SharePoint
Aide
Faire avancer la sûreté nucléaire

La Recherchev2

Publications

Rupture complexity of a moderate intraplate earthquake in the Alps: the 1996 M5 Epagny-Annecy earthquake.


Fermer

Authentification

Email :

Mot de passe :

Courboulex, F; Deichmann, N; Gariel, JC. GEOPHYSICAL JOURNAL INTERNATIONAL, 139: (1) 152-160.

Type de document > *Article de revue

Mots clés > séismes, instrumentation, séismes

Unité de recherche > IRSN/DEI/SARG/BERSSIN

Auteurs > GARIEL Jean-Christophe

Date de publication > 01/10/1999

Résumé

The magnitude 5 Epagny-Annecy earthquake of 1996 July 15 is the largest seismic event to have occurred in the Alps since the introduction of modern digital instrumentation. This strike-slip event was located on the Vuache Fault, near the town of Annecy, in the northern French Alps. The aim of our work was to retrieve the main parameters of the rupture process of this earthquake from seismograms recorded at local and regional distances (20-300 km). To eliminate path and site effects from the seismograms, we compared the main shock recordings at each station with those of the largest aftershocks nearby. We used a combination of techniques, including pulse-width measurements and cross-correlation of velocity traces, comparison of P-wave displacement pulses, and empirical Green's function deconvolution, to retrieve the apparent duration of the rupture process as seen at each station. Our results demonstrate that, in the absence of on-scale data, P-wave pulse-width measurements on clipped signals can be misleading if the rupture process is complex. In the case of the Annecy earthquake, comparisons of on-scale P-wave displacement seismograms and the empirical Green's function deconvolutions show that the rupture process consisted of at least two subevents separated by 0.2-0.3 s, and with a total duration of about 0.5 s. The systematic azimuthal dependence of both the shape and duration of the apparent source-time function is consistent with a nearly unilateral propagation of the main rupture phase in a southeast direction along the fault plane and parallel to the direction of slip. An isochron analysis reveals that the first subevent occurred slightly to the northwest of the nucleation point but that the second subevent was located further to the southeast, thus confirming the overall rupture directivity towards the southeast. An interpretation of our results in light of the previously documented aftershock distribution and of observations of ground cracks in the epicentral area suggests that the main shock occurred on the Vuache Fault, and that rupture in a northwest direction was inhibited by a right-lateral stepover in the fault. Accordingly, the vast majority of the subsequent aftershocks, which include several magnitude 3-4 events, occurred on a fault segment that is slightly offset from the inferred surface trace of the Vuache Fault and that was activated by the main shock.