LABORATORY MEASUREMENTS OF ULTRASONIC WAVE VELOCITIES IN ROCKS FROM THE CAMPI-FLEGREI VOLCANIC SYSTEM AND THEIR RELATION TO OTHER FIELD DATA

The Campi Flegrei caldera, located near the city of Naples, Italy, has shown signs of unrest, characterized by large ground deformation, sei smic activity, and changes in the hydrothermal system since 1970. Any attempt to model the processes giving rise to these phenomena requires some knowledge of the subsurface structure and of the physical proper ties of the rocks at depth. In this study we report the results of lab oratory measurements of ultrasonic compressional and shear wave veloci ties in rock samples cored in 4 different geothermal wells reaching a maximun depth of 3 km. The wells are located in the San Vito (SV) and Mofete (MF) areas, at about 3 km north and 4 km west, respectively, fr om the center of the caldera. We selected 10 core samples, 7 from the SV wells and 3 from the MF wells, the attention being focused on the m ost important rock formations underneath Campi Flegrei and on the lowe r sections of the wells where data had been lacking. The measurements involved the transmission of ultrasonic waves along three orthogonal d irections through dry and water-saturated rocks at room conditions. Th e samples axe only moderately anisotropic (13% for one SV sample and < 8% for all the other samples). Wave velocities generally increase with the depth of the sample. The range of velocities is 3-5.4 km s-1 for P waves and 1.5-2.9 km s-1 for S waves. For the SV samples, water satu ration generally implies slightly larger (<10%) V(P) values and smalle r (<20%) V(S) values as compared with ''dry'' velocities. For the MF s amples, water saturation implies both higher V(S) and V(S) velocities (up to 35% and 18%, respectively). These results may be explained by t he different type of porosity of the SV samples (vesicular) and the MF samples (fissural). The ultrasonic measurements on water-saturated sa mples are slightly higher than the available in situ V(P) sonic measur ements (25% for one SV sample and <15% for all the other samples). The se differences can be explained by velocity dispersion between ultraso nic and sonic frequencies and by the different pressure-temperature co nditions existing at depth.