Detecting clusters in spatially correlated waveforms

Seismic networks often record signals characterized by similar shapes that provide important information according to their geographic positions. We propose an approach to identify homogeneous clusters of seismic waves, combining analysis of waveforms with metadata and spectrogram information. In waveforms clustering, cross-correlation measures between signals may presents some limitations, so we refer to more recent contributes relating data-depth based clustering analysis. The mechanism for alignment is also an important topic of the analysis: warping (or aligning) procedures identify nuisance effects in phase variation, that, if ignored, may result in a possible loss of information and the immediate consequence is that the underlying pattern could not be retained. The effectiveness of the approach is investigated by mean of real data. The data consist of a set of recordings of 21 earthquakes in the Centre of Italy with magnitude more than 5.5 mw, provided by the seismic network RAN (Rete Accelerometrica Nazionale) managed by the Italian Department of Civil Protection, are obtained from ESM/ITACA database (esm.mi.ing.it; itaca.mi.ingv.it).The signals were recorded by stations, whose distances from the epicenter are in the range from 50 to 100 km. The goal is dividing the spatial domain into homogeneous clusters and extracting information from the shapes of the underlying curves. This work is supported by National grant MIUR, PRIN-2015 program, Prot.20157PRZC4: Complex space-time modeling and functional analysis for probabilistic forecast of seismic events.