Rayleigh Friction to CCPP #1245

Katetc · 2025-02-04T00:57:01Z

I need to merge this one up too, but here's the PR for Rayleigh Friction to CCPP!

nusbaume

Looks good @Katetc! Just had one request with regards to the CAM subroutine names.

src/physics/cam/rayleigh_friction_cam.F90

cacraigucar

I took a quick peek back at this PR, and realized that it does not have the update to .gitmodules to bring in the atmospheric_physics PR which will come in along with this one

cacraigucar

I realized that I never did a full review, so I have done that now. @nusbaume - please chime in your thoughts on the _cam routines. I believe they add an unnecessary layer, but it might be a convention which are trying to implement?

cacraigucar · 2025-04-04T22:15:27Z

src/physics/cam/rayleigh_friction_cam.F90

+subroutine rayleigh_friction_cam_tend(                                     &
+     ztodt    ,state    ,ptend    )


I believe this routine does not need to be made and the call to rayleigh_friction_run can be made directly in physpkg. See the important note below on the individual elements in the call.

My proposed changes have been made assuming the elimination of this routine.

cacraigucar · 2025-04-04T22:20:16Z

src/physics/cam/rayleigh_friction_cam.F90

-      ptend%v(:ncol,k) = c1 * state%v(:ncol,k)
-      ptend%s(:ncol,k) = c3 * (state%u(:ncol,k)**2 + state%v(:ncol,k)**2)
-   enddo
+   call rayleigh_friction_run(ncol, pver, ztodt, state%u, state%v, ptend%u, ptend%v, ptend%s, errmsg, errflg)


I believe these arrays need to be subset by :ncol in the appropriate dimensions. For instance, state%u(:ncol,:), instead of just state%u. I'm surprised this is working without this subsetting.

I believe this works because the subsetting is done inside the repo, but I agree that it would be safest to do it at the calling-level as well!

I don't believe the CCPP level should be doing the subsetting. I've made that request in the atmospheric_physics PR and then this subsetting will be required. I'm indicating where this call to rayleigh_friction_run should go if this routine is eliminated.

cacraigucar · 2025-04-04T22:21:18Z

src/physics/cam/rayleigh_friction_cam.F90


-subroutine rayleigh_friction_init()
+subroutine rayleigh_friction_cam_init()


I would suggest that this routine should be eliminated and rayleigh_friction_init be called directly in the physpkg.F90 routines

My proposed changes have been made assuming the elimination of this routine.

nusbaume

Looks great to me now (assuming you also make @cacraigucar's changes, as I agree just calling the CCPP-ized subroutines directly is probably the best approach in this case)!

cacraigucar

I believe I have found all the lines that need to be changed, but I might have missed something. If you encounter problems with these changes, let me know and I can help you debug.

cacraigucar · 2025-04-08T22:25:42Z

src/physics/cam/rayleigh_friction_cam.F90

-      ptend%v(:ncol,k) = c1 * state%v(:ncol,k)
-      ptend%s(:ncol,k) = c3 * (state%u(:ncol,k)**2 + state%v(:ncol,k)**2)
-   enddo
+   call rayleigh_friction_run(ncol, pver, ztodt, state%u, state%v, ptend%u, ptend%v, ptend%s, errmsg, errflg)


I don't believe the CCPP level should be doing the subsetting. I've made that request in the atmospheric_physics PR and then this subsetting will be required. I'm indicating where this call to rayleigh_friction_run should go if this routine is eliminated.

cacraigucar · 2025-04-08T22:40:22Z

src/physics/cam/physpkg.F90

+    if (trim(cam_take_snapshot_before) == "rayleigh_friction_tend") then
+       call cam_snapshot_all_outfld_tphysac(cam_snapshot_before_num, state, tend, cam_in, cam_out, pbuf,&
+            fh2o, surfric, obklen, flx_heat)
+    end if


Add:
call physics_ptend_init(ptend, state%psetcols, 'rayleigh friction', ls=.true., lu=.true., lv=.true.)

! Initialize ptend variables to zero
!REMOVECAM - no longer need these when CAM is retired and pcols no longer exists
ptend%u(:,:) = 0._r8
ptend%v(:,:) = 0._r8
ptend%s(:,:) = 0._r8
!REMOVECAM_END

cacraigucar · 2025-04-08T22:41:11Z

src/physics/cam/physpkg.F90

+       call cam_snapshot_all_outfld_tphysac(cam_snapshot_before_num, state, tend, cam_in, cam_out, pbuf,&
+            fh2o, surfric, obklen, flx_heat)
+    end if
+    call rayleigh_friction_cam_tend( ztodt, state, ptend)


Call rayleigh_friction_run here.

call rayleigh_friction_run pver, ztodt, state%u(:ncol,:), state%v(:ncol,:), ptend%u(:ncol,:), ptend%v(:ncol,:), ptend%s(:ncol,:), errmsg, errflg)

Updated with appropriate subsetting

cacraigucar · 2025-04-08T22:44:26Z

src/physics/cam7/physpkg.F90

+       call cam_snapshot_all_outfld_tphysac(cam_snapshot_before_num, state, tend, cam_in, cam_out, pbuf, &
+            fh2o, surfric, obklen, flx_heat, cmfmc, dlf, det_s, det_ice, net_flx)
+    end if
+    call rayleigh_friction_tendO( ztodt, state, ptend)


Replace this line with:
call rayleigh_friction_run pver, ztodt, state%u(:ncol,:), state%v(:ncol,:), ptend%u(:ncol,:), ptend%v(:ncol,:), ptend%s(:ncol,:), errmsg, errflg)

I couldn't find the code for rayleigh_friction_tendO, but it exists, I would suggest it be removed

I changed the name to "_cam_tend" for Jesse's review, but forgot to update the cam7 physpkg. I have updated that file and tested both cam7 and camdev physics.

cacraigucar · 2025-04-08T22:45:17Z

src/physics/cam7/physpkg.F90

@@ -877,7 +877,7 @@ subroutine phys_init( phys_state, phys_tend, pbuf2d, cam_in, cam_out )

    call gw_init()

-    call rayleigh_friction_init()
+    call rayleigh_friction_initO()


Call rayleigh_friction_init directly.

