European Seismological Commission, 32nd General Assembly, Sept. 6-10, 2010, Montpellier, France
Caccavale M., Satriano C., Convertito V., Di Crosta M., Emolo A., Zollo A., Gallovic F.
Ground-motion shaking maps represent a promising tool to recognize the areas mainly involved in damage and losses after an earthquake. These maps, showing the spatial distribution of some parameters like peak ground acceleration, velocity, spectral acceleration and/or instrumental intensity, play a fundamental role for decision maker to properly direct rescue teams and organize the emergency service operations.
Shakemaps could be particularly useful in those areas with high-seismicity risk levels, such as the Campania-Lucania region in southern Italy. In these areas a very dense and wide dynamic seismic network (ISNet) has been recently installed and ShakeMap® is applied to generate ground shaking map during the earthquake occurrence.
ISNet covers an area of approximately 100 km x 70 km along the southern Apennine chain, and it has been deployed to monitor the active fault system that generated the 1980, MS6.9, Irpinia earthquake. The ISNet configuration comprises an extended star topology that has been designed to ensure fast and robust data recording, transmission and analysis. To ensure a high-dynamic range and to avoid signal saturation, each station is equipped with a 3-C strong-motion accelerometer and 3-C velocimeter. The inter-station distances vary from 10 km to 20 km. ISNet has 29 seismic stations that are grouped into six sub-nets. To evaluate the influence of the station density and the performance of the ShakeMap® implementation, a massive synthetic waveforms data-base has been produced for possible M6.9 earthquakes scenarios located at the center of ISNet. The synthetic waveforms have been computed using a hybrid approach for source modeling, which combines the integral approach (based on the evaluation of the representation theorem) at low frequencies (<1 Hz) and the composite approach at high frequencies (1–20 Hz). Both approaches are based on a common set of sub-sources providing ‘‘k-squared’’ slip distribution. The source model is coupled with full-wavefield Green’ s functions computed, for a 1D layered crustal model, by the discrete wave-number method. A total of 300 scenarios for the M6.9 earthquake have been computed by combining variable nucleation points in the lower half of the fault, and different final slip and rupture velocity distributions over the fault plane. The scenarios provide relatively large variability of the synthetics. For each scenario event, the ShakeMap has been computed allowing us to analyze the variability of acceleration, velocity and instrumental intensity.