Cookbook for mantle convection model using tomography and other geophysical data. #5256

alarshi · 2023-07-10T22:42:11Z

This PR is a follow up from #PR 4727, which adds a cookbook to setup a mantle convection model using tomography and other geophysical data. I addressed most of the comments left on #PR 4727, but need help if the cookbook is clear enough and makes sense for a 2D spherical shell model. Thank you!

Before your first pull request:

I have read the guidelines in our CONTRIBUTING.md document.

For all pull requests:

I have followed the instructions for indenting my code.

For new features/models or changes of existing features:

I have tested my new feature locally to ensure it is correct.
I have created a testcase for the new feature/benchmark in the tests/ directory.
I have added a changelog entry in the doc/modules/changes directory that will inform other users of my change.

gassmoeller · 2023-07-12T23:45:32Z

/rebuild

gassmoeller

Just some first comments to let you get started, I still need to look into a lot of the files.

cookbooks/equilibrium_grain_size/doc/Fig_composition.png

cookbooks/equilibrium_grain_size/doc/tomography_based_model.md

cookbooks/equilibrium_grain_size/input_data/1D_reference_profiles/prem.txt

cookbooks/equilibrium_grain_size/plugins/CMakeLists.txt

gassmoeller · 2023-07-13T00:17:20Z

doc/sphinx/user/cookbooks/geophysical-setups.md

@@ -106,5 +106,6 @@ cookbooks/global_melt/doc/global_melt.md
 cookbooks/mid_ocean_ridge/doc/mid_ocean_ridge.md
 cookbooks/kinematically_driven_subduction_2d/doc/kinematically_driven_subduction_2d.md
 cookbooks/inclusions/doc/inclusions.md
+cookbooks/equilibrium_grain_size/doc/tomography_based_model.md


Here you name the section 'tomography based model', but the folder is called 'equilibrium grain size'. I think both are not very good names for this cookbook (tomography based model is quite unspecific, and the model has not much to do with grain size anymore). I would suggest: "Tomography based plate motions" or "tomography_based_plate_motions.prm. I know this is a lot of work, but please rename the folder and the file, and all references to the old names in all files (search and replace is your friend :-)).

Got it, thanks for suggestion!

alarshi · 2023-07-13T18:34:31Z

Thank you @gassmoeller for the review! I addressed your comments in the latest commits. We now have a different name for the model and I have added the changelog entry related to this PR.

gassmoeller

Some more comments. Still havent gotten to the main files.

cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.md

cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.part.prm

cookbooks/tomography_based_plate_motions/input_data/LLNL_model_cropped_cratons_faults.txt

cookbooks/tomography_based_plate_motions/input_data/upper_mantle_TM1_2D.txt

cookbooks/tomography_based_plate_motions/models/gypsum_model/grain-size-test.prm

cookbooks/tomography_based_plate_motions/models/tests/2D_slice_with_faults_and_cratons.prm

cookbooks/tomography_based_plate_motions/plugins/reference_profile.h

cookbooks/tomography_based_plate_motions/plugins/tomography_based_plate_motions.h

bangerth · 2023-10-06T17:15:58Z

There are two warnings from sphinx:

/home/docs/checkouts/readthedocs.org/user_builds/aspect-documentation/checkouts/5256/doc/sphinx/user/cookbooks/cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.md:51: WARNING: Include file '/home/docs/checkouts/readthedocs.org/user_builds/aspect-documentation/checkouts/5256/doc/sphinx/user/cookbooks/cookbooks/tomography_based_plate_motions/doc/tomography_based_model.part.prm' not found or reading it failed
/home/docs/checkouts/readthedocs.org/user_builds/aspect-documentation/checkouts/5256/doc/sphinx/user/cookbooks/geophysical-setups.md:93: WARNING: toctree contains reference to nonexisting document 'user/cookbooks/<<<<<<< HEAD'

cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.md

cookbooks/tomography_based_plate_motions/doc/input_tomography.part.txt

cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.md

cookbooks/tomography_based_plate_motions/models/2D_slice_with_faults_and_cratons.prm

bangerth · 2023-10-06T17:38:09Z

cookbooks/tomography_based_plate_motions/models/2D_slice_with_faults_and_cratons.prm

+# We use intial temperatures from an ASCII file but we also use adiabatic
+# boundary plugin such it returns very small temperatures, and on taking


Something is grammatically not quite right here.

I rephrased what I wanted to say, I hope it makes more sense now.

cookbooks/tomography_based_plate_motions/models/2D_slice_with_faults_and_cratons.prm

cookbooks/tomography_based_plate_motions/plugins/reference_profile.cc

cookbooks/tomography_based_plate_motions/plugins/reference_profile.h

bangerth

I have not looked in detail through the material model implementation, but assume that it is largely copied and pasted from other material models.

