Looking for curves similarity could be a complex issue characterized bysubjective choices related to continuous transformations of observed discrete data(Chiodi, 1989). Waveforms correlation techniques have been introduced to charac-terize the degree of seismic event similarity (Menke, 1999) and in facilitating moreaccurate relative locations within similar event clusters by providing more precisetiming of seismic wave (P and S) arrivals (Phillips, 1997).In this paper functional analysis (Ramsey, and Silverman, 2006) is considered tohighlight common characteristics of waveforms-data and to summarize these charac-teristics by few components, by applying a variant of a classical clustering method torotated data (Sangalli et al., 2010) according to the direction of maximum variance(i.e. based on PCA rotation of data).
|Numero di pagine||1|
|Stato di pubblicazione||Published - 2010|