Comment on “Interpretation of Kappa and fmax filters as source effect”, by Igor A. Beresnev

Beresnev (2019) advocates the use of an earthquake slip function that produces an ω-2.5 high-frequency falloff of Fourier displacement spectra in the far field, where ω denotes the angular frequency. He argues that the observed high-frequency decay of earthquake spectra can be adequately modeled by this ω-2.5 falloff, without needing to include high frequency attenuation at shallow depth under the site, commonly characterized as fmax or kappa. In his abstract, Beresnev (2019) describes source models with falloffs intermediate between ω-2 and ω-3 as “providing natural high-cut filtering exclusively as a source effect.” In many studies to date, observed spectra are modeled using an ω-2 source spectrum combined with attenuation along the propagation path, including strong attenuation at shallow depths (< 1 km) beneath a site. It is not clear whether Beresnev (2019) is claiming that path effects (including site attenuation) are unimportant to ground motions or if he is proposing a simple, pragmatic method to fit the high-frequency decay of earthquake spectra.