updating unified MYNN-EDMF driver with packaged variables by joeolson42 · Pull Request #280 · ufs-community/MPAS-Model

joeolson42 · 2026-08-28T11:32:59Z

This PR brings in a unified (WRF/MPAS) driver with some advancements towards unification with CCPP. Additional changes include:

packaged 3D variables, which should help save memory usage when additional output/tke budgets are not activated.
additional CI tests for restarts (in MYNN-EDMF submodule)
additional CI tests for smoke/dust mixing (in MYNN-EDMF submodule)
use of hydrostatic pressure at interfaces, which should help provide more computational stability. total pressure is still brought in for saturation checks
MYNN-EDMF repository has been restructured to move all the code into a src directory and it now has a hot new logo

The testing was done in a commit prior to the stochastic physics, so the base already requires updating at our leisure.

Mandatory Questions

Does this PR include any additions or changes to external inputs (e.g., microphysics lookup tables, static data for gravity-wave drag -- things like that)?
- no
Does this PR require updating one or more baselines for the CI tests? If so, what?
- no

Reviews

Is this PR currently draft, still being worked on, or ready for review?
- The MYNN-EDMF, pbldriver, Makefile, and packages are ready for review, but updating the base is to come soon.
For when this PR begins review, please list the developers/collaborators you'd like to prioritize for review; e.g.,:
- @AndersJensen-NOAA
- @clark-evans

Copilot

Pull request overview

This PR updates MPAS atmosphere core plumbing to support a unified MYNN-EDMF driver integration by introducing new package flags for optional diagnostics (e.g., TKE budget and MF diagnostics), adjusting driver data movement accordingly, and updating build system paths to match the MYNN-EDMF repository’s new src/ layout.

Changes:

Added new MYNN-EDMF packages in Registry.xml to separate TKE-budget and diagnostic-output variables for conditional allocation/output.
Updated PBL package setup and PBL driver code paths to use the new config/package gating and revised MYNN-EDMF driver argument list (including hydrostatic interface pressure).
Updated Makefile/CMake source paths to reflect MYNN-EDMF code relocation into a src/ directory.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
src/core_atmosphere/Registry.xml	Adds new MYNN-EDMF diagnostic packages and updates related namelist/variable package metadata.
src/core_atmosphere/physics/mpas_atmphys_packages.F	Activates new MYNN-EDMF packages based on config flags during physics package setup.
src/core_atmosphere/physics/mpas_atmphys_driver_pbl.F	Gates MYNN-EDMF diagnostic pool transfers on config flags and updates the MYNN-EDMF driver call signature.
src/core_atmosphere/physics/Makefile	Updates build recipe and include paths for the MYNN-EDMF `src/` restructure.
src/core_atmosphere/CMakeLists.txt	Updates MYNN-EDMF source file paths to align with the new `src/` layout.

Suppressed comments (4)

src/core_atmosphere/physics/mpas_atmphys_driver_pbl.F:855

Same as above: copying EDMF diagnostics from MPAS pools is currently gated on config_mynn_edmf_output == 1, which will skip the copy for other nonzero modes described in Registry.xml.

       if(config_mynn_edmf_output == 1)then
         do j = jts,jte
         do k = kts,kte
         do i = its,ite
            edmfa_p(i,k,j)    = edmf_a(k,i)
            edmfent_p(i,k,j)  = edmf_ent(k,i)
            edmfqc_p(i,k,j)   = edmf_qc(k,i)
            edmfqt_p(i,k,j)   = edmf_qt(k,i)
            edmfthl_p(i,k,j)  = edmf_thl(k,i)
            edmfw_p(i,k,j)    = edmf_w(k,i)
            subthl_p(i,k,j)   = sub_thl(k,i)
            subqv_p(i,k,j)    = sub_qv(k,i)
            detthl_p(i,k,j)   = det_thl(k,i)
            detqv_p(i,k,j)    = det_qv(k,i)
         enddo
         enddo
         enddo
       endif

src/core_atmosphere/physics/mpas_atmphys_driver_pbl.F:1270

Same gating issue when copying EDMF diagnostics back into MPAS pools: config_mynn_edmf_output == 1 will skip updates for other nonzero modes described in Registry.xml.

