Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-16T09:54:47.401Z Has data issue: false hasContentIssue false

Acoustic–gravity waves from multi-fault rupture

Published online by Cambridge University Press:  29 March 2021

Byron Williams
Affiliation:
School of Mathematics, Cardiff University, Cardiff CF24 4AG, UK
Usama Kadri*
Affiliation:
School of Mathematics, Cardiff University, Cardiff CF24 4AG, UK
Ali Abdolali
Affiliation:
NWS/NCEP/Environmental Modeling Center, National Oceanic and Atmospheric Administration (NOAA), College Park, MD 20740, USAI.M. Systems Group, Inc. (IMSG), Rockville, MD 20852, USAUniversity of Maryland, College Park, MD 20742, USA
*
†Email address for correspondence: kadriu@cardiff.ac.uk

Abstract

The propagation of wave disturbances from a complex multi-fault submarine earthquake of slender rectangular segments in a sea of constant depth is discussed, accounting for both water compressibility and gravity effects. It is found that including gravity effects the modal envelopes of the modified two-dimensional acoustic waves and the tsunami are governed by the Schrödinger equation. An explicit solution is derived using a multi-fault approach that allows capturing the main peak of the tsunami. Moreover, a linear superposition of the solution allows solving complicated multi-fault ruptures, in particular in the absence of dissipation due to large variations in depth. Consequently, the modulations of acoustic waves due to gravity, and of tsunami due to compressibility, are governed simultaneously and accurately, which is essential for practical applications such as tsunami early warning systems. The results are validated numerically against the mild-slope equation for weakly compressible fluids.

Information

Type
JFM Papers
Copyright
© The Author(s), 2021. Published by Cambridge University Press
Figure 0

Figure 1.Sketch of slender fault.

Figure 1

Table 1.Constants and parameters used in validation of current model with gravity.

Figure 2

Figure 2.Comparison of current model against numerical model and Stiassnie (2010). (a) Bottom pressure signals predicted by current model (top), Stiassnie (2010) (middle) and numerical model (bottom). Coordinates are $x=1000$, $y=0$ km. (b) Surface elevation plots generated by current model (stationary-phase approximation inclusive of compressibility), Stiassnie (2010) and the numerical model. Coordinates are $x=1000$, $y=0$ km.

Figure 3

Figure 3.Panel (a) indicates the fault dimensions ($L=100$ and $b=10$ km), the numerical domain extent and the coordinates of the virtual point observations. The time series of bottom pressure calculated from the current model (b) and extracted from the numerical model (c). Only the first mode is considered in order to keep computation time manageable.

Figure 4

Figure 4.Location and orientation of a slender fault cluster relative to a hydrophone – axes and orientation of the $i$th fault indicated.

Figure 5

Figure 5.Comparison of current model against observation during Tohoku 2011. (a) Depth transect between Tohoku epicentre and DART buoy 21418. (b) Surface elevation comparison between current model (general compressible) and data recorded by DART buoy 21418 (red trace) for Tohoku 2011 event.

Figure 6

Table 2.Constants and parameters used in the calculation of predicted surface elevation at DART buoy 21418 for Tohoku 2011 event.

Figure 7

Table 3.Parameters used for Sumatra 2004 event – ten faults in total. Includes $\zeta$ – the vertical displacement.

Figure 8

Figure 6.Overview of area considered for the bottom pressure map. The section to the left of the black dashed line is that used in the calculations for figure 8. The origin of $x, y$ coordinates is at the earthquake epicentre (yellow star). Fault centroids are shown by blue stars and the faults delineated by rectangles. Depth below sea level is indicated by the colour bar with the white areas at 4 km depth. The four points used to construct the time series of figure 7 are labelled 1, 2, 3 and 4. The transect AB is shown with a dashed line.

Figure 9

Figure 7.(a–d) Dynamic pressure time series for points 1, 2, 3 and 4 (from figure 6). The black trace is the current model (constant depth), the blue trace is a depth-integrated numerical model (constant depth) and the red trace is depth-integrated numerical but with variable depth.

Figure 10

Figure 8.Snapshots of bottom pressure fields at $t=1260$ s (a–c), $t=1800$ s (d–f ) and $t=2340$ s (g–i) from the current model, (5.1) (a,d,g), numerical model for the case of constant depth of 4 km (b,e,h) and numerical model for the case of variable depth (c,f ,i). The domain extent is shown in figure 6 and the boundary forcing is imposed along the dashed line – also figure 6. The dynamic pressure variation is indicated with reference to the colour bar where white corresponds to 0 Pa. The transect AB is shown with dashed line in each subplot.

Figure 11

Figure 9.Maximum absolute values of the bottom pressure ($P$) of the acoustic wave generated by the Sumatra 2004 event during the first 1 h since rupture.

Figure 12

Figure 10.(a) The ocean profile along section AB (as shown in figures 6 and 8). (b–d) Bottom pressure anomalies along transect AB at $t = 1260, 1800$ and 2340 s from the current model (black), numerical model with constant depth (blue) and numerical model with variable depth (red).

Figure 13

Figure 11.(a) The time series of bottom displacement for linear, half-sine ($\zeta _s$) and exponential ($\zeta _e$) functions. (b) The time series of surface elevation ($\eta$) and (c) bottom pressure signals. Coordinates are $x= 1000$, $y = 0$ km.

Williams et al. supplementary movie 1

Bottom pressure fields from the numerical model for a single fault (L= 100 km and b= 10 km). The parameters are given in Figure 3. the numerical domain extent and the coordinates of the virtual point observations are shown.

Download Williams et al. supplementary movie 1(Video)
Video 72.2 MB

Williams et al. supplementary movie 2

Bottom pressure fields from the current model, Eq. (5.1) (left column), numerical model for the case of constant depth of 4 km (middle column) and numerical model for the case of variable depth (right column).

Download Williams et al. supplementary movie 2(Video)
Video 28 MB
Supplementary material: PDF

Williams et al. supplementary material

Supplementary data

Download Williams et al. supplementary material(PDF)
PDF 153.2 KB