Simulation-Based Probabilistic Tsunami Hazard Analysis: Empirical and Robust Hazard Predictions
Simulation-Based Probabilistic Tsunami Hazard Analysis: Empirical and Robust Hazard Predictions
复制标题
DOI：
10.1007/s00024-017-1588-9
复制
发表时间：
2017-08-01
影响因子：
2
通讯作者：
Goda, Katsuichiro
中科院分区：
文献类型：
--
作者：
De Risi, Raffaele;Goda, Katsuichiro
概率海啸危险性分析（PTHA）是严格的风险评估的先决条件，因此，有关风险缓解战略的决策。本文提出了一种新的基于模拟的方法海啸灾害评估的特定网站的工程项目沿着海岸，或更广泛地说，为更广泛的海啸易发地区。该方法结合了许多不确定的参数，这些参数与地球物理过程有关，通过采用新的海啸地震区的比例关系。通过所提出的方法，有可能获得单一地点的海啸危险曲线，即海啸强度量度（如淹没深度）与其平均年发生率的关系，或海啸危险图，表示在给定时间窗内发生特定概率的地理区域内的预期海啸强度量度。除了传统的海啸危险曲线是基于经验统计表示的模拟为基础的PTHA结果，本研究提出了一个强大的海啸危险曲线，这是基于贝叶斯拟合方法。鲁棒的方法允许显着减少模拟的数量，因此，减少计算工作量。这两种方法都产生了危险的中心估计值以及置信区间，便于危险不确定性的严格量化。
Probabilistic tsunami hazard analysis (PTHA) is the prerequisite for rigorous risk assessment and thus for decision-making regarding risk mitigation strategies. This paper proposes a new simulation-based methodology for tsunami hazard assessment for a specific site of an engineering project along the coast, or, more broadly, for a wider tsunami-prone region. The methodology incorporates numerous uncertain parameters that are related to geophysical processes by adopting new scaling relationships for tsunamigenic seismic regions. Through the proposed methodology it is possible to obtain either a tsunami hazard curve for a single location, that is the representation of a tsunami intensity measure (such as inundation depth) versus its mean annual rate of occurrence, or tsunami hazard maps, representing the expected tsunami intensity measures within a geographical area, for a specific probability of occurrence in a given time window. In addition to the conventional tsunami hazard curve that is based on an empirical statistical representation of the simulation-based PTHA results, this study presents a robust tsunami hazard curve, which is based on a Bayesian fitting methodology. The robust approach allows a significant reduction of the number of simulations and, therefore, a reduction of the computational effort. Both methods produce a central estimate of the hazard as well as a confidence interval, facilitating the rigorous quantification of the hazard uncertainties.