Site condition evaluation using horizontal-to-vertical response spectral ratios of earthquakes in the NGA-West 2 and Japanese databases Hadi Ghofrani n , Gail M. Atkinson Department of Earth Sciences, Western University, London, Ontario, Canada N6A 5B7 article info Article history: Received 27 September 2013 Received in revised form 13 August 2014 Accepted 31 August 2014 Keywords: Horizontal-to-vertical (H/V) spectral ratio The time-averaged shear wave velocity to a depth of 30 m (V S30 ) Site amplification NGA-West 2 database Standard H/V response spectral ratio Ground motion Site classification abstract We evaluate the usefulness of the horizontal-to-vertical (H/V) response spectral ratio as a parameter to describe site response. In particular we compare its effectiveness with that of the time-averaged shear- wave velocity to 30 m (V S30 ), and also look at how those two measures are correlated. The evaluation is performed considering two major compiled ground-motion databases, one being the international NGA- West 2 database, and the other being a comprehensive database from Japan. A uniform procedure is applied to pick peak frequency (f peak ) and peak amplitude (A peak ) from the averaged H/V response spectral ratio for each site in the database. The H/V peak parameters are then grouped by their behavior, and the relationship of these peak parameters to V S30 are investigated by region. We conclude that: (1) H/V offers at least as much site information as V S30 ; (2) H/V is of more descriptive value than V S30 for deep soil sites, having f peak r1 Hz; and (3) the averaged H/V response spectral ratio for a site peaks at a specific frequency which is related to the depth of deposit, and has a stable peak amplitude of 0.45 log10 units, independent of the region. We conclude that H/V has significant advantages over V S30 as a site description variable, and note that it is in general easier (less expensive) to obtain. & 2014 Elsevier Ltd. All rights reserved. 1. Introduction It has long been recognized that site amplification is a key factor in determining the degree of damage caused by earth- quakes, and thus it has been the subject of much scientific research, both empirical and theoretical (e.g., [50,13,5,8,9]). Among several methods introduced for estimating site response, one is that provided by the observed horizontal-to-vertical com- ponent ratio (H/V) of ground–motion data. Most typically, H and V are the Fourier amplitude spectra of the horizontal and vertical components of ground-motion, respectively. A variant of the spectral ratio method, employed in this study, utilizes the earth- quake's response spectrum (pseudo-spectral acceleration [PSA] or pseudo-velocity). These ratios are much smoother functions of frequency, due to the narrowband filtering effect of the single degree of freedom damped oscillator [48]; throughout the manu- script, by the H/V ratio we refer to the response spectral ratio of horizontal to vertical components. This may be considered as an approximation for the corresponding smoothed Fourier spectral ratios. The H/V ratio is believed to reflect the total amplification effects at a site, which include amplification due to the crustal velocity gradient and any resonant amplification caused by the near-surface soil layer [42,39,43,7]. The underlying hypothesis is drawn from an extension of Nakamura's [42] technique. For a soft layer overlying a half-space, the incident horizontal-component amplitudes will be amplified by the soft layer. There may be amplification and attenuation of the vertical component also, but for the vertical component any ampli- fication effects are counterbalanced by the effects of refracting the ray path toward the vertical. It is reasonable to postulate that the net amplification effects for the vertical component are relatively small in comparison to the horizontal component, and can thus be neglected. Consequently the H/V ratio is a rough measure of the amplification of the horizontal component due to the soft layer. The H/V ratio on soft soil sites generally exhibits a clear, stable peak that is correlated with the fundamental resonant frequency [46,40,25,37,27,18,21,47]. The site fundamental frequency can be estimated using either ambient noise measurements or H/V spectral ratios obtained from earthquake data e.g., [8]. However, it is known that H/V typically underestimates the amplitude of the site amplifi- cation [25,8]. The H/V method is attractive from a practical point of view, because it is a single-station estimation which can be obtained cheaply and easily from a micro-tremor measurement at the site. For seismographic sites, it has the advantage of offering site information without the need for additional site surveys. Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/soildyn Soil Dynamics and Earthquake Engineering http://dx.doi.org/10.1016/j.soildyn.2014.08.015 0267-7261/& 2014 Elsevier Ltd. All rights reserved. n Corresponding author. Tel.: þ519 661 2111x84715; fax: þ519 661 3198. E-mail address: hghofra@uwo.ca (H. Ghofrani). Soil Dynamics and Earthquake Engineering 67 (2014) 30–43