cam6_4_206: Moving-mountain GWD: MMTAUE/W/S/N diagnostics and atmos_phys bump for tilt/precip source options by JulioTBacmeister · Pull Request #1684 · ESCOMP/CAM

JulioTBacmeister · 2026-09-24T17:19:15Z

Summary

Adds directional (E/W/S/N) Reynolds-stress history diagnostics
(MMTAUE/MMTAUW/MMTAUS/MMTAUN) for the moving-mountain gravity wave
source, and bumps src/atmos_phys to pick up new opt-in flow-dependent
tilt/precipitation source formulations for that scheme.

This PR is blocked on, and must not be merged before, the companion
atmos_phys PR: ESCOMP/atmospheric_physics#448
The src/atmos_phys submodule pointer here currently points at the tip of
that PR's branch on my personal fork (a temporary placeholder) rather than
a real upstream commit, since that PR hasn't merged yet. It will be
updated to the real, tagged upstream commit once #448 merges.

gw_drag_cam.F90: wires the scheme's new taucd_west/east/south/north
outputs into 4 new history fields (MMTAUE/W/S/N, interface levels,
units Pa), named by direct analogy to the existing Beres BTAUE/W/S/N
diagnostics but with a plain MM prefix, manually added rather than
routed through the generic scheme-prefix infrastructure (per author
preference — the existing infra was judged more general than needed
here). Also adds TILT_MOVMTN/PSTEER_MOVMTN/PLAUNCH_MOVMTN
diagnostics and threads a new prect input to the scheme.
Also fixes several pre-existing outfld calls that passed ncol
instead of pcols as the array-bound argument (CAM convention);
affects only fill values in unused padding columns of those diagnostic
fields, not the prognostic state.
namelist_definition.xml: doc-string-only update for movmtn_source
describing the new valid values 3-5; default (1) and valid_values
unchanged.
No answer changes for any existing configuration: movmtn_source
defaults to 1, which exercises only the pre-existing, unchanged
vorticity_flux_src path; new outputs are write-only diagnostics.

AI involvement disclosure

This PR was produced through heavy, sustained collaboration with Claude
(Anthropic's Claude Code), across every stage: reconciling two weeks of
local development against current cam_development via 3-way merge,
diagnosing and fixing a merge-introduced compile error via an actual test
build, adding the new diagnostics, and drafting this PR/ChangeLog text.
All changes were reviewed and directed by the human author (Julio
Bacmeister) throughout, but the mechanical and drafting work was
substantially done by Claude, not just lightly assisted.

Test plan

Compiled successfully (QPC7 compset, ne3pg3_ne3pg3_mt232 grid, casper
machine): 0 warnings/errors in the modified files.
Full aux_cam regression suite (derecho/intel, derecho/nvhpc,
izumi/nag, izumi/gnu) has not yet been run against this branch —
opening as draft pending that and pending part of cam6_3_033: Increase default number of tasks on cheyenne for WACCMX #448.
Reconciled via git merge-file 3-way merge against the true upstream
base; confirmed the only real upstream drift was two cosmetic/
precision-suffix changes unrelated to this branch.

🤖 Generated with Claude Code

- Add local arrays and outfld/addfld calls for MMTAUE, MMTAUW, MMTAUS, MMTAUN: cardinal-direction Reynolds stresses for the moving-mountain gravity wave source, named by direct analogy to Beres's existing 'B'-prefixed outputs (manually named, not routed through the generic scheme-prefix infrastructure). - Wire the 4 new taucd_west/east/south/north intent(out) args from gravity_wave_drag_moving_mountain_run into gw_drag_cam.F90. - Update namelist_definition.xml doc string for movmtn_source to describe all 5 valid values (1=vorticity, 2=PBL momentum flux, 3=tilt-layer-mean, 4/5=tilt+precip fits); default and valid_values unchanged. - Bump src/atmos_phys submodule pointer to the branch tip on JulioTBacmeister/atmospheric_physics as a temporary placeholder, pending merge of the companion atmos_phys PR: ESCOMP/atmospheric_physics#448 Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

