Bulletin of the Seismological Society of America, 96, 2, 754-756, 2006
V. Convertito and A. Herrero
We thank F. O. Strasser, V. Montaldo, J. Douglas, and J. J. Bommer for the interest they have shown in our article (Convertito and Herrero, 2004). Strasser et al. (2006) present a critical comment of our work arguing that the solution proposed by Bommer et al. (2003) is a better solution. Note that the authors are nearly the same in both article and comment, except for V. Montaldo. Because this brief article is a reply, we will focus on the arguments directly concerning our article. The main objection supported by Strasser et al. (2006) is that the method we proposed is not appropriate to "styleof-faulting" correction. We completely agree with this assertion because it is simply not the scope of our article. We speak about "focal mechanism" intended as radiation pattern and nothing else. This point is clearly stated in the introduction of Convertito and Herrero (2004): "in this article we consider that the focal mechanism influence is only expressed by radiation pattern changes. In particular we do not consider any tectonic influence, stress drop variation or dynamic effects." The style-of-faulting parameter, even if its identity is blurred (e.g., Bommer et al., 2003), is an empirical definition of a complex set of physical conditions including the tectonic regime, the medium behavior, rock mechanics, rupture dynamics, and so on. In our opinion, the style of faulting is simply too complex to be used directly in our approach. Because the scope of our article is to show how it is possible to insert inside the main equation of probabilistic seismic-hazard analysis (PSHA; e.g., Cornell, 1968), simple physical parameters of the seismic source, that is, how it is possible to integrate deterministic parameters inside a probabilistic approach, we have chosen a small target, limiting ourselves only to the radiation pattern. We believe that the same approach can be used to insert many other parameters of the seismic source inside PSHA by using only a theoretical approach such as the fault strike (which has already been shown by Convertito, 2004), the directivity and stress drop. The second important argumentation is that a method based on regression (i.e., Bommer et al., 2003) is better than the method we propose. Once again we agree with Strasser et al. (2006) and this is clearly stated in the conclusion of our article: "when an attenuation law, including a faulting style parameter, is available for a given region, the use of this attenuation law gives a more reliable estimate of the hazard than the one obtained using the corrective coefficient we propose in this article."