[CIG-SEISMO] SPECFEM3D - determining the effectiveness of absorbing boundaries

[Dimitri Komatitsch]

Hi Shahar,

You should switch to CPML; Stacey-Clayton-Enquist is obsolete.
Later today I'll send you a draft that my group has just submitted about 
CPML for forward and adjoint problems, in which we compare sensitivity 
kernels based on CPML and based on Stacey and show that the CPML version 
is much cleaner.

Best regards,
Dimitri.

On 11/27/2013 09:39 PM, Shahar Barak wrote:
> Hi,
>
> I understand from the references in the manual that the absorbing
> boundaries in SPECFEM3D is not perfect. For just using the absorbing
> boundaries in the code ( ABSORBING_CONDITIONS Set to .true. to turn on
> Clayton-Enquist absorbing boundary conditions ), i.e., without adding
> special constraints on the edges such as larger attenuation at the
> edges, are there any quantitatively guidelines/suggestions on how far
> away should the mesh edges be, given a source type and moment
> magnitude?
> Also, is the effectiveness of absorbing boundaries scalable? i.e., if
> I run tests on a small source within a small volume and find the size
> of volume needed in order to eliminate boundary returns, would that
> scale by a constant to a larger mesh and a larger source, keeping the
> size of mesh elements with respect to the size of the volume the same?
> Has anyone done such tests for SPECFEM3D? are there publications?
>
> Thanks,
>
> Shahar Barak
> =======================================
> Crustal Geophysics Research Group
> Department of Geophysics, Stanford University
> 397 Panama Mall, Stanford CA 94305-2215
> Mitchell Building, room 463
> Office: (650) 725 8229
> Cell: (650) 276 6169
> shaharb at stanford.edu
> _______________________________________________
> CIG-SEISMO mailing list
> CIG-SEISMO at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo
>

-- 
Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France    http://komatitsch.free.fr