GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 2.1 GNGTS 2024 Uncertainty in source parameters and seismic hazard estimation Stefano Parolai Department of Mathematics, Informatics and Geosciences, University of Trieste, Via E. Weiss 2, 34128 Trieste, Italy Seismic hazard assessment requires as a key ingredient a definition of earthquake magnitude both when using a probabilistic approach and through scenario calculations. The estimation of the size of an earthquake event, represented by the magnitude, in fact makes it possible to calculate the level of ground shaking expected at a certain distance from the rupture (or epicenter, or hypocenter, depending on the metric used), and to calibrate the probability of occurrence of events of different magnitudes in a certain time period and in a certain space. Over the years, thanks in part to the gradual introduction of new instruments, the development of seismic networks and the refinement of techniques for analyzing recorded data, different magnitude scales have gradually been proposed. For the same event, the magnitude estimates obtained may differ, being made on different frequency bands (period), of the recorded seismic signal, thus related to different processes of the seismic source. Magnitude is also used to estimate, through empirical relationships, seismic energy. Moment magnitude (Mw), introduced to obviate the saturation problem that affected previous magnitudes (e.g., Ms, mb, Ml, Md), is determined using the long period (low frequency) amplitudes of the source Fourier spectra, controlled by the average dislocation on the rupture. However, these spectral periods are little affected by changes in the stress drop that determines the radiated energy in high-frequency seismic waves, which is extremely relevant to seismic action on a wide range of structures. In this presentation, a brief review of some of the magnitude scales discussed above and their differences with a view to engineering applications will be proposed. The scales generally used for seismic hazard assessment and mainly now based on low-frequency spectral ordinates, scale reasonably well, at least within the Italian territory. In fact, the well-known saturation effects, for example for Magnitude Ms, occur only at values close to those of historically known maximum magnitudes. The reasonable scaling persists even when considering the level of uncertainty in magnitude estimation due to the limited observations and the observed shaking variations caused by propagation in the heterogeneous crust. However, this uncertainty must be taken into account for an appropriate treatment of this parameter. Finally, the effects that, due to the limitation of existing scales in capturing source processes, contribute to 'aleatoric uncertainty in ground shaking models will be illustrated.