|Authors||S. Glimsdal, G. K. Pedersen, K. Atakan, C. B. Harbitz, H. P. Langtangen and F. Løvholt|
|Title||Propagation of the Dec. 26, 2004 Indian Ocean Tsunami: Effects of Dispersion and Source Characteristics|
|Afilliation||Scientific Computing, Scientific Computing|
|Publication Type||Journal Article|
|Year of Publication||2006|
|Journal||Int. J. Fluid Mech. Res.|
This work presents numerical simulations of the tsunami generated by the Dec. 26, 2004 Sumatra-Andaman earthquake. The numerical models employed include the linear shallow water equations, a weakly nonlinear and dispersive model (Boussinesq equations), and ray theory for linear hydrostatic waves. Four different tsunami sources, constructed from inversion models based on seismological recordings, are studied. We have investigated the sensitivity to the choice of mathematical model, grid resolution, source parameters, and delay of tsunami generation at the northern part of the source area. The results are compared to surface elevation recordings. Numerical simulations show that the effect of dispersion may modify the waves (slightly) during long propagation times only, and dispersion is not observed in the tsunami generation phase. In some shallow regions, on the other hand, nonlinear steepening of the wave front may enhance dispersion, and undular bores may be produced, which cannot be modeled by the standard shallow water equations commonly used for tsunami simulation. The sensitivity analysis results provide important insights to the source complexity of the Dec. 26, 2004 earthquake.