I think I've looked through everything else, though.

cookbooks/tomography_based_plate_motions/plugins/reference_profile.h

doc/sphinx/user/cookbooks/geophysical-setups.md

bangerth · 2023-10-06T19:17:30Z

cookbooks/tomography_based_plate_motions/plugins/reference_profile.cc

+      double uppermost_mantle_thickness = 0.0;
+      if (Plugins::plugin_type_matches<const MaterialModel::TomographyBasedPlateMotions<dim>>(this->get_material_model()))
+        {
+          const MaterialModel::TomographyBasedPlateMotions<dim> &material_model
+            = Plugins::get_plugin_as_type<const MaterialModel::TomographyBasedPlateMotions<dim>> (this->get_material_model());
+
+          uppermost_mantle_thickness = material_model.get_depth_of_base_of_uppermost_mantle();
+        }


You initialize uppermost_mantle_thickness to zero if that plugin is not found. Is that the correct value? Or are you using the 0.0 value to express that the value is uninitialized and hope that you never use the value below? It seems to me the former, right?

Yes, it it the former case. If we use densities from the material model (in this case, tomography_based_plate_motions) , we will simply use the densities from the material model until the uppermost_mantle_thickness depth. We initialize it to zero so that by default the density profile is from the input reference density profile.

bangerth · 2023-10-06T19:18:28Z

cookbooks/tomography_based_plate_motions/plugins/reference_profile.cc

+            }
+          else
+            {
+              // We only want to use the PREM densities in the part of the model that also


Is this a copy-paste error? I thought you're not using PREM, or did I midunderstand?

Not really, we are using PREM for defining our reference profile against the tomography. But, "reference densities" may be more appropriate.

cookbooks/tomography_based_plate_motions/plugins/tomography_based_plate_motions.h

cookbooks/tomography_based_plate_motions/plugins/tomography_based_plate_motions.cc

bangerth · 2023-10-06T19:35:34Z

cookbooks/tomography_based_plate_motions/plugins/tomography_based_plate_motions.cc

+      // The update() function updates the profile that stores the laterally averaged viscosity.
+      // This is needed to compute the viscosity in the material model (since the viscosities are rescaled
+      // so that the lateral average matches the reference profile). Since the viscosity depends on both
+      // temperature and velocity, we want to update the lateral average profile after each temperature
+      // and Stokes solve.
+      this->get_signals().post_stokes_solver.connect([&](const SimulatorAccess<dim> &,
+                                                         const unsigned int ,
+                                                         const unsigned int ,
+                                                         const SolverControl &,
+                                                         const SolverControl &)
+      {
+        this->update();
+      });
+
+      this->get_signals().post_advection_solver.connect([&](const SimulatorAccess<dim> &,
+                                                            const unsigned int ,
+                                                            const unsigned int ,
+                                                            const SolverControl &)
+      {
+        this->update();
+      });


If I understand it right, your model is instantaneous and so you never actually solve for the temperature. In that case, the comment is not quite right, nor is it necessary to connect to the post_advection_solver.

This is mostly for clarification to myself: You could update the comment and remove the second connect() call. But perhaps it is best to leave it in place anyway just in case someone wanted to extend our model to the time-dependent case?

cookbooks/tomography_based_plate_motions/plugins/tomography_based_plate_motions.cc

alarshi · 2023-10-07T00:12:38Z

Thank you for the review! I have addressed all your comments, except one, in the latest commits. I will discuss the remaining one with you tomorrow.

gassmoeller

Just a few typos. Afterwards ready to merge.

cookbooks/tomography_based_plate_motions/doc/tomography_based_plate_motions.md

cookbooks/tomography_based_plate_motions/models/2D_slice_with_faults_and_cratons.prm

alarshi · 2023-10-07T19:57:21Z

Thank you for the review! I addressed your comments in the latest commit and added the plugin in the tests/ directory. Hopefully, the test works now :)

gassmoeller · 2023-10-07T20:49:35Z

cookbooks/tomography_based_plate_motions/models/2D_slice_with_faults_and_cratons.prm

The tester is failing, because this .prm file is not in the base directory of the cookbook, but in the models/ subdirectory. You can either move the .prm file there, or create a custom rule for this cookbook in cookbooks/check.mk. I would prefer moving the file into the base folder of the cookbook (also check if you have to adjust the test .prm for that.

Add test data files; implement equilibrium grain size Add scripts from Dannberg et al, 2017 Cleanup and add main() Add compile command to repo Temporarily store reference result from python script for comparison with c++ Add constant grain size model model: remove nullspace add matrix-free.prm Make constant grain size models converge. Improve setup. fix equilibrium grain size and add a test file Add prem profile Becker viscosity model. parent f1fd292 author Arushi Saxena <asaxena1@login01.expanse.sdsc.edu> 1622820847 -0700 committer asaxena <saxena.arushi@ufl.edu> 1653082535 -0400 parent f1fd292 author Arushi Saxena <asaxena1@login01.expanse.sdsc.edu> 1622820847 -0700 committer asaxena <saxena.arushi@ufl.edu> 1653082518 -0400 parent f1fd292 author Arushi Saxena <asaxena1@login01.expanse.sdsc.edu> 1622820847 -0700 committer asaxena <saxena.arushi@ufl.edu> 1653082460 -0400 Added no surface rotation in nullspace. Moved files according to aspect data structure. Removed unnecessary files. fix a file while rebase. remove unnecessary files. Indentation fix. Addressed initial suggestions from the PR. Changes to remove reference profile in adiabatic conditions. Fix for using adiabatic conditions. Addressed most comments on the souce files. Fixed the unit style. Remove adiabatic conditions from reference profile. Changes with recent ASPECT version. Fix seg fault from composition manager. Parameter file and plugin changes consistent with recent ASPECT. Remaining suggestions from PR. Modifications in the initial temperature pointer, consistent with current main. fix indent. Add references and check if the cookbook documentation works. Indent changes. Add figures and documentation related files. Add the reference file. Fix a bib entry. Fix formatting of headings and small doc changes. Fix the path to a file.

