Towards Incorporation of Seismicity Clustering in Probabilistic Seismic Loss Estimation

In current practice, probabilistic seismic risk assessment of distributed portfolios of structures involves the generation of stochastic event catalogues, usually assuming a Poissonian occurrence model. Such models are typically developed based on “declustered” historical and instrumental catalogues of past events. This practice is justifiable for design but the notion that mainshock events can be considered representative of the damage potential of seismic sequences is dubious. Herein, we model seismicity as an epidemic-type aftershock sequence (ETAS) process in order to investigate the effects of seismicity clustering on regional risk assessment. Loss estimates are derived by ETAS-generated stochastic event catalogues for the region of Umbria in Central Italy, and are compared with those obtained with a conventional Poisson-based model. The potential gains of utilizing a model capable of capturing the spatio-temporal clustering features of seismicity are illustrated and offered along with a discussion on the various details and challenges of such considerations.