Comparison of SSR and H/V microtremor data analysis techniques

The destructions caused by many recent earthquakes shows that the subsurface structure directly affects the ground motions. Especially in areas with possible seismic hazard expectation, expansion of cities has come up with detailed vulnerability analysis. Therefore, site response investigation is a key step in every microzonation study. One of the best approaches for understanding the geological structure is to observe the seismic ground motions on the ground surface. Although mostly qualitative measures are presented, the relationship between the destruction caused by earthquakes and site effect has been proven by many researchers. So, it is widely known that the site response is the most important information needed in site evaluation. So, there is a great need to evaluate the dynamic site response characteristics such as resonant frequency and amplification factor efficiently and cost-effectively. These data are not only useful for earthquake engineers but geotechnical engineers and seismologists.
Nowadays, microtremors are being used in site effect and microzonation studies. Microtremors are the ubiquitous, weak, low amplitude vibrations which may be recorded on the surface of the earth. Microtremors are used as passive seismic methods which are cost effective, so t¬¬hey are being used increasingly. Studies by many scientists (e.g. Aki, 1957; Nakamura, 1989; Lachet and Bard, 1994; Lermo and Chavez-Garcia, 1994) about nature of microtremor sources and effect on ground identify the atmospheric disturbances and meteorological phenomena over the land or the sea as well as human activities as generators of microtremors. The relationship between local geological structure of surface layers and microtremor spectral characteristics has been proven by the above mentioned scientists. The most commonly used methods to quantify this relationship, are the well known Standard Spectral Ratio technique which requires the simultaneous measurement of local and remote reference data on bedrock (Borcherdt and Gibbs, 1976) and the more recent Horizontal to Vertical Spectral Ratio (Nakamura, 1989), each of which has its own advantages and disadvantages. A difficulty of SSR technique is selection of an appropriate reference site on the outcrop of bedrock, which is also free of topographic effects. The last mentioned point is vital for analyzing and interpreting data. The H/V technique assumes that the vertical component is not amplified by the local geology.
In the present paper microtremor data has been analyzed using Standard Spectral Ratio (SSR) and horizontal to vertical Spectral ratio (H/V) techniques using FFT spectral analysis method, in Kamyaran city using a developed code in Matlab software. Kamyaran city is located in Kordestan province in west of Iran with an area of 2950 square kilometers and its elevation from sea level is 1400 meter. It is located in the longitude of 34.793 and the latitude 46.936. Due to the high seismicity of Kamyaran city,a national project was carried out in order to measure the microtremors and evaluate site response and estimate the relation between ground motions and geological structure i.e. amplification level and resonant frequencies. In other words, quantifying and comprehending the local geology and ground motions in Kamyaran City is a main task of this project. Using geological and seismological approaches, we can achieve such a task.
In order to carry out this, two cost effective seismological techniques were used; Horizontal to vertical ratio (H/V)(Nakamura, 1989; Lachet and Bard, 1994; Lermo and Chaez- Garcia, 1994) and Standard Spectral ratio (SSR)(e.g. Borcherdt and Gibbs, 1976;) for processing the recorded microtremor data in studied area. Also, a comparison is made between the results of the two techniques.
Ambient noise data were collected using SSR1 sensor in Kamyaran city; ten appropriate stations were selected among all recorded stations for site effect evaluation.
The H/V technique should be more concentrated, because powerful near field noise causes amplification of both the horizontal and vertical components similarly. Practically, it results in underestimation of real H/V ratio, because of the distortion caused by leakage. In coincidence of the noise and real ground resonant frequencies, these effects are more noticeable. At the end for better interpretation the geological and geotechnical data are used.
H/V technique appears to be more useful in site response evaluation, because it yields precise resonant frequency and amplification factor and it has a greater correspondence with the available geological information. On the other hand, SSR technique (with respect to a reference rock-site station) appears to be useful in sites close to the reference site and it can be used in H/V calibration.
As mentioned before, generally, the results obtained from both techniques are relatively similar in many sites of the studied area. Detailed comparison yields that H/V techniques shows a greater consistency to the available geological data (not mentioned here) than SSR technique. On the other hand, in sites with a close distance to the reference site, SSR technique yields data similar to H/V technique. This result demonstrates the advantage of application of the SSR in interpretation of the obtained H/V ratios. Also, a good way to solve the problems of these methods is applying both techniques and interpretation of the results together, which has been proved by other authors as well.