If the routine rayleigh_friction_init0 exists, I suggest it be removed

I changed the name to "_cam_init" for Jesse's review but forgot about the camdev physpkg. It's all changed over to the ccpp code now.

cacraigucar · 2025-04-08T22:47:30Z

src/physics/cam7/physpkg.F90

+    if (trim(cam_take_snapshot_before) == "rayleigh_friction_tend") then
+       call cam_snapshot_all_outfld_tphysac(cam_snapshot_before_num, state, tend, cam_in, cam_out, pbuf, &
+            fh2o, surfric, obklen, flx_heat, cmfmc, dlf, det_s, det_ice, net_flx)
+    end if


Add:
call physics_ptend_init(ptend, state%psetcols, 'rayleigh friction', ls=.true., lu=.true., lv=.true.)

! Initialize ptend variables to zero
!REMOVECAM - no longer need these when CAM is retired and pcols no longer exists
ptend%u(:,:) = 0._r8
ptend%v(:,:) = 0._r8
ptend%s(:,:) = 0._r8
!REMOVECAM_END

Made these changes

cacraigucar · 2025-04-08T22:56:41Z

src/physics/cam/rayleigh_friction_cam.F90


-subroutine rayleigh_friction_init()
+subroutine rayleigh_friction_cam_init()


My proposed changes have been made assuming the elimination of this routine.

cacraigucar · 2025-04-08T22:56:59Z

src/physics/cam/rayleigh_friction_cam.F90

+subroutine rayleigh_friction_cam_tend(                                     &
+     ztodt    ,state    ,ptend    )


My proposed changes have been made assuming the elimination of this routine.

Katetc added 7 commits December 9, 2024 09:34

Code for RF snapshots

a15f13a

Give correct lable to snapshots

53dd7a4

Adding support for RF snapshots

a46b2f7

Merge branch 'cam_development' into RFccpp_updates_1224

5c2bf28

New rf interface to use ccpp code

58d3800

Change needed to get RF to run at all

a9ed66a

Changes for CAM7

a9c58fb

Katetc requested review from peverwhee and nusbaume February 4, 2025 00:57

Katetc self-assigned this Feb 4, 2025

Merge branch 'cam_development' into RFccpp_updates_1224

5580696

nusbaume added the ccpp-conversion label Feb 28, 2025

Merge up to cam6_4_072

5679374

nusbaume requested changes Mar 10, 2025

View reviewed changes

src/physics/cam/rayleigh_friction_cam.F90 Outdated Show resolved Hide resolved

cacraigucar requested changes Mar 24, 2025

View reviewed changes

Katetc added 2 commits April 2, 2025 11:38

Jesses last review suggestions were good ones

05246a7

Merge branch 'cam_development' into RFccpp_updates_1224

fa0a8bd

Katetc requested a review from nusbaume April 4, 2025 21:06

cacraigucar requested changes Apr 4, 2025

View reviewed changes

nusbaume approved these changes Apr 6, 2025

View reviewed changes

cacraigucar requested changes Apr 8, 2025

View reviewed changes

Katetc added 3 commits April 10, 2025 16:45

Changed to call raleigh_friction_run ccpp code directly from physpkg.F90

7c60852

Merge branch 'cam_development' into RFccpp_updates_1224

ef7628c

Removing cam wrappers for rf ccpp init and run proceedures

4f2e8c1

Katetc requested a review from cacraigucar April 11, 2025 22:37

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Rayleigh Friction to CCPP #1245

Rayleigh Friction to CCPP #1245

Katetc commented Feb 4, 2025

nusbaume left a comment

cacraigucar left a comment

cacraigucar left a comment

cacraigucar Apr 4, 2025

cacraigucar Apr 8, 2025

cacraigucar Apr 4, 2025

nusbaume Apr 6, 2025

cacraigucar Apr 8, 2025

cacraigucar Apr 4, 2025

cacraigucar Apr 8, 2025

nusbaume left a comment

cacraigucar left a comment

cacraigucar Apr 8, 2025

cacraigucar Apr 8, 2025

Katetc Apr 11, 2025

cacraigucar Apr 8, 2025

Katetc Apr 11, 2025

cacraigucar Apr 8, 2025

Katetc Apr 11, 2025

cacraigucar Apr 8, 2025

Katetc Apr 11, 2025

cacraigucar Apr 8, 2025

Katetc Apr 11, 2025

cacraigucar Apr 8, 2025

cacraigucar Apr 8, 2025

		subroutine rayleigh_friction_cam_tend( &
		ztodt ,state ,ptend )


		subroutine rayleigh_friction_init()
		subroutine rayleigh_friction_cam_init()

Rayleigh Friction to CCPP #1245

Are you sure you want to change the base?

Rayleigh Friction to CCPP #1245

Conversation

Katetc commented Feb 4, 2025

nusbaume left a comment

Choose a reason for hiding this comment

cacraigucar left a comment

Choose a reason for hiding this comment

cacraigucar 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

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nusbaume left a comment

Choose a reason for hiding this comment

cacraigucar 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

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

Choose a reason for hiding this comment

Choose a reason for hiding this comment