gw_drag_cam_init requested the TTEND_DP pbuf index whenever (use_gw_convect_dp .or. use_gw_movmtn_pbl), but convect_deep.F90 only ever registers that field when use_gw_convect_dp .and. deep_scheme=='ZM'. Running with the moving-mountain source on and use_gw_convect_dp=.false. therefore failed at init looking up a field that was never added to pbuf. gravity_wave_drag_moving_mountain.F90 confirms ttend_dp/netdt is only read in the use_gw_movmtn_pbl=.false. (legacy, non-PBL) branch of gw_movmtn_src; every currently-supported movmtn_source path requires use_gw_movmtn_pbl=.true. and never touches it. So the fix is simply to stop requesting the index unless the actual producer (deep convection GW) is enabled; the existing "if (ttend_dp_idx > 0)" fallback already zero-fills the array otherwise. Verified: case.build (0 warnings/errors) and a 5-day case.submit run with use_gw_convect_dp=.false. + use_gw_movmtn_pbl=.true. completed successfully (CaseStatus: "model execution success"), where it previously failed at init. Also bumps the src/atmos_phys submodule pointer to 2a5b2df (the comment-restoration commit already pushed to atmospheric_physics#448), which had been left stale after that push. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

… PRECT - Sum PREC_DP + PREC_SH + PREC_STR through the atmos_phys interstitial compute_total_precipitation_rate_run, the same routine CAM-SIMA runs, so the summation order matches between the two hosts. - PREC_STR replaces PREC_SED + PREC_PCW (PREC_STR = PREC_PCW + PREC_SED in physpkg). - Drop the errcode/index guards: all three fields always exist wherever gw_drag_cam runs (moist physics only), so a missing field should abort. Assisted-by: claude-opus:5.5

…ectional stress diagnostics (#448) ## Summary Adds flow-dependent tilt/precipitation source options to the moving-mountain gravity wave drag scheme, plus new cardinal-direction Reynolds-stress diagnostics. This is a companion PR to ESCOMP/CAM#1684, which adds the corresponding `MMTAUE/W/S/N` history outputs and namelist documentation. - New opt-in `movmtn_source` values 3-5 (tilt-layer-mean; tilt+precip 6-parameter fit; tilt+precip PySR cx17 fit). Existing values 1 (vorticity, default) and 2 (PBL momentum flux) are unchanged. - New diagnostic-only outputs threaded through as `intent(out)`: `taucd_west/east/south/north` (directional Reynolds stresses from `gw_common`'s `calc_taucd`), plus `tilt`, `p_steer`, `p_launch`. - New `prect` input, used only by the new precip-dependent source options. - None of the new pathways are reached unless a user explicitly sets `movmtn_source` to 3/4/5; default behavior and answers are unchanged. ## AI involvement disclosure This PR was produced through heavy, sustained collaboration with Claude (Anthropic's Claude Code), across every stage: reconciling two weeks of local development against current `main` via 3-way merge, diagnosing and fixing a merge-introduced compile error via an actual test build, adding the new diagnostic outputs, and drafting this PR/ChangeLog text. All changes were reviewed and directed by the human author (Julio Bacmeister) throughout, but the mechanical and drafting work was substantially done by Claude, not just lightly assisted. ## Test plan - Compiled successfully (QPC7 compset, `ne3pg3_ne3pg3_mt232` grid, casper machine): 0 warnings/errors in the modified files. - Full `aux_cam` regression suite (derecho/intel, derecho/nvhpc, izumi/nag, izumi/gnu) has **not** yet been run against this branch — opening as draft pending that. - No answer changes expected for any existing configuration, since `movmtn_source` defaults to 1. 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com> Co-authored-by: Haipeng Lin <hplin@ucar.edu>

jimmielin

Thanks @JulioTBacmeister

Ready for tagging.

JulioTBacmeister and others added 4 commits September 24, 2026 11:18

Merge branch 'cam_development' into movmtn-mmtau-diagnostics

cