A 1D P-wave velocity model of the Gargano promontory (south-eastern Italy)

Abstract

We investigate the elastic properties of the crust in the Gargano promontory, located in the northern part of the Apulia region (Southeastern Italy). Starting on April, 2013, a local-scale seismic network, composed of 12 short-period (1 Hz) seismic stations, was deployed on the Gargano promontory. Starting on October, 2013, the network was integrated with the recordings of nine seismic stations managed by the Italian Institute of Geophysics and Volcanology (INGV). The network recorded more than 1200 seismic events in about 15 months of data acquisition, with more than 700 small magnitude events localized in the Gargano promontory and surrounding areas. A Wadati-modified method allowed us to infer VP/VS = 1.73 for the area. A subset of about 400 events having a relatively smaller azimuthal gap (<200°) was selected to calibrate a 1D P-wave velocity model of the area, using the VELEST inversion code. The preferred model was obtained from the average of ten velocity models, each of them representing the inversion result from given initial velocity models, calibrated on previous geological and geophysical studies in the area. The results obtained under the assumption that VP could decrease with depth are unstable, with very different depths of the top of low-velocity layers. Therefore, the velocity model was obtained from the average of the results obtained under the assumption that VP cannot decrease with depth. A strong reduction of both RMS (about 58%) and errors on the location of the events was obtained with respect to the starting model. The final velocity model shows a strong velocity gradient in the upper 5 km of the crust and a small increase (from 6.7 to 7 km) at 30 km of depth. The epicenters of relocated events do not show clear correlations with the surface projection of known seismic faults. A cluster of the epicenters of the relocated events intersects almost perpendicularly the Candelaro fault trace at the surface.