earthquakes will occur more or less frequently than in the past, we must face the fact that it only takes one event like the Northridge earthquake of 1994 to ruin our day. So the bottom line remains, "be prepared"!

Contacts:

Dr. Thomas Henyey, Director, Southern California Earthquake Center
(213) 740-0323 (email henyey@usc.edu)

References:

Jackson, D. D., K. Aki, C. A. Cornell, J. H. Dieterich, T. L. Henyey, M. Mahdyiar, D. Schwartz and S. N. Ward, Seismic Hazards in Southern California: Probable Earthquakes, 1994-2024, Bulletin of the Seismological Society of America, 85, no. 2, pp. 379-439, 1995.

Stirling, M. W. and S. G. Wesnousky, Do historical rates of seismicity in southern California require the occurrence of earthquake magnitudes greater than would be predicted from fault length? Bulletin of the Seismological Society of America, 87, pp. 1662-1666, 1997.

beginning of article