Articoli in Libri
Iterative tomographic analysis based on automatic refined picking
Conception, verification and application of innovative techniques to study active volcanoes, Edited by Warner Marzocchi e Aldo Zollo, ISBN 978-88-89972-09-0, 2008.
C. Satriano, A. Zollo, C. Rowe
We developed a new, iterative scheme for 3D seismic tomography built on semi-automatic phase picking of first arrivals.
Based on the original work of Rowe et al. (2002), the technique uses cross-correlation to examine waveform similarity and to perform picking adjustment. Starting from a preliminary set of picks as reference, we cross-correlate each pair of traces, following a station-by-station approach. The maximum correlation time is an estimate of the relative time-lag between traces, while the maximum correlation amplitude quantifies waveform similarity and is used as a weighting function in the process of pick adjustment using a smaller correlation window. Additionally, maximum correlation values can be used to build clusters of similar records. Adopting a clustering process can be advantageous in cases where “a priori” assumptions regarding the waveform similarity (such as the distance between sources and receivers) may be inappropriate. Automatic picks, obtained from the first step of the process, are used to build an initial 3D tomographic model through a standard inversion code (here we use Benz et al., 1996). Travel times calculated from the resulting model are then taken as reference picks for the next step of automatic pick adjustment. The result is a tomographic image, refined at each step of the procedure.We tested this procedure on the waveform data set recorded during the 2001 SERAPIS active seismic survey in the gulfs of Naples and Pozzuoli, Italy. We show that the technique can be effectively employed for fast analysis of large data-sets to support rapid decisions such as re-orienting a seismic survey or setting-up automatic 4D tomographic investigations.