Probabilistic seismic hazard analysis (PSHA) (Cornell 1968) is currently the standard approach adopted worldwide for hazard assessment within design building codes, such as in Italy (NTC18). Key ingredients of the PSHA are the ground motion models (GMMs) that provide the median predictions and corresponding uncertainty of the shaking parameters. Although GMMs are widely calibrated worldwide due to the growing number of recorded data, they are still poorly constrained in near-source conditions or volcanic regions. However, recent studies have underlined the importance of correctly quantifying these effects in the hazard estimates, as design seismic actions can often be exceeded, for example in the epicentral area for moderate to high magnitude earthquakes. This thesis aims to investigate the impact, within a probabilistic seismic hazard calculation, of new Italian GMMs specifically calibrated for shallow crustal earthquakes and for volcanic areas, or adjusted with empirical factors, to account for the global effects of near-source ground motion. In the analyses, particular attention is also given to assessing the variability associated with the models used as well as to the impact on hazard estimates of the vertical ground motion component, which can be significantly larger than its horizontal counterpart in the near-source region at short periods. Special care was devoted as well to the used of a specific volcanic model and to its impact in the PSHA, to consider peculiar characteristics of ground motion in areas mainly affected by shallow volcanic events, such as in Sicily. The main findings of this work have shown that significant differences may arise in UHS when different GMMs are used, depending on the distance and the vibration periods at play, while the type of modelling of the uncertainties of the GMMs had limited influence. A limited influence was shown as well for the way the vertical component was modelled inside the PSHA. The impact of the near source correction factors of the GMMs was proved to be meaningful for certain scenarios and periods ranges. For volcanic areas, the use of a specific GMM provided to have limited influence on the seismic hazard, due to the dominance of the tectonic seismicity of the adjacent source zones, particularly at short periods.
L'analisi probabilistica della pericolosità sismica (PSHA) (Cornell 1968) è attualmente l'approccio standard adottato in tutto il mondo per la valutazione della pericolosità all'interno dei codici di progettazione degli edifici, come in Italia (NTC18). I principali ingredienti della PSHA sono i modelli di moto del suolo che forniscono le previsioni mediane e la corrispondente incertezza dei parametri di scuotimento. Sebbene i GMM siano ampiamente calibrati in tutto il mondo, grazie al crescente numero di dati registrati, sono ancora scarsamente vincolati in condizioni prossime alla sorgente o in regioni vulcaniche. Tuttavia, studi recenti hanno sottolineato l'importanza di quantificare correttamente questi effetti nelle stime di pericolosità, poiché le azioni sismiche di progetto possono spesso essere superate, come ad esempio avviene in area epicentrale per terremoti di magnitudo medio elevati. Questa tesi si propone di indagare l'impatto, all'interno di un calcolo probabilistico della pericolosità sismica, di nuovi GMM italiani specificamente calibrati per i terremoti di crosta superficiale o regolati con fattori empirici, per tenere conto degli effetti globali di campo vicino del moto del suolo. Nelle analisi, particolare attenzione è stata data anche alla valutazione della variabilità associata ai modelli utilizzati e all'impatto sulle stime di pericolosità della componente verticale del moto del suolo, che può essere significativamente più grande della sua controparte orizzontale in campo vicino, a corti periodi. Particolare interesse è stato dedicato anche all'utilizzo di uno specifico modello vulcanico e al suo impatto nella PSHA, per considerare le caratteristiche peculiari del moto del suolo in aree interessate da eventi vulcanici poco profondi, come in Sicilia. I principali risultati di questo lavoro hanno mostrato che possono emergere differenze significative nell'UHS quando vengono utilizzati diversi GMM, a seconda della distanza e dei periodi di vibrazione in gioco, mentre il tipo di modellazione delle incertezze dei GMM ha avuto un'influenza limitata. Un'influenza limitata è stata evidenziata anche per il modo in cui la componente verticale è stata modellata all'interno del PSHA. L'impatto dei fattori di correzione delle sorgenti vicine dei GMM si è dimostrato significativo per alcuni scenari e periodi. Per le aree vulcaniche, l'uso di uno specifico GMM ha dimostrato di avere un'influenza limitata sulla pericolosità sismica, a causa della dominanza della sismicità tettonica delle zone sorgente adiacenti, in particolare a brevi periodi strutturali.
Impact of italian ground motion models in seismic hazard assessment: case studies in near-source region and volcanic context
Cipriano, Arianna
2021/2022
Abstract
Probabilistic seismic hazard analysis (PSHA) (Cornell 1968) is currently the standard approach adopted worldwide for hazard assessment within design building codes, such as in Italy (NTC18). Key ingredients of the PSHA are the ground motion models (GMMs) that provide the median predictions and corresponding uncertainty of the shaking parameters. Although GMMs are widely calibrated worldwide due to the growing number of recorded data, they are still poorly constrained in near-source conditions or volcanic regions. However, recent studies have underlined the importance of correctly quantifying these effects in the hazard estimates, as design seismic actions can often be exceeded, for example in the epicentral area for moderate to high magnitude earthquakes. This thesis aims to investigate the impact, within a probabilistic seismic hazard calculation, of new Italian GMMs specifically calibrated for shallow crustal earthquakes and for volcanic areas, or adjusted with empirical factors, to account for the global effects of near-source ground motion. In the analyses, particular attention is also given to assessing the variability associated with the models used as well as to the impact on hazard estimates of the vertical ground motion component, which can be significantly larger than its horizontal counterpart in the near-source region at short periods. Special care was devoted as well to the used of a specific volcanic model and to its impact in the PSHA, to consider peculiar characteristics of ground motion in areas mainly affected by shallow volcanic events, such as in Sicily. The main findings of this work have shown that significant differences may arise in UHS when different GMMs are used, depending on the distance and the vibration periods at play, while the type of modelling of the uncertainties of the GMMs had limited influence. A limited influence was shown as well for the way the vertical component was modelled inside the PSHA. The impact of the near source correction factors of the GMMs was proved to be meaningful for certain scenarios and periods ranges. For volcanic areas, the use of a specific GMM provided to have limited influence on the seismic hazard, due to the dominance of the tectonic seismicity of the adjacent source zones, particularly at short periods.| File | Dimensione | Formato | |
|---|---|---|---|
|
2020_12_Cipriano.pdf
accessibile in internet per tutti
Dimensione
9.78 MB
Formato
Adobe PDF
|
9.78 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/197115