GNGTS 2021 - Atti del 39° Convegno Nazionale

GNGTS 2021 S essione 2.1 176 AN ANALYSIS OF IMPULSIVE SIGNALS PRODUCED BY EARTHQUAKES IN ITALY D. Ertuncay 1* , P. Malisan 2 , G. Costa 1 , S. Grimaz 2 1 Department of Mathematics and Geosciences, SeisRaM Working Group, University of Trieste, Trieste, Italy 2 Safety and Protection Intersectoral Laboratory - SPRINT-Lab, University of Udine, Udine, Italy The number of seismic stations installed nearby active fault lines increased in the last two decades. These stations allow acquiring information on seismic ground motion in near-fault regions and proved that features of the ground motion in the near-fault regions could be very different from those of far-field motions (Grimaz and Malisan, 2014). In near-fault regions, large-amplitude ground motions could occur with larger periods with respect to far-field. Such signals are called impulsive signals (or impulses) (Somerville et al, 1997) and can be identified by analyzing the velocity-time history recorded by the seismic stations. There are various source characteristics that may create impulsive signals. First is the forward directivity. When the rupture front propagates to a site of interest with the slip direction facing to the site, seismic energy accumulates to a small-time interval and arrives at the station in a short time. Another reason for the impulsive motions is the permanent displacement of the ground, also known as the fling step effect. Displacement-time history of a seismic station that recorded the fling step effect also has peculiar shapes with respect to other seismic records. In this study we want to identify the impulsive signals produced by earthquakes in Italy. To identify the impulsive motion, we apply three different algorithms, namely: Shahi, S. K. and Baker, J. W. (2014); Chang et al. (2016); and Ertuncay D. and Costa G. (2019). Shahi, S. K. and Baker, J. W. (2014) investigate the horizontal components together, whereas the other two algorithms analyze the ground motion signals individually along with the three components. For this reason, Chang et al. (2016) and Ertuncay D. and Costa G. (2019) can be used to analyze the features of the vertical motions. We used the earthquake catalogue and the seismic waveforms stored in the ITACA database (Luzi et al., 2016). We analyzed overall 93 earthquakes with magnitudes larger than M w 5.0. Since the impulsive motions can be detected in fault normal and fault parallel components, we rotated the horizontal channels to the strike angle of the fault, if presented in the database. Table 1 summarizes the numbers of impulsive signals identified in all the analysed earthquakes (only events with at least one impulsive signal have been reported). Data are divided considering if the impulsive signal has been identified along the fault normal (FN), fault parallel (FP) and vertical (Z) direction.