- © 2012
It might seem odd to be writing about confirmation of the validity of probabilistic seismic hazard assessment (PSHA) in 2011, given that the method has been successfully applied in countless studies worldwide over the last 40 years. However, the fact that papers still occasionally find their way into print attacking the method as mathematically invalid seems to indicate that there is still some requirement, if small, to demonstrate the soundness of the method. A number of mathematical arguments have been advanced over the last few years purporting to show mathematical or logical flaws in the standard PSHA methodology that invalidate the results. A comprehensive summary of these objections can be found in Klügel (2008).
It is not the purpose of this short paper to enumerate these claims or analyze them in detail. Rather, the intention is to point out that the results of a PSHA study using the Cornell-McGuire method can be duplicated by a completely different route, and without any but the simplest mathematics, using a quasi-observational approach. The fact that two completely different approaches to computing the answer to the question “What is the probability of ground motion Y at a given site?” yield the same answer suggests the validity of both methods. If the basis of conventional PSHA were flawed, the results could not agree with the results of another study that shared none of the same procedures.
It seems that many seismic hazard practitioners are unaware of this “refutation of refutations”; hence the present paper may be timely. It used to be received opinion that computing hazard in different ways necessarily led to different results (e.g., Makropoulos 1993), which would certainly raise questions as to which results were “right.” The statement “based on these assumptions, the annual probability of a PGA of 0.2 g …