Search

Ambient noise measurements acquired in Yalova, which was highly damaged during the 1999 Izmit earthquake, are analyzed to explore the site characteristics. The region of Yalova is governed by complex geological and geomorphological structures consisting of river beds extending from the mountains to the sea, ridges between them, plains in front of them with different size, and the sea coast. As a result of these shallow geological features, the H/V curves exhibit complex patterns. Clear peaks in the H/V curves, which can be interpreted as reliable site resonance frequency, are observed only at about half of the measurement sites. At the remaining sites industrial peaks, broad peaks, or flat responses dominate the spectral ratio graphs. We observed that man-made noises generated by marble cutting machines in Hersek delta mask the site resonance frequencies or can be misinterpreted as a resonance frequency. In total, we identified three anthropogenic noise sources at fundamental frequencies of 1.3, 1.5, and 1.7 Hz along with their two- and threefold harmonics. The parts of H/V curves showing unusual low scattering can be a clue to identify anthropogenic effects. In the assessment of H/V curves, the site location and the similarity of the near surface geology were taken into account. The Laledere plain with thick and soft sediment sequence surprisingly displays flat responses due to a possible low impedance contrast. The Ciftlikkoy and Hacimehmet plains exhibit clear resonance peaks at nearly 1 Hz possessing the largest amplitudes. These sites experienced the highest damage in Yalova during the Izmit earthquake. In contrast, the Cinarcik region which was also exposed to high damage, do not show any obvious amplifications on the H/V curves. Generally, the H/V curves for valley and ridge sites in Yalova reveals a resonance peak at approximately 1 Hz and almost flat curves, respectively. However, several sites on the ridges and valleys portray different patterns. [ABSTRACT FROM AUTHOR]

The site response at 15 stations in the Adana-Ceyhan region (Southern Turkey) is calculated from the recordings of aftershocks of June 27, 1998 Adana-Ceyhan earthquake (M S=6.2) by using the Standard Spectral Ratio (SSR) and the Horizontal-to-Vertical Spectral Ratio (HVSR) methods. While the two methods are in good harmony at a few stations in determining the site effects, they show differences on the estimated amplifications or on the site resonance frequencies at most stations. It was not clear which one of the two methods underestimates or overestimates the amplification values. We observe that at some stations, where the local site conditions are rather complex, the vertical component records are strongly influenced from the local soil conditions. Thus, the HVSR method fails at these stations. The SSR method underestimates the amplifications at some stations since the rock site, selected as reference site, has its own site response and/or the path correction we applied, considering the geometrical spreading factor only, is insufficient. At the sites where high intensity values were observed, we found high amplifications. The fundamental soil frequencies characterize the damage properties observed in the Adana-Ceyhan earthquake. The fundamental soil frequency is nearly at 1.1 Hz at the Ceyhan site, where severe damage was observed in the 5-6 story buildings, while the fundamental soil frequency is between 3-6 Hz at the Adana site, where damage was in the low-story buildings. Therefore, in addition to inefficient construction practices, it is clear that the resonance effects have also contributed to the observed damage. [ABSTRACT FROM AUTHOR]

Ground motion amplifications in the Dinar basin, and contributions of the surface waves generated from basin edges are investigated in frequency and time domains. Amplification functions are computed from the aftershock data of the October 1, 1995 Dinar earthquake (M W=6.4) using the Standard Spectral Ratio method which requires a pair of instruments; one located at the site under investigation (generally on alluvium) and the other on a reference site, preferably a nearby rock site. First, a time window covering the whole signal is used to compute the amplification function, and, successively, the noise, P wave, S wave and the surface wave time windows are used in computation to observe their contributions to the amplification function. It is seen that the maximum amplifications observed at about 2.0 Hz on the amplification functions of the stations located in the basin are largely due to basin edge induced surface waves. These waves have significantly increased the duration of signals recorded within the basin. Particularly, on the vertical component records, the amplitudes of surface waves are larger than the S-wave amplitudes. The periods of waves amplified maximally due to the basin structure coincide with the natural periods of 4-6-story buildings which were heavily damaged in Dinar. This indicates that the site effects may have been important regarding the damage which occurred during the Dinar earthquake of October 1995. [ABSTRACT FROM AUTHOR]