The eastern Aleutian tsunami of 1 April 1946 exhibited large waves and caused extensive damage throughout the eastern Pacific, yet the source region inferred from aftershocks was only 80 km wide. Near-source run-ups showed a rapid variation with distance, consistent with a small source. Reconciling the near-source variation with the transpacific reach of the tsunami seems impossible for any sensible earthquake source. We have earlier argued that a landslide source fits the combined near-field and far-field observations better than an earth- quake. Here we show that the motion of the landslide was fast enough to couple efficiently with the tsunami to produce large waves in the far field. Perfect coupling, however, was never accom- plished because the slide motion was always less than celerity. Waves in the reverse direction to slide motion were therefore not suppressed, resulting in 35 m run-up at Scotch Cap on Unimak Island. We further show that the arrival time of the tsunami at Scotch Cap demands a landslide at the edge of the Aleutian Shelf at precisely the location of the 1200 km 2 Ugamak Slide.
This chapter presents three great natural hazards: earthquakes, volcanic eruptions, and tsunamis. These three natural hazards pose increasing risks as economic activity and population continues to grow around the Pacific and eastern Indian Oceans, which are surrounded by convergent plate margins. The concept of convergent plate margin is similar to the concept of a convergent plate boundary except that the former encompasses broader regions on either side of the boundary itself. These are generally associated with subduction zones and pose a severe risk to nearby human populations. The chapter outlines basic plate tectonic concepts, and the most important features of convergent plate margins and subduction zones. The most spectacular surficial manifestations of a convergent margin are the trenches, which mark the plate boundary and where the greatest depths in any ocean are found. Another distinctive feature of most convergent plate margins is a chain of great volcanoes, parallel to and set back from the trench.
The authors propose a method to search and detect the impact of tsunami disaster by integrating numerical modeling, remote sensing and GIS technologies.This method consists of regional hazard/damage mapping, identifying exposed population, and satellite image interpretation in terms of structural damage.The method is implemented to the recent tsunami event, the 2009 tsunami in American Samoa, to identify the structural damage by the tsunami.
For the computation of synthetic seismograms in a generally anisotropic layered earth it is necessary to find the eigenvectors and eigenvalues of the first order elastic system matrix A for many values of wavenumber and frequency. The analytical formulas used to construct the eigenvectors of A in the isotropic case are not available in the general anisotropic case so one must use numerical methods whose speed often depends on an efficient use of the properties of A. First we review the symmetries of A and the conditions under which A is not semi-simple. Then we construct a perturbation theory for the eigenvectors of A. Finally we show how to make A symmetric so that special techniques for symmetric matrices, such as Jacobi iteration, can be used. All the results given here remain valid when the medium is attenuating, i.e. when the elastic coefficients are complex.
The very high compressional and shear velocity gradients of marine sediments may result in continuous interconversion between P (compressional) and S (shear) types of motion at low frequencies. Since the ray theories commonly used in modeling acoustic interaction with the ocean bottom implicitly assume that P and S are decoupled, the importance of such phenomena must be assessed. The problem is investigated here through theoretical studies of the reflectivity function for representative models of the ocean bottom. The only practical approach for including all such wave phenomena in a determination of reflectivities involves numerical solution of the wave equation. For a depth-varying structure, the most efficient numerical scheme is the classical approximation by homogeneous layers. This procedure can be readily modified to isolate the effects of gradient-induced coupling by forcing the P- and S-wave potentials to be independent and studying the effects of this on the reflectivity. Unfortunately, the results of this analysis, being in frequency-wavenumber space, are so difficult to interpret that the physics of the coupling process is obscured. Some insight can be gained by transforming the frequency dependence of the reflectivity function to a time dependence. The resulting function (the plane-wave response) is more amenable to physical interpretation and shows clearly that the principal consequence of coupling is the conversion of shear to compressional motion. As a result, coupling reduces the amplitude of shear arrivals and draws out the tails of compressional arrivals. Above 1 Hz, however, the effects of this coupling are extremely small, so for most marine acoustics applications, the phenomenon can be ignored.
Seismic anisotropy in oceanic layer 2 resulting from a preferred alignment of fractures has been widely recognized, but all experiments to date have sought to measure only the weak azimuthal variation of elastic properties resulting from tectonically controlled systems of vertical fractures. From ocean drilling data, however, especially from DSDP Hole 504B, we know that layer 2 is composed of interleaved massive flows and breccia units, and that the massive units have a very strong concentration of horizontal fractures. Layer 2's pronounced horizontal fabric of low-velocity “layers” (fractures and/or breccia zones) permeating an otherwise high-velocity matrix, will cause P-waves to travel faster horizontally than vertically. This anisotropy has no azimuthal expression, and so cannot easily be recognized in seismic data, but it may lead to overestimation of the thickness of upper crustal layers by as much as 30% in young crust. Further, the anisotropy affects P and S waves differently, so where shear-wave data are available, Poisson's ratio may be substantially underestimated. The widespread observation of a low Poisson's ratio zone in the upper few kilometers of young crust is almost certainly an artifact of ignoring anisotropy. As the crust ages, fractures and voids are filled by chemical alteration and precipitation, the velocity contrast between rock and void-filling material is reduced, and the anisotropy decreases. The errors introduced by assuming isotropy thus show an inverse relationship to crustal age, so that thickness measurements from old crust are probably no more than 10% in error. This explains a long-standing enigma of marine seismology: the apparent thinning of upper crustal layers with age.
