[cig-commits] [commit] devel: commented out the call to meshfem3D_models_getatten_val(), which was making the code unreliable at least in the case of s362ani + attenuation (see https://github.com/geodynamics/specfem3d/issues/8 ) (b19fbb4)
cig_noreply at geodynamics.org
cig_noreply at geodynamics.org
Mon Apr 21 14:42:35 PDT 2014
Repository : ssh://geoshell/specfem3d_globe
On branch : devel
Link : https://github.com/geodynamics/specfem3d_globe/compare/9419f59565e2229712f8b5429d09a2dc53fcec62...573eae5c27d0d961432a410249e596fde6e3c0e9
>---------------------------------------------------------------
commit b19fbb4952b2586ec6c51e61683ef372956d0a55
Author: Dimitri Komatitsch <komatitsch at lma.cnrs-mrs.fr>
Date: Mon Apr 21 23:39:47 2014 +0200
commented out the call to meshfem3D_models_getatten_val(), which was making the code unreliable at least in the case of s362ani + attenuation (see https://github.com/geodynamics/specfem3d/issues/8 )
>---------------------------------------------------------------
b19fbb4952b2586ec6c51e61683ef372956d0a55
src/meshfem3D/get_model.f90 | 23 +++++++++++++++--------
src/meshfem3D/meshfem3D_models.f90 | 16 ++++++++++++++++
2 files changed, 31 insertions(+), 8 deletions(-)
diff --git a/src/meshfem3D/get_model.f90 b/src/meshfem3D/get_model.f90
index 14f8762..7b40d8e 100644
--- a/src/meshfem3D/get_model.f90
+++ b/src/meshfem3D/get_model.f90
@@ -174,16 +174,23 @@
call exit_mpi(myrank,'error get_model values')
endif
- !> Hejun
- ! New Attenuation assignment
- ! Define 3D and 1D Attenuation after moho stretch
- ! and before TOPOGRAPHY/ELLIPCITY
+ !> Hejun, New attenuation assignment
+ ! Define 3D and 1D attenuation after Moho stretch
+ ! and before TOPOGRAPHY / ELLIPTICITY
!
!note: only Qmu attenuation considered, Qkappa attenuation not used so far...
- if( ATTENUATION ) &
- call meshfem3D_models_getatten_val(idoubling,xmesh,ymesh,zmesh,r_prem, &
- tau_e,tau_s,T_c_source, &
- moho,Qmu,Qkappa,elem_in_crust)
+!! DK DK BUG
+!! DK DK BUG routine meshfem3D_models_getatten_val() is unsafe and for instance breaks s362ani + attenuation
+!! DK DK BUG (see the so-called mcmodel=medium bug at https://github.com/geodynamics/specfem3d/issues/8 )
+!! DK DK BUG and thus for now I comment it out here.
+!! DK DK BUG
+!! DK DK BUG It is also not clear to me if the if() statement below is sufficient or if the routine should
+!! DK DK BUG be called in the case of 3D Earth models only.
+!! DK DK BUG
+!! DK DK BUG if( ATTENUATION ) &
+!! DK DK BUG call meshfem3D_models_getatten_val(idoubling,xmesh,ymesh,zmesh,r_prem, &
+!! DK DK BUG tau_e,tau_s,T_c_source, &
+!! DK DK BUG moho,Qmu,Qkappa,elem_in_crust)
! define elastic parameters in the model
diff --git a/src/meshfem3D/meshfem3D_models.f90 b/src/meshfem3D/meshfem3D_models.f90
index d4e7444..d76c108 100644
--- a/src/meshfem3D/meshfem3D_models.f90
+++ b/src/meshfem3D/meshfem3D_models.f90
@@ -814,6 +814,13 @@
tau_e,tau_s,T_c_source, &
moho,Qmu,Qkappa,elem_in_crust)
+!! DK DK BUG
+!! DK DK BUG
+!! DK DK BUG routine meshfem3D_models_getatten_val() is unsafe and for instance breaks s362ani + attenuation
+!! DK DK BUG (see the so-called mcmodel=medium bug at https://github.com/geodynamics/specfem3d/issues/8 )
+!! DK DK BUG
+!! DK DK BUG
+
! sets attenuation values tau_e and Qmu for a given point
!
! note: only Qmu attenuation considered, Qkappa attenuation not used so far in solver...
@@ -841,6 +848,15 @@
double precision :: r_used
double precision, parameter :: rmoho_prem = 6371.d0 - 24.4d0
+!! DK DK BUG
+!! DK DK BUG
+!! DK DK BUG routine meshfem3D_models_getatten_val() is unsafe and for instance breaks s362ani + attenuation
+!! DK DK BUG (see the so-called mcmodel=medium bug at https://github.com/geodynamics/specfem3d/issues/8 )
+!! DK DK BUG and thus for now I add a stop statement here.
+!! DK DK BUG
+!! DK DK BUG
+ stop 'DK DK BUG routine meshfem3D_models_getatten_val() is unsafe and for instance breaks s362ani + attenuation, thus removed'
+
! initializes
tau_e(:) = 0.0d0
More information about the CIG-COMMITS
mailing list