Local amplification and subsoil structure at a difficult site: Understanding site effects from different measurements

We present a detailed site effects study at a site (TYF) close to the Thermaikos gulf coast in Thessaloniki, northern Greece. Different types of data recorded by different instruments are analyzed. Empirical amplification is estimated using spectral ratios relative to a reference station (SSR) and horizontal to vertical spectral ratios (HVSR) using earthquake data. In addition, seismic noise records from different arrays were analyzed using HVSR. Our results show that earthquake data SSR fails for our data. The reason is the poor signal to noise ratio of our records. Better results were obtained using HVSR for earthquake data. Seismic noise HVSR were not useful due to the particular soil profile at TYF, with the exception of HVSR of seismic noise recorded in one of our arrays that were able to reflect a significant change in the coast line at our site. Although amplification at site TYF is relatively small, it is large enough to originate a difference of one intensity unit relative to firm ground motion. Amplification at TYF is caused by a deep soil structure (over 350 m in depth) and therefore cannot be captured using measures like Vs30.