Comment on "Statistical Features of Short-Period and Long-Period Near-Source Ground Motions" by Masumi Yamada, Anna H. Olsen, and Thomas H. Heaton

Abstract

Yamada et al. (2009) collected recorded ground motions from the near-source region of large earthquakes that occurred before 2005 and investigated to what extent this historical record could provide a basis for estimating future ground motions at similar sites. Their main conclusion, which they refer to as a paradox, is that the largest peak ground acceleration (PGA) observed in the past can be considered representative of future ground motions with the largest PGA (even though at present earth scientists cannot reliably simulate PGA values by means of numerical models). In contrast to this, the largest peak ground displacement (PGD) observed in the past may not be representative of the future largest PGD (even though this is easier to estimate through numerical simulations). The previous statement, supported by Yamada et al. (2009) with theoretical considerations based on the physics of earthquake sources, stems from the widely recognized weak (if any) dependence of PGA on magnitude and from the still-debated saturation of PGA in the near field of large earthquakes. This is generally mirrored into the functional forms of many state-of-the-art ground-motion-prediction equations (GMPEs), see e.g., Abrahamson et al. (2008) in the special issue of Earthquake Spectra devoted to the Next Generation Attenuation (NGA) Project. The same does not occur for PGD values that, being strongly correlated to magnitude, would tend to follow a uniform probability distribution for ![Graphic][1] , so that, quoting Yamada et al. (2009), the estimation of the largest PGD from future earthquakes is not clear. We first remark here that a statistical analysis of near-source ground motions would have at least required (1) a homogeneous distribution of records, taking into account that about 2/3 of those used in Yamada et al. (2009) come from only two earthquakes and, as acknowledged by the authors themselves, some ranges of magnitudes are not represented at all; (2) a … [1]: /embed/inline-graphic-1.gif