Geophysical Exploration of the Campi Flegrei (Southern Italy) Caldera' Interiors: Data, Methods and Results. Edited by Aldo Zollo, Paolo Capuano, Margherita Corciulo, 2006
P. Klinc, E. Priolo, L. D’Auria, A. Zollo, G. Seriani, A. Vuan
Several microearthquake waveforms were recorded in the Campi Flegrei area during the 1983-84 seismic swarms caused by an episode of bradisism. The data were acquired by a portable digital network deployed by Univ. of Wisconsin and Vesuvius Observatory and collected into a data-set (Aster and Meyer, 1988). In order to reproduce a synthetic equivalent of that data-set, a 3-D digital model of the structure beneath the Campi Flegrei area has been build-up. The model reproduces the likely, but still hypothetical, velocity structure of the caldera and simulates the presence of the magmatic chamber at the bottom and some isolated bodies representing magmatic chimneys within the model. Being filled by fluid magma, these structural units are opaque to S-waves propagation (i.e. the S-wave velocity approaches to zero). The ultimate goal of this study is to build-up a synthetic data-set of waveforms and arrival times to be used as an input by earthquake tomography methods to estimate the actual resolution power of the “Winsconsin” microearthquake data-set. The computation of such a data-set involves a number of difficulties, such as the design and definition of the 3D digital geo-model, the input preparation and execution of the numerical simulations of seismic fullwave propagation in a 3D complex structure, and the estimation of P- and Sarrival times. In particular, the main difficulty of the last point was that the arrival times should have been picked from a data-set of several thousands of complex waveforms manually. To overcome this problem, we have chosen to compute the theoretical travel times by an independent method. The model design and input preparation tasks have been performed using the Gocad© software tools, whereas the data-set of full-waveforms has been computed using the 3-D Fourier staggered pseudo-spectral method originally developed within this project. Finally, the data-set of P- and S-arrival times has been computed using a numerical solution of the eikonal equation (Podvine and Lecomte, 1991) and NonLinLoc software (Lomax, 1999).