1984
DOI: 10.1111/j.1365-246x.1984.tb01944.x
|View full text |Cite
|
Sign up to set email alerts
|

Matrix methods in synthetic seismograms

Abstract: Some methods are examined for the matrix problems arising in the computation of seismograms in a stratified medium. In particular, the Thornson-Haskell and Kennett algorithms are placed in a common setting, and their connections with standard methods in numerical analysis are given. Finally, a new formulation is presented which has advantages both for physical insight and for numerical accuracy.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
22
0

Year Published

1990
1990
2020
2020

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 46 publications
(22 citation statements)
references
References 20 publications
0
22
0
Order By: Relevance
“…The problem with the exponential dichotomy arises in the high-frequency range, when some of the eigenvalues of A have large real parts with opposite signs, resulting in rapidly decaying as well as growing exponentials (Chin et al, 1984). These are associated with standing waves attenuated in both senses along the axial direction.…”
Section: Solution Of the State Equationmentioning
confidence: 99%
“…The problem with the exponential dichotomy arises in the high-frequency range, when some of the eigenvalues of A have large real parts with opposite signs, resulting in rapidly decaying as well as growing exponentials (Chin et al, 1984). These are associated with standing waves attenuated in both senses along the axial direction.…”
Section: Solution Of the State Equationmentioning
confidence: 99%
“…(4) for the transfer matrix by using the geometric integrator concept. However, it is well known the transfer matrix method becomes inherently unstable due to the coexistence of exponentially decaying and growing matrix elements associated with nonhomogeneous waves, the problem that has been known as 'exponential dichotomy' in the numerical solution of linear matrix di erential equations (Chin et al, 1984;Wang and Rokhlin, 2004a). The sti ness matrix method is unconditionally stable (Wang and Rokhlin, 2001;Rokhlin and Wang, 2002).…”
Section: Outline Of the Approachmentioning
confidence: 99%
“…This new method will be called 'invariant embedding' based on the nomenclature in the review paper of Chin, Hedstrom & Thigpen (1984). We demonstrate the superior numerical stability and computational efficiency of the invariant embedding technique compared with the propagator formulation using a variety of numerical examples.…”
Section: Introductionmentioning
confidence: 97%