Theoretical and Computational Contributions to the Modeling of Global Tsunamis

Salaree, Amirmahmood

doi:https://doi.org/10.21985/n2-5wmh-s691

Work

Theoretical and Computational Contributions to the Modeling of Global Tsunamis

Public

Download PDF

I. The distribution of tsunami amplitudes in the open ocean is controlled by source mechanism and bathymetry geometry. Although detailed studies have considered heterogeneity effects in earthquake tsunami sources, little or no attention has been paid to the effects of physical resolution of detailed bathymetry on tsunami waveforms in the far field. Regardless of the simulation method, on one hand, it is desirable to include detailed bathymetry features in the simulation grids in order to predict tsunami amplitudes as accurately as possible, but on the other hand, large detailed grids result in long simulation times. It is therefore of interest to investigate the amount of detail in bathymetric grids that control the most important features in tsunami amplitudes, to assess what constitutes sufficient level for grids in numerical simulations. In this context, we consider the real bathymetry of the Pacific basin and use two different smoothing techniques to decrease the physical resolution of the propagation medium. First, we use a spherical harmonics series approach to decompose the bathymetry of the Pacific Ocean into its components down to a resolution of 4° (l=100) and create bathymetric grids by summing the resulting terms. Secondly, we use a moving average technique and quantify smoothness by assigning a maximum self-similarity threshold. We then use these grids to simulate the tsunami behavior from pure thrust events of different sizes around the Pacific using the MOST algorithm. Application of four different metrics, namely MT (Metric Tsunami), correlation coefficient, frequency domain analysis, and entropy reveal that for large megathrust events (M_0=10^29 dyn-cm), one only needs to consider the sum of the first ~20 coefficients (equivalent to a resolution of ~2000 km, or ~1% surface smoothness of the Pacific grid, in order to reproduce the main components of the true distribution of tsunami amplitudes. This would result in simpler simulations, and faster computations in the context of tsunami warning algorithms. II. We study individual tsunamis of landslide or atmospheric origin, through extensive field surveys and numerical modeling, to unravel the exact mechanism of their generation. In the Caspian Sea (1990) and in the case of an aftershock of the 1923 Kamchatka earthquake, we document generation from seismically triggered landslides, with significant implications in terms of tsunami hazard for these provinces. In the Persian Gulf, we attribute the 2017 Dayyer rogue wave to a meteotsunami presumably correlated with a large scale atmospheric disturbance. Although these non-tectonic events occurred on much smaller scales compared to earthquake tsunamis, we investigate their respective source mechanisms and propagation patterns by compiling appropriate geological/morphological datasets in order to better constrain their modeling as hydrodynamic events. In this respect, we will also be able to quantify the tsunami hazard of the various basins in which these events took place.

Creator