revisits source time functions, source angle and acoustic term sign conventions; adds axisymmetric monopole source factor

[danielpeter]

No description provided.

[codecov]

Codecov Report

Attention: Patch coverage is 52.87356% with 41 lines in your changes missing coverage. Please review.

Project coverage is 53.12%. Comparing base (138fa18) to head (b372dfe).
Report is 16 commits behind head on devel.

Files with missing lines	Patch %	Lines
src/specfem2D/noise_tomography.f90	41.37%	17 Missing ⚠️
src/specfem2D/prepare_initial_field.F90	20.00%	8 Missing ⚠️
src/specfem2D/compute_arrays_source.f90	25.00%	6 Missing ⚠️
src/specfem2D/axisymmetric_routines.f90	0.00%	5 Missing ⚠️
.../specfem2D/compute_add_sources_electromagnetic.f90	0.00%	3 Missing ⚠️
src/shared/read_source_file.f90	33.33%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##            devel    #1259   +/-   ##
=======================================
  Coverage   53.12%   53.12%           
=======================================
  Files         171      171           
  Lines       31877    31902   +25     
=======================================
+ Hits        16935    16949   +14     
- Misses      14942    14953   +11     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[danielpeter]

@Rohit-Kakodkar @icui @lsawade - acoustic source term & source time function changes (to be more consistent with 3D Cartesian)

[lsawade]

@danielpeter,
For me to recap:

• the source time function definitions remain unchanged
• prepare stf removes sign change
• acoustic force angle is computed from positive vertical clockwise
• compute_add_viscoelastic for SH remains unchanged resulting in net negative sign?

correct?

I know we have discussed this, but that means the definition of a gaussian in the code is inverted, which -- to me -- is weird. But I guess the point here is to keep the "natural" Ricker signature and integrations of that.

[danielpeter]

probably correct - you should have asked 2 months ago :)

and agree, there is a rather unusual Gaussian in the 2D version due to this USE_TRICK_FOR_BETTER_PRESSURE feature. note that in 2D you can now have a "normalized" Gaussian (comp_source_time_function_Gaussian_norm; setting time_function_type == 0 in DATA/SOURCE file) which is the typical Gaussian (positive) with a normalization factor, as used also in the 3D version. and then there is this unusual Gaussian, as the second integral of the Ricker wavelet which has negative values. the 2D version used that unusual Gaussian as the default Gaussian. the 3D version used the normalized, more typical Gaussian as default. well, i guess there are plenty of inconsistencies between 2D and 3D versions... you chose whatever default you think is more suitable.

note, if you get rid of that USE_TRICK_FOR_BETTER_PRESSURE feature, then there is no need for the unusual Gaussian.