       if(config_mynn_edmf_output == 1)then
          do j = jts,jte
          do k = kts,kte
          do i = its,ite
             edmf_a(k,i)     = edmfa_p(i,k,j)
             edmf_ent(k,i)   = edmfent_p(i,k,j)
             edmf_qc(k,i)    = edmfqc_p(i,k,j)
             edmf_qt(k,i)    = edmfqt_p(i,k,j)
             edmf_thl(k,i)   = edmfthl_p(i,k,j)
             edmf_w(k,i)     = edmfw_p(i,k,j)
             sub_thl(k,i)    = subthl_p(i,k,j)
             sub_qv(k,i)     = subqv_p(i,k,j)
             det_thl(k,i)    = detthl_p(i,k,j)
             det_qv(k,i)     = detqv_p(i,k,j)
          enddo
          enddo
          enddo
       endif

src/core_atmosphere/physics/mpas_atmphys_packages.F:244

config_mynn_edmf_output allows values 1 (updrafts) and 2 (downdrafts) per Registry.xml, but the package is only activated when the value is exactly 1. This prevents the output package from being enabled for the documented nonzero modes.

 elseif(config_pbl_scheme == 'bl_mynnedmf') then
    bl_mynnedmf_in = .true.
    if(config_mynn_edmf_output ==1) bl_mynnedmf_output_in = .true.
    if(config_mynn_tkebudget ==1) bl_mynnedmf_tkebudget_in = .true.

src/core_atmosphere/physics/mpas_atmphys_driver_pbl.F:1216

config_mynn_edmf_output supports nonzero modes beyond 1 per Registry.xml, but this guard only allocates/gets EDMF diagnostic arrays when the value is exactly 1. Use > 0 so the driver can retrieve arrays for all enabled output modes.

       if(config_mynn_edmf_output == 1) then
          call mpas_pool_get_array(diag_physics,'edmf_a'     ,edmf_a        )
          call mpas_pool_get_array(diag_physics,'edmf_ent'   ,edmf_ent      )
          call mpas_pool_get_array(diag_physics,'edmf_qc'    ,edmf_qc       )
          call mpas_pool_get_array(diag_physics,'edmf_qt'    ,edmf_qt       )
          call mpas_pool_get_array(diag_physics,'edmf_thl'   ,edmf_thl      )
          call mpas_pool_get_array(diag_physics,'edmf_w'     ,edmf_w        )
          call mpas_pool_get_array(diag_physics,'sub_thl'    ,sub_thl       )
          call mpas_pool_get_array(diag_physics,'sub_qv'     ,sub_qv        )
          call mpas_pool_get_array(diag_physics,'det_thl'    ,det_thl       )
          call mpas_pool_get_array(diag_physics,'det_qv'     ,det_qv        )
       endif

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

clark-evans · 2026-08-28T14:24:06Z

-       call mpas_pool_get_array(diag_physics,'det_thl'   ,det_thl   )
-       call mpas_pool_get_array(diag_physics,'det_qv'    ,det_qv    )
-
+       if(config_mynn_edmf_output == 1) then


@joeolson42 since 1=updrafts and 2=downdrafts now, I presume Copilot's recommendation is reasonable since these variables would seem to apply for both updrafts and downdrafts?

yes, this is a bug. Fixing with next commit.

clark-evans · 2026-08-28T14:26:36Z

-core_mynnedmf: core_physics_init core_physics_mmm
-	(cd physics_noaa/MYNN-EDMF; cp ./MPAS/Makefile .; cp ./MPAS/module_bl_mynnedmf_driver.F90 .; cp ./MPAS/module_bl_mynnedmf_common.F90 .; $(MAKE) all)
+core_mynnedmf: core_physics_init
+	(cd physics_noaa/MYNN-EDMF/src; ln -s ./MPAS/Makefile; ln -s ./MPAS/module_bl_mynnedmf_common.F90; $(MAKE) all COREDEF="$(COREDEF)")


@joeolson42 what is your thinking for changing from cp to ln here, particularly when all of the other physics (granted, those we've added) are set up to use cp?

