French-Japanese Workshop on Earthquake Source, Paris-Orléans, France 5-9 October 2009
G. Festa, A. Emolo, E. Lucca, V. Convertito, J. Zahardnik and A. Zollo
On April 6, 2009, a M 6.3 earthquake shook L’Aquila city, in central Apennines,Italy, causing about 300 deaths and large damages to public and private buildings in the city and its outskirts. The earthquake originated on a low-angle normal fault, located just underneath the city, and propagated in a complex
geological structure, with a shallow alluvial basin bounded by steep mountains. Such a strong structural heterogeneity reflected into a large spatial variability of the recorded ground motion and the damage. Along a profile crossing the valley in the dip-parallel direction, the recorded PGA increased from 0.5 g at the center of the valley to more than 1 g on the top of the mountain; the displacement was as high as 20 cm on the vertical component of a station located in the city center; generally the maximum ground motion amplitudes in the fault vicinity were reached on the vertical component rather than on the horizontal ones. Such a near-fault dataset is therefore expected to assess the effects of the near-source radiation in the fault vicinity in a broad range of frequencies. Discriminating if most of the damages were caused by near-source effects, local amplification of waves or high vulnerability of structures is also the objective of the working group of the Naples University which combines geophysicists, geologists, geotechnical and structural engineers. It aims at giving indications about the redefinition of the seismic input to be introduced in the code for building construction regulation.
Specifically, this study is aimed at investigating the role of the geometry, of the focal mechanism, of the up-dip directivity, of the near-source radiation and its coupling with the propagation medium and of the amplification of waves in the basin in building up the broad range of frequencies observed in the data. This will be done by kinematic inversion of regional strong motion data to retrieve the characteristics of the slip and rupture velocity at large wavelengths, by trial-and-error 1D modeling to retrieve high-frequency patches of slip on the fault plane and by direct 3D- Spectral element modeling coupling the complexity of the rupture with the shape of the basin and steep topography.