[CIG-SEISMO] Results difference from rev 18194 to rev 20451 for a simple configuration

[Camille Mazoyer]

Hello all,

Last year my collegues and I tried SPECFEM3D (rev 18194, April 2011) for a simple configuration. 
This week, we wanted to continue our work. I downloaded rev 20451 to run again my configuration. I was very surprised because the results are very differents. I don't know if I made a mistake in my data files, or if a bug were corrected in between times.

There are very few things to change in the data files between rev 18194 and rev 20451 (can you confirm?).
Here are the only modifications (the variables) I added:

In Par_file:
MODEL = default

In Mesh_Par_file
CREATE_VTK_FILES = .false.    

I send you the plot I had with gnuplot, for two stations named in my example K400 and K401 (K400.CE.CXZ.semd, K401.CE.CXZ.semd).
You can see results are quite differents, eg for a station that I called K400:
rev 18194: http://pageperso.univ-brest.fr/~mazoyer/guennou/specfem/rev18194/K400_rev18194.jpg
rev 20451: http://pageperso.univ-brest.fr/~mazoyer/guennou/specfem/rev20451/K400_%20rev20451.jpg

I send you the data files in case you want to have to take a look on them, or rerun the config. 
Here is the link where you can download all: http://pageperso.univ-brest.fr/~mazoyer/guennou/specfem
 
Any help would be very appreciate,
Thanks a lot,

Camille Mazoyer

-- 
Camille Mazoyer-Kervazo
UMS3113
Institut Universitaire Européen de la Mer
Rue Dumont d'Urville
29280 Plouzane

Tél Brest : 02 98 49 88 09, Bureau B116
Tél Toulon: 04 94 14 25 26
Tél Portable: 06 63 15 13 41

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/cig-seismo/attachments/20120709/d06153f5/attachment.htm