A carregar...
Publicação
Avaliação da consistência da deformação sísmica nos modelos de recorrência usados no cálculo probabilista da perigosidade sísmica
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.11 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Para avaliar a perigosidade sísmica de uma região é muitas vezes utilizada uma abordagem probabilista, PSHA (Probabilistic Seismic Hazard Assessment). Em áreas de deformação lenta, como a Ibéria, os catálogos sísmicos estão incompletos e as estimativas de perigosidade sísmica estão propensas a grandes incertezas. Este facto pode explicar por que razão alguns estudos recentes de PSHA mostraram, para Portugal continental, resultados muito diferentes, levando a preocupações entre a comunidade científica, dúvidas entre os decisores políticos e o público em geral. A existência de vários estudos de perigosidade sísmica a nível europeu e nacional, que apresentam mapas de perigosidade para a região de Portugal continental bastante diferentes, motivou a realização desta tese. Aqui, em resposta a essas preocupações, investigamos as origens dessas diferenças, concentrando-nos nas duas primeiras etapas do PSHA: divisão de zonas sismogénicas e modelos de recorrência sísmica. Os cinco estudos aqui escolhidos, que apresentam diferentes modelos de divisão de zonas sismogénicas foram o estudo que resultou da elaboração dos Anexos Nacionais do Eurocódigo 8 (EC8); o estudo no âmbito do projecto Estudo do Risco Sísmico e de Tsunamis do Algarve (ERSTA); o projecto europeu Seismic Hazard Harmonization in Europe (SHARE); o projecto QREN financiado por várias instituições nacionais e, a proposta apresentada por Vilanova e Fonseca em 2007. Para restringir este estudo foram escolhidas duas zonas sismogénicas bastante relevantes, a zona do Vale Inferior do Tejo (VIT), por englobar a cidade de Lisboa onde está estimado um risco sísmico elevado, e o Golfo de Cádis onde se gerou o sismo e tsunami de 1755, sendo este o sismo de referência para Portugal continental, e também tendo em conta que esta área é de grande complexidade geológica. Utilizámos dois métodos para avaliar a consistência de cada modelo de geração escolhido com a deformação tectónica, um teste simplificado e outro complexo. No teste simples, o momento sísmico total libertado é convertido numa velocidade relativa entre blocos, para uma fácil interpretação. O método complexo, tendo por base os resultados da modelação neotectónica de Neres et al., (2016), fornece uma distribuição da deformação que pode ser convertida em taxas de actividade sísmica para uma determinada zona sismogénica. O cálculo é realizado utilizando uma versão modificada do programa “Long Term Seismicity” (Bird et al., 2010), que considera os parâmetros da zona fornecidos pelo utilizador, e usa a lei “Tapered” de Gutenberg-Richter. Foi, também, tida em conta a utilização de um esquema em Árvore Lógica pelos modelos de Vilanova et al. (2007) e SHARE, e o método de Monte-Carlo na proposta de Vilanova et al. (2007), avaliando assim as incertezas associadas num estudo de PSHA. Os resultados obtidos possibilitaram uma boa avaliação da consistência, ou inconsistência, dos vários modelos de geração estudados, com a deformação tectónica, inferindo assim a confiança a depositar nos modelos de geração sísmica.