Addressed remaining suggestions from the PR.

Address some comments in the PR. Address remaining suggestions.

Addressed comments from the PR. fix indent

Address comments.

Fix the test output.

gassmoeller

Thanks, good to go when the testers are done.

gassmoeller · 2023-10-08T03:53:41Z

The relevant testers passed, so let's get this merged. Congrats @alarshi!

bangerth · 2023-10-08T15:49:21Z

Yay, very nice!

alarshi force-pushed the material_model_branch branch 3 times, most recently from bdece88 to 288b84a Compare July 12, 2023 23:07

gassmoeller added the ready to test label Jul 12, 2023

gassmoeller reviewed Jul 13, 2023

View reviewed changes

alarshi force-pushed the material_model_branch branch from 38155f7 to 2cc2f18 Compare July 13, 2023 18:31

gassmoeller reviewed Jul 13, 2023

View reviewed changes

bangerth changed the title ~~Material model branch~~ Cookbook for mantle convection model using tomography and other geophysical data. Sep 29, 2023

alarshi force-pushed the material_model_branch branch from 68a2ce5 to 8bd1374 Compare September 29, 2023 18:57

bangerth reviewed Oct 6, 2023

View reviewed changes

gassmoeller reviewed Oct 7, 2023

View reviewed changes

gassmoeller and others added 11 commits October 7, 2023 20:01

Copy the grain size material model

e63dcf7

Fix errors in documentation.

5686b9d

fix indent

820c9cd

Addressed initial comments on the PR.

d47f5fe

Addressed remaining suggestions from the PR.

Fix indent.

3fca713

Addressed comments from the PR.

ea005a1

Address some comments in the PR. Address remaining suggestions.

Add a test case for the cookbook.

39b503d

Addressed comments from the PR. fix indent

Add the plugin in the tests.

e200a87

Address comments.

Fix the test.

fba5d83

Fix the test output.

remove a duplicate reference relative to main.

b7597de

alarshi force-pushed the material_model_branch branch from 7f3c94a to b7597de Compare October 8, 2023 00:11

gassmoeller added the ready to merge label Oct 8, 2023

gassmoeller approved these changes Oct 8, 2023

View reviewed changes

gassmoeller merged commit 4024c3d into geodynamics:main Oct 8, 2023
6 of 7 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Cookbook for mantle convection model using tomography and other geophysical data. #5256

Cookbook for mantle convection model using tomography and other geophysical data. #5256

alarshi commented Jul 10, 2023

gassmoeller commented Jul 12, 2023

gassmoeller left a comment

gassmoeller Jul 13, 2023

alarshi Jul 13, 2023

alarshi commented Jul 13, 2023

gassmoeller left a comment

bangerth commented Oct 6, 2023

bangerth Oct 6, 2023

alarshi Oct 6, 2023

bangerth left a comment

bangerth Oct 6, 2023

alarshi Oct 6, 2023

bangerth Oct 6, 2023

alarshi Oct 6, 2023

bangerth Oct 6, 2023

alarshi commented Oct 7, 2023

gassmoeller left a comment

alarshi commented Oct 7, 2023

gassmoeller Oct 7, 2023

gassmoeller left a comment

gassmoeller commented Oct 8, 2023

bangerth commented Oct 8, 2023

		# We use intial temperatures from an ASCII file but we also use adiabatic
		# boundary plugin such it returns very small temperatures, and on taking

Cookbook for mantle convection model using tomography and other geophysical data. #5256

Cookbook for mantle convection model using tomography and other geophysical data. #5256

Conversation

alarshi commented Jul 10, 2023

Before your first pull request:

For all pull requests:

For new features/models or changes of existing features:

gassmoeller commented Jul 12, 2023

gassmoeller left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

alarshi commented Jul 13, 2023

gassmoeller left a comment

Choose a reason for hiding this comment

bangerth commented Oct 6, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bangerth left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

alarshi commented Oct 7, 2023

gassmoeller left a comment

Choose a reason for hiding this comment

alarshi commented Oct 7, 2023

Choose a reason for hiding this comment

gassmoeller left a comment

Choose a reason for hiding this comment

gassmoeller commented Oct 8, 2023

bangerth commented Oct 8, 2023