GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.1 GNGTS 2023 MPS04 map with a probability of exceedance of 10% in 50 years (http://zonesismiche.mi.ingv.it/mappa_ps_apr04/italia.html ). As discussed in Magrin et al. (2017), we consider for the comparison this PSHA map because, under the Poisson assumption, it corresponds to a ground motion with a rate of λ =0.0021 (1/y), namely an average occurrence time T = 1/ λ = 475 y (often erroneously labelled as "return period"), which according to the Nyquist–Shannon sampling theorem is compatible with the length of the available earthquake catalog (about 1000 years). Even accounting for specific soil conditions, i.e., considering the ground shaking estimates for soil type C, the peak ground acceleration (PGA) predicted according to the PSHA map is 0.21 g (Gallipoli et al., 2014), a value still lower than that observed. Another large earthquake that struck the Italian territory a few years before the Emilia event, was the L’Aquila earthquake (Mw = 6.3, 2009). Although the L’Aquila earthquake scores as a failure in the earthquake prediction experiment, the time-dependent ground-shaking scenario defined for the period 1 March 2009 to 1 May 2009 correctly predicted the macroseismic intensities, as large as IX (MCS), observed for this earthquake. The latest confirmation of the applied methodology for time-dependent seismic hazard assessment is provided by the recent Amatrice (Mw 6.2) and Norcia (Mw 6.6) earthquakes, which struck the central Italy region on 24 August 2016 and on 30 October 2016, respectively. Specifically, the events took place within the previously identified seismogenic nodes prone to earthquakes with magnitude M ≥  6.0 (Gorshkov et al., 2002), an area that has experienced other destructive earthquakes in the past, including the large Mw 6.9 Valnerina earthquake in 1703. The time-dependent ground-shaking scenario associated with the central region defined for the period 1 November 2012 to 1 September 2016 was very consistent with the ground shaking recorded for the earthquake (Peresan et al., 2016, Panza and Peresan, 2016). In fact, the acceleration range predicted by the then-current NDSHA map (Panza et al., 2012) was 0.3–0.6 g, which contains the values recorded on rock as high as 0.45 g (RAN data: http://ran.protezionecivile.it/IT/dettaglio_evid.php?evid=340867 ). Similarly, the time-dependent scenario defined for the period 1 September 2016 to 1 November 2016, associated with the continued alert by the applied CN algorithm for the central region, correctly anticipated the even stronger ground shaking recorded for the larger Norcia earthquake, reaching 0.6 g at some sites ( http://ran.protezionecivile.it/IT/dettaglio_evid.php?evid=345980) . Since the NDSHA time-dependent scenarios are regularly computed, from 2006, four large earthquakes have struck the Italian territory, Amatrice and Norcia (2016) along with L’Aquila (2009) and Emilia (2012), and in all cases the integrated method correctly predicted the observed ground motion. According to the results described above, we can conclude that the following statement from the recent paper by Slejko et al (2022): "As all approaches may undergo validation against real data, it should be recalled that the NDSHA estimates by Zuccolo et al. (2011) failed to predict the ground motion recorded close to Norcia during the 2016 seismic sequence in central Italy.” is erroneous, as it largely overlooks existing literature and related papers published during the last decade, including the updated NDSHA map by Panza et al. (2012), which naturally replaces all earlier maps,