GNGTS 2021 - Atti del 39° Convegno Nazionale

69 GNGTS 2021 S essione 1.1 of the near-field appears larger for strike-slip faults (≈1000 km 2 ), with respect to thrusts (≈650 km 2 ) and normal faults (≈400 km 2 ). Considering the InSAR fingerprints as the surface projection of the mobilized seismogenic volume, the volume can be estimated for normal and thrust fault events (geometrically is not quantifiable for strike-slip faults) considering the deformed area as the base of an inverted prism multiplied to the maximum seismogenic depth. Since the epicentral area increases with earth- quake magnitude, the same consequently occurs for the involved volumes. The estimated vol- umes compared with the observed magnitudes doubles by an increase of 0.5 point of magnitude for normal earthquakes or quadruples for thrust earthquakes. It follows that the deformed vol- ume exponentially increases with increasing magnitude and is twice as high in case of thrust fault earthquakes as in the normal fault earthquakes (Petricca et al ., 2021). Focusing on coseismic ground shaking, we analyzed available data recorded by the Ital- ian accelerometric network for events that are also studied in terms of SAR and included in Table 1. Namely they are: i) Colfiorito 1997 (Mw 6.0); ii) L’Aquila 2009 (Mw 6.3); iii) Emilia 2012 1st event (Mw 5.9); iv) Emilia 2012 2nd event (Mw 5.8); v) Amatrice 2016 (Mw 6.1); vi) Norcia 2016 (Mw 6.5). We analyzed the maximum PGA value of accelerometric waveforms recorded at the coseismic stage for each event as follows: i) maximum PGA of the horizontal and vertical com- ponents versus epicenter-station distance and versus the peak of maximum vertical deformation detected by InSAR; ii) vertical-horizontal PGA ratio vs epicenter-station distance. We classified accelerometric stations in two categories: those lying within the polygon that encloses the de- formed area recognized by SAR analysis, i.e., the InSAR epicentral area and those that fall outside the same polygon but within 35 km away (radial distance) from the epicenter. Both the vertical and horizontal components of the maximum PGA always increase approaching the depocenter. The increase of the maximum PGA recorded moving toward the epicenter is also observed. The stations record PGAs 2-3 times greater within the epicentral area identified by the InSAR (red and light-red symbols in Fig. 3) than outside (blue and light-blue symbols in Fig. 3). Local effect of site amplification can be excluded from our observations because the bedrock characteristics (dis- criminated through capital letters in Fig. 3) are homogeneous within the sub-network considered for each event. Comparing PGA components recorded at the same station emerges that their ratio leans towards the vertical direction (Petricca et al ., 2021). 4. Conclusion InSAR data figure out that the epicentral area or near-field has routinely an elliptic shape for thrust and normal faults, whereas it tends to quadrilobate around strike-slip faults, thus provid- ing a fingerprint of the occurred earthquake. The ratio between the long and short axis of the ellipse increases with magnitude in all tectonic settings. In all tectonic settings, the dimension of the deformed epicentral area is determined by the vertical projection of the mobilized hanging wall (thrusts and normal faults) or adjacent (strike slip) that is related to the fault length, depth and dip. Therefore, the coseismically deformed area increases exponentially with the magni- tude size. Unlike the common assumption that the near-field is referred to the distance from the epicenter and from the fault, here we show that the largest shaking is not necessarily along the fault, but where the cumulate ground deformation recorded by InSAR has been maximum and not always coincident with the epicenter. The dimension of the epicentral area is well described by the InSAR data that illuminate the fingerprint of the so-called near-field. Being these areas larger than 100 km 2 for M6 earthquakes, the dimension is definitely too large to be neglected for seismic hazard assessments and the hazard maps should be calibrated for the near-field, i.e, the epicentral area (Grimaz and Malisan, 2014), where stronger shaking is expected to occur within