First European Conference on Earthquake Engineering and Seismology (a joint event of the 13 th ECEE & 30 th General Assembly of the ESC) Geneva. Switzerland. 3-8 September 2006 Paper Number: 1249 SITE EFFECTS ESTIMATION USING EARTHQUAKE AND AMBIENT NOISE DATA: THE CASE OF LEFKAS TOWN (W. GREECE) Petros TRIANTAFYLLIDIS 1 , Nikos THEODULIDIS 1 , Alexandros SAVVAIDIS 1 Christos PAPAIOANNOU 1 and Petros DIMITRIU 1 SUMMARY The town of Lefkas is assigned zone III in the current Greek seismic code and has been repeatedly hit in the past by strong nearby earthquakes, the last one occurring on 14 August 2003 (M6.2). An accelerograph, installed almost thirty years ago in the town of Lefkas, recorded the mainshock with a PGA≈0.42g. After this strong earthquake, four digital, high-resolution accelerographs were installed at selected soil sites of the town and one at a reference “rock” site. During three months of continuous operation, this array recorded several hundreds of aftershocks, providing a large dataset suitable for site effects study in the town of Lefkas. This dataset consists mainly of low- amplitude recordings (PGA<0.05g) and represents the weak-motion dataset. At the aforementioned strong-motion station sites, single-station ambient-noise measurements were performed using a broadband sensor with a high-resolution digitiser. These measurements constitute the ambient-noise dataset. We processed the recordings from both datasets by applying the SESAME-software and a common procedure. First, at all soil sites we evaluated the standard spectral ratio (SSR) and horizontal to vertical component spectral ratio - the so called receiver function (RF) - using the weak motion dataset. Then we used ambient-noise data to estimate HVSR at all stations. Comparison between SSR and RF ratios for earthquake recordings reveals similarity in the evaluated fundamental frequencies but striking differences in their amplitudes, in agreement with previous results. At one soil site the mainshock strong-motion RF is found to be shifted to lower frequencies with respect to weak-motion and ambient-noise ones, indicating non- linear behavior of the corresponding soil materials. An attempt is made to explain the obtained experimental results with 1D modeling. 1. INTRODUCTION Experimental site effects studies are of high importance to microzonation and land use planning in urban areas. In regions of moderate to high seismicity seismic recordings are easily acquired following a mainshock in the vicinity of a residential area by deploying a strong motion network of high resolution recorders. Various techniques are used to better confine the site effects throughout the study area depending on the available number of instruments. The best procedure for determining the site response of a particular sedimentary location is to record ground motion during an earthquake and calculate the spectral ratio of the sediment site recording to a nearby reference site located on rock (e.g. Borcherdt, 1970). However, to gather sufficient data using this technique in a reasonable period of time may be difficult, especially in regions of low seismicity. On the other hand, the numerical prediction of site effects with a reasonable confidence level requires a detailed knowledge of some key geotechnical and geophysical parameters, which is also a difficult task. An alternative approach to characterizing site response in urban environment involves the use of the H/V 1 Institute of Engineering Seismology and Earthquake Engineering, P.O. Box 53, 55102 Thessaloniki, Greece [e-mail : ntheo@itsak.gr ] 1