Frequency variation in site response over long and short time scales, as observed from strong motion data of the L’Aquila (2009) seismic sequence

Abstract

Previous works based mainly on strong-motion recordings of large Japanese earthquakes showed that site amplification and soil fundamental frequency could vary over long and short time scales. These phenomena were attributed to non-linear soil behaviour due to inelastic, softening non-linearity: the starting fundamental frequency and amplification were both decreasing and not recovering for a time varying from few hours to several months. The recent April 6th 2009 earthquake (MW 6.3), occurred in the L'Aquila district (central Italy), gave us the possibility to test hypotheses on time variation of amplification function and soil fundamental frequency, thanks to the recordings provided by a preexisting strong-motion array and by a large number of temporary stations. We performed spectral ratio studies for the permanent stations of the Aterno Valley array where a reference station was available. The temporary stations and permanent ones were studied using time-frequency analyses through the S-Transform approach (Stockwell et al., 1996). Finally, analyses on noise recordings were performed, in order to study the soil behaviour in linear conditions. The results provided puzzling evidences. Concerning the long time scale, little variation was observed at the permanent stations of the Aterno Valley array. As for the short time-scale variation, the evidence was often contrasting, with some station showing a time-varying behavior, while others did not change their frequency with respect to the one evaluated from noise HVSR. Even when a time-varying fundamental frequency was observed, it was difficult to attribute it to a classical, softening non-linear behaviour. Even for the strongest recorded shocks, with PGA reaching 0.7 g, variations in frequency and amplitude seems not relevant from building design standpoint. The only exception seems to be the site named AQV, where the analyses evidence a fundamental frequency of the soil, shifting from 3 Hz to about 1.5 Hz during the mainshock.