Time series pre-processing techniques for improving the fundamental frequency assessment in H/V seismic spectral ratios

In the context of seismic site effect characterisation, the H/V technique is able to identify efficiently the fundamental frequency of a soft surface layer upon a bedrock. Compared to other techniques for such experimental investigations, the H/V approach is one of the cheapest and therefore most widely used. The direct application of the H/V spectral ratio, however, provides results that are not always easy to analyse; for this purpose, it can be extremely useful to look for pre-processing algorithms that help highlight the characteristics of the signal we are most interested in, in particular, the position of the peak associated to the fundamental frequency. In this paper we summarise the main lines of research developed by our research group at the �Universit� degli Studi di Udine�, i.e. the Singular Spectrum Analysis (SSA) and the wavelet analysis (WASEE). A particular focus is put on the application of these techniques in combination with a previous rotation of the coordinate system in order to maximise the efficiency provided by the SSA and/or the WASEE.