Investigating source directivity for the 2012 Ml5.9 Emilia (Northern Italy) earthquake by jointly using High-rate GPS and Strong motion data

Abstract

On May, 20th 2012, the Ferrara and Modena provinces (Emilia Romagna, Northern Italy) were struck by a moderate magnitude earthquake (Ml 5.9). The focal mechanism is consistent with a ~E-W-striking thrust fault. The mainshock was recorded by 29 high-rate sampling (1-Hz) continuous GPS (HRGPS) stations belonging to scientific or commercial networks and by 55 strong motion (SM) stations belonging to INGV (Istituto Nazionale di Geofisica e Vulcanologia) and RAN (Rete Accelerometrica Nazionale) networks, respectively. The spatial distribution of both HRGPS and SM stations with respect to the mainshock location allows a satisfactory azimuthal coverage of the area. To investigate directivity effects during the mainshock occurrence, we analyze the spatial variation of the peak ground displacement (PGD) measured either for HRGPS or SM sites, using different methods. For each HRGPS and SM site, we rotated the horizontal time series to the azimuth direction and we estimated the GPS-related and the SM-related peak ground displacement (G-PGD and S-PGD, respectively) retrieved by transverse component. However, in contrast to GPS displacements, the double integration of the SM data can be affected by the presence of drifts and, thus, they have to be corrected by quasi-manual procedures. To more properly compare the G-PGDs to the S-PGDs, we used the response spectrum. A response spectrum is simply the response of a series of oscillators of varying natural frequency, that are forced into motion by the same input. The asymptotic value of the displacement response spectrum is the peak ground displacement. Thus, for each HRGPS and SM site, we computed the value of this asymptotic trend (G-PGDrs and S-PGDrs, respectively). This method allows simple automatic procedures. The consistency of the PGDs derived from HRGPS and SM is also evaluated for sites where the two instruments are collocated. The PGDs obtained by the two different methods and the two different data types suggest a source directivity effect in the SE (~120°-150°N) direction.