GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 2.1 GNGTS 2024 Earthquake sequences and long-term seismic hazard maps: an oximoron? P. Bazzurro University School for Superior Studies (IUSS) in Pavia, Pavia, Italy Seismic hazard maps for regional or national applications have no interest per se. Their importance resides in their critical support to earthquake risk assessment and, if needed, risk mitigation of specific structures or portfolios of structures. The oldest of the many possible applications of seismic hazard maps involves the definition of the ground motion loads to be used for designing new buildings in such a way that they possess the required level of safety or, equivalently, that have sufficiently low chance of becoming unfit for purpose in a given period of time. Another application, and arguably a more challenging one, involves their use at the basis for assessing the risk that existing buildings of different age and structural typology have to become unfit for occupancy or even to be destroyed by earthquakes in a given period of time. For these applications, and others not mentioned above, it is customary to require that these maps provide a long-term stationary view of the seismic hazard. This has been achieved by adopting a mainshock-only view of the earthquake phenomenon, a tenet that underpins essentially every single hazard map developed worldwide. Evidence has shown, however, that in most parts of the world, including Italy, earthquakes occur in sequences and that large damaging earthquakes not preceded or followed by other nearby events closely spaced in time are a rarity rather than the norm. The larger amount of damage that sequences cause when compared to the damage inflicted by the mainshock only has been apparent for a long time and the Central Italy sequence of 2016-17 is only one recent example. Therefore, given that hazard assessment should serve risk calculations and risk estimates are impacted by the occurrence of all damaging earthquakes, regardless of their label, it is clear that future seismic hazard maps should include the contribution of all earthquakes, not just the mainshocks. Several methods have been proposed to include earthquakes “other” than mainshocks in the hazard/risk calculations, some more elegant than others. We will present a method that allows the development of hazard maps that include the occurrence of realistic sequences that (i) are statistically consistent in time and space with those occurred in the region; (ii) include events with magnitude lower than that of the mainshock of the sequence and that may or may not occur along the same rupture of the mainshock. Such a method, if appropriately managed statistically, may yield maps that still reflect a long-term view estimate of the seismic hazard, but heightened if compared to the traditional hazard estimates that accounts only for mainshocks. The underestimation of the traditional seismic hazard due to the consideration of only mainshocks is, of course, more significant in regions where prolific sequences occur more often. From the risk side more refined engineering models that are able to capture damage accumulation in buildings due to multiple shocks are under development. There is no doubt that these maps are the way of the, hopefully, close future