To evaluate the seismic hazard of a region, a Probabilistic Seismic Hazard Assessment (PSHA) is often used. In slowly deforming areas, like Iberia, the earthquake catalogues are incomplete and earthquake hazard estimates are prone to large uncertainties. This fact may explain why some recent Probabilistic Seismic Hazard Assessment (PSHA) studies for Portugal mainland have shown very different results leading to concerns among the scientific community and doubts among decision makers and the public in general. The existence of several seismic hazard studies at European and national level, which present very different hazard maps for Portugal mainland, was the motivation for this thesis. Here, as a response to these concerns we investigate the sources of these differences concentrating on the first two steps of PSHA: the earthquake source zoning and the earthquake recurrence models. The five studies chosen here, which present different models of division of seismogenic zones were the study that resulted from the preparation of the National Annexes of Eurocode 8 (EC8); the study in the scope of the Project of the Seismic Risk and Tsunamis of the Algarve (ERSTA); the European project Seismic Hazard Harmonization in Europe (SHARE); the QREN project financed by several national institutions, and the proposal presented by Vilanova and Fonseca in 2007. In order to restrict this study, two very important seismogenic zones were chosen, the Vale Inferior do Tejo (VIT) zone, since it encompasses the city of Lisbon where a high seismic risk is estimated, and the Gulf of Cadiz where the 1755 earthquake and tsunami were generated, this being the reference earthquake for Portugal mainland, and also considering that this area is of great geological complexity. We used two methods to evaluate the consistency of each recurrence model chosen, with tectonic deformation, a simplified test and a complex one. In the simple method, the total seismic moment released is converted into a relative velocity between blocks, for easy interpretation. The complex method, based on the results of the neotectonic modelling of Neres et al., (2016), provides a strain distribution that can be converted into earthquake rates (seismic activity rates) for a given seismogenic zone. The computation is done by a modified version of the Long-Term Seismicity code (Bird et al., 2010), that considers the zone parameters as provided by the user and uses the “Tapered” Gutenberg-Richter law. It was also considered the use of a Logic Tree by the models of Vilanova et al., (2007) and SHARE, and the Monte Carlo approach by Vilanova et al. (2007), in order to assess the associated uncertainties in a PSHA study. The results obtained allowed a good evaluation of the consistency, or inconsistency, of the various earthquake recurrence models studied, with the tectonic deformation, thus inferring the confidence to be deposited in those seismic models.
To evaluate the seismic hazard of a region, a Probabilistic Seismic Hazard Assessment (PSHA) is often used. In slowly deforming areas, like Iberia, the earthquake catalogues are incomplete and earthquake hazard estimates are prone to large uncertainties. This fact may explain why some recent Probabilistic Seismic Hazard Assessment (PSHA) studies for Portugal mainland have shown very different results leading to concerns among the scientific community and doubts among decision makers and the public in general. The existence of several seismic hazard studies at European and national level, which present very different hazard maps for Portugal mainland, was the motivation for this thesis. Here, as a response to these concerns we investigate the sources of these differences concentrating on the first two steps of PSHA: the earthquake source zoning and the earthquake recurrence models. The five studies chosen here, which present different models of division of seismogenic zones were the study that resulted from the preparation of the National Annexes of Eurocode 8 (EC8); the study in the scope of the Project of the Seismic Risk and Tsunamis of the Algarve (ERSTA); the European project Seismic Hazard Harmonization in Europe (SHARE); the QREN project financed by several national institutions, and the proposal presented by Vilanova and Fonseca in 2007. In order to restrict this study, two very important seismogenic zones were chosen, the Vale Inferior do Tejo (VIT) zone, since it encompasses the city of Lisbon where a high seismic risk is estimated, and the Gulf of Cadiz where the 1755 earthquake and tsunami were generated, this being the reference earthquake for Portugal mainland, and also considering that this area is of great geological complexity. We used two methods to evaluate the consistency of each recurrence model chosen, with tectonic deformation, a simplified test and a complex one. In the simple method, the total seismic moment released is converted into a relative velocity between blocks, for easy interpretation. The complex method, based on the results of the neotectonic modelling of Neres et al., (2016), provides a strain distribution that can be converted into earthquake rates (seismic activity rates) for a given seismogenic zone. The computation is done by a modified version of the Long-Term Seismicity code (Bird et al., 2010), that considers the zone parameters as provided by the user and uses the “Tapered” Gutenberg-Richter law. It was also considered the use of a Logic Tree by the models of Vilanova et al., (2007) and SHARE, and the Monte Carlo approach by Vilanova et al. (2007), in order to assess the associated uncertainties in a PSHA study. The results obtained allowed a good evaluation of the consistency, or inconsistency, of the various earthquake recurrence models studied, with the tectonic deformation, thus inferring the confidence to be deposited in those seismic models.
Descrição
Tese de mestrado, Ciências Geofísicas (Geofísica Interna) Universidade de Lisboa, Faculdade de Ciências, 2018
Palavras-chave
PSHA Vale Inferior do Tejo Geração do sismo de 1755 Modelos de recorrência Modelação Neotectónica Teses de mestrado - 2018