Investigating the nature of microtremor H/V spectral ratio peak in south of Tehran

By growing the population and need for settlement, many cities have been built on soft sediments and seismic areas. It emphasizes the need for a careful and reliable assessment of site effect phenomena. Beyond the methods of studying site effects, microtremor recordings has become popular over the last decades as it offers a convenient, practical and low cost tool to be used in urbanized areas. Besides, in areas of low to moderate seismicity which gathering a significant number of recordings with satisfactory signal to noise ratio is a time-consuming task, microtermor studies are more useful.
Lack of accurate knowledge about the nature of microtremor wave field, would lead to misinterpretation of site effects, hence, investigating microtremor wave field is an important goal to achieve. Two techniques are predominantly used to determine microtremor wave field: the array techniques (such as SPAC and F-K methods) and the single station horizontal to vertical spectral ratio (H/V). Array studies have shown that surface waves dominate microtremor wave field, but the relative proportion of Rayleigh and love waves has still been unclear.
 In this study, single station horizontal to vertical spectral ratio method was applied to investigate the nature of microtremors in south of Tehran. Theoretical aspects of this method, has always been a considerable issue for researchers in this field. Regarding the dominance of fundamental Rayleigh wave mode on vertical component of microtremors, some researchers believe that, if impedance contrast between surface layers and the bedrock tend to be high, ellipticity curves (ellipticity at each frequency is defined as the ratio between the horizontal and vertical displacement eigenfunctions in the P-SV case, at the free surface) of fundamental Rayleigh wave mode shows a conspicuous peak around the site resonance frequency. It is due to the vanishing of vertical component corresponding to reversal rotation of fundamental Rayleigh wave from retrograde to prograde, In contrast, some other researchers do believe that the SH resonance in surficial layers (removing the effects of surface waves) accounts for H/V ratio peaks.
Data used in this study was recorded by Haghshenas et al. (2003) using continuous recording for a period of five months in 13 seismological stations. The results of two stations are shown here. Geopsy software (www.geopsy.org) is used to analyze microtremors.
First, particle motion behavior in microtremor wave filed was studied. The results showed an elliptical behavior that can be related to predominance of Rayleigh waves in microtremor wave field. It should be mentioned that if body waves dominate the wave filed, the particle motion will show a linear behavior which is not observed in our study. Then, spectrum amplitude curves were obtained. To compute H/V for each time window, root mean square of two horizontal amplitude spectra is divided by vertical amplitude spectra at each point. H/V curves showed that site resonance frequency varies from 0.3 to 5 Hz in south of Tehran. Our study revealed that the peaks at site resonance frequency were localized by minima in vertical amplitude spectra as well as by maxima in horizontal amplitude spectra. To study dispersion property of layers beneath each station, shear wave velocity variation with respect to depth was investigated. Dispersion curves were obtained based on earth models. It was shown that the under-structure layers are dispersive. Take into consideration that Tehran has a complicated geological state and lack of borehole information, Jica & Cest report (2000) was used to obtain earth models. Jica & Cest report (2000) includes only thickness and type of layers, while, the shear and longitudinal wave velocity is needed. Jica & Cest report (2000) contains Standard Penetration Test values according to the type and thickness of layers and experimental relations between these values and shear wave velocity. These relations were used to compute shear wave velocity under each station. Longitudinal velocities were computed by the relation proposed by Lay and Wallace (1995).
Finally, ellipticity curves of Rayleigh waves were modeled and compared with H/V curves. The same earth models of previous step were used to model ellipticity curves. The ellipticity curves showed a conspicuous peak around the site resonance frequency. These could be due to the reversal motion of fundamental Rayleigh wave mode from retrograde to prograde. To sum up, it could be said that in south of Tehran, fundamental mode of Rayleigh wave accounts for H/V ratio peak.