I will take copilot's suggestion and use "ln -sf". I switched to linking from copying because I was sick of modifying the copied file for testing, then cleaning & recompiling (wiping out modified version), and then testing the old version. Commit fix in minutes.

clark-evans · 2026-08-28T14:42:21Z

    bl_mynn_in = .true.
 elseif(config_pbl_scheme == 'bl_mynnedmf') then
    bl_mynnedmf_in = .true.
+    if(config_mynn_edmf_output ==1) bl_mynnedmf_output_in = .true.
+    if(config_mynn_tkebudget ==1) bl_mynnedmf_tkebudget_in = .true.


This seasons reasonable to me, though I'm not sure if many/any of the other variables like this one have similar definitions.

yes, this fix is also coming.

clark-evans · 2026-08-28T14:41:42Z

       call mpas_pool_get_config(configs,'config_mynn_mixqt'      ,bl_mynn_mixqt      )
       call mpas_pool_get_config(configs,'config_mynn_tkeadvect'  ,bl_mynn_tkeadvect  )
       call mpas_pool_get_config(configs,'config_mynn_tkebudget'  ,bl_mynn_tkebudget  )
-       call mpas_pool_get_config(configs,'config_icloud_bl'       ,icloud_bl          )
       call mpas_pool_get_config(configs,'config_spp_pbl'         ,spp_pbl            )


@joeolson42 if I'm interpreting this correctly, I think Laura's MYNN-EDMF implementation uses icloud_bl, such that entirely removing its definition breaks that implementation. However, the CI comparison tests all completed successfully (though the compile tests are failing at the GNU autotools stage -- which shouldn't be affected by this PR), so 🤷‍♂️

icloud_bl is hard-coded to zero in mpas_atmphys_vars.F and it (and/or 'config_icloud_bl') does not exist as a namelist option in the Registry, so it needs to be removed. I think, in it's current form, this is a benign bug.

updating unified MYNN-EDMF driver with packaged variables

7026365

joeolson42 requested review from clark-evans and dustinswales as code owners August 28, 2026 11:33

github-project-automation Bot added this to MPAS-A PRs to process Aug 28, 2026

github-project-automation Bot moved this to Not assigned in MPAS-A PRs to process Aug 28, 2026

clark-evans moved this from Not assigned to In progress in MPAS-A PRs to process Aug 28, 2026

clark-evans assigned joeolson42 Aug 28, 2026

clark-evans requested review from AndersJensen-NOAA and a lite review from Copilot August 28, 2026 14:11

Copilot started reviewing on behalf of clark-evans August 28, 2026 14:11 View session

Copilot AI reviewed Aug 28, 2026

View reviewed changes

addressing copilots comments/suggestions

6b54b29

Navigation Menu

Uh oh!

updating unified MYNN-EDMF driver with packaged variables - #280

updating unified MYNN-EDMF driver with packaged variables#280
joeolson42 wants to merge 2 commits into
ufs-community:noaa/developfrom
joeolson42:mynn_20260822

joeolson42 commented Aug 28, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

clark-evans Aug 28, 2026

Uh oh!

joeolson42 Aug 30, 2026

Uh oh!

clark-evans Aug 28, 2026

Uh oh!

joeolson42 Aug 30, 2026

Uh oh!

clark-evans Aug 28, 2026

Uh oh!

joeolson42 Aug 30, 2026

Uh oh!

clark-evans Aug 28, 2026

Uh oh!

joeolson42 Aug 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

joeolson42 commented Aug 28, 2026

Mandatory Questions

Reviews

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants