Non-uniqueness in surface-wave inversion and consequences on seismic site response analyses

Guidelines for the good practice of surface wave analysis: a product of the InterPACIFIC project

“…Right corresponding 1D soil response. The differences between acceleration response spectra at the surface are limited (Foti et al 2009) A key parameter for ground response analyses is material damping, which is typically introduced on the basis of laboratory tests. However they may be not sufficiently representative of the response of the soil deposit.…”

Comment on “Effect of surface wave inversion non-uniqueness on 1D seismic ground response analysis” by Roy et al.

“…They refer to a previous study (Foti et al, 2009) in which it was shown that the impact of solution non-uniqueness on seismic response simulations is negligible for profile having high impedance contrast. They also refer to another study (Boaga et al, 2011) in which it was reported that, in the case of a gradual velocity increase with depth, solution non-uniqueness deeply affects the accuracy of seismic response analyses: for low impedance contrast the effect is much more pronounced than for high impedance contrast, for which the equivalent solutions have a very little influence.…”

“…We also address the numerical simulation of seismic site response. We apply a consistent framework to one real record of the same earthquake to show that, contrary to what is claimed in the considered study, the solution nonuniqueness has negligible effect in amplification and acceleration response spectra.Comment Roy et al (2013) study the impact of solution non-uniqueness of surface wave inversion on seismic site response of soil column using near-source and far-source earthquake records.They refer to a previous study (Foti et al, 2009) in which it was shown that the impact of solution non-uniqueness on seismic response simulations is negligible for profile having high impedance contrast. They also refer to another study (Boaga et al, 2011) in which it was reported that, in the case of a gradual velocity increase with depth, solution non-uniqueness deeply affects the accuracy of seismic response analyses: for low impedance contrast the effect is much more pronounced than for high impedance contrast, for which the equivalent solutions have a very little influence.…”

“…In real cases, if the investigation depth of surface waves does not reach the seismic bedrock, this information should be introduced into the process as an a priori information inferred on the basis of other surveys (e.g. seismic reflection/refraction) as done for instance by Foti et al, 2009. Applying this procedure we implemented a 1D equivalent linear elastic site response analysis, carried out using the software EERA (Bardet et al, 2000), considering the same 15 best fitting profiles adopted by Roy et al and the same earthquake occurred on 2012/03/27: the selected time history is the surface EW record of the Tarou station at epicentral distance of 37 km. Input motion was obtained using a deconvolution procedure at bedrock outcrop beneath the Tarou station considering the Vs profile reported in Figure 2 (details in K-NET and KiK-net database) and with modulus reduction and damping curves taken from the literature (Seed and Idriss, 1984).…”

Characterization of site conditions (soil class, VS30, velocity profiles) for 33 stations from the French permanent accelerometric network (RAP) using surface-wave methods

“…They attribute this to surface wave depth resolution issues, non-uniqueness of the inversion procedures and/or the energy content of the signals used. The consequences of the uncertainties related to nonuniqueness of site responses have been a matter of discussion in recent years Foti et al 2009;Socco et al 2012;Boaga et al 2013;Jakka et al 2014;Comina and Foti 2015;Griffiths et al 2016;Cox and Teague 2016), with some authors claiming that the great uncertainty in Vs profiles leads to high variability of the site responses, although others do not agree. However, considering that the bedrock velocity is well known (e.g., through large arrays or other indirect information) and the velocity profiles are coherent with other indicators (e.g., the ambient vibration H/V frequency peak), the impact of nonuniqueness on site responses remains acceptable.…”