How big, how bad, how often: are extreme events accounted for in modern seismic hazard analyses?Ivan Wong The 2004 moment magnitude (M) 9.2 Sumatra and 2011 M 9.0 Tohoku, Japan earthquakes can be regarded as “extreme” events because in the past 200 years, less than 10 earthquakes have reached M 9.0 or greater. Both earthquakes can also be regarded as extreme because of the devastating loss of life. We can also consider the 2010 M 7.0 Haiti earthquake as an extreme event with possibly more than 10,000 deaths (possibly much more) even though it was of moderate size. Thus earthquakes can be extreme either because of the resulting hazard (e.g., ground shaking, tsunami) or the resulting losses or both. It is important to distinguish between hazard and losses when defining extreme events. So the question above must be asked in the context of hazard and risk analyses. Except for possibly extreme magnitudes, we know, by and large, where there will be extreme losses on a global scale: in earthquake-prone areas where there is a large concentrated population that is vulnerable because of poorly engineered buildings. We were aware of the situation in Haiti. In terms of predicting seismic hazard, was the 2011 earthquake considered in analyses in Japan? Apparently not but in retrospect, it should have been. Although there was no well documented case of a previous M 9 earthquake occurring along the northern Japan subduction zone, the 869 Jogan earthquake was suggested in 2005 to have been a predecessor of the 2011 event. Why wasn't this possibility considered in probabilistic seismic hazard analyses (PSHA) for the Fukishima nuclear power plant? In the U.S., guidelines have been established for the proper performance of PSHAs for critical facilities. They stress the importance of (1) proper and full incorporation of uncertainties and (2) inclusion of the range of diverse technical interpretations that are supported by data. I would suggest that the 2011 earthquake would have been considered in PSHAs in Japan if the philosophy behind these guidelines had been followed. However, what if no evidence of the 869 earthquake had become available? It might still be argued that a M 9 earthquake should have been addressed by looking at other subduction zones worldwide where apparently “segmented” subduction zones had multiple segments rupturing together in a larger than anticipated earthquake, such as the 2004 Sumatra earthquake. The uncertainties in our knowledge of earthquake behavior need to be adequately included in PSHAs. However, even then there is no guarantee that all extreme events will be recognized; there are always unanticipated surprises. We must recognize that the results of even the best PSHAs have a limited “guarantee”. Although stability is sought in hazard predictions, the record suggests that hazard estimates may only be stable for a decade at best. |
