Comment on "How Can Seismic Hazard in the New Madrid Seismic Zone Be Similar to That in California?" by Arthur Frankel

Abstract

Frankel (2004) argued that seismic hazard, defined as the maximum shaking predicted for sufficiently low probabilities of recurrence, or equivalently over sufficiently long time intervals, is comparable for sites in the New Madrid zone to that for sites in California. Many aspects of this argument are debatable, because it relies on assumptions about the size, recurrence, and shaking from future earthquakes, none of which is well known (Newman et al. , 2001), and then makes inferences about the extreme and hence uncertain “tails” of the probability distribution presumed to characterize future ground motion (Wang and Ormsbee, 2005). Even more seriously, as discussed here, it relies on a definition of seismic hazard that is of little use in formulating public policy such as building codes because it considers the maximum shaking at a geographic point over 500 or 2,500 years (10% or 2% probability in 50 years) rather than over the much shorter (50-100 year) life of typical structures. Estimating seismic hazard involves both making assumptions about future earthquakes and choosing a definition of hazard. The latter is even more arbitrary and at least as significant as the former. Hence, even if the probabilities of future earthquake occurrence and ground motion were well known—which is far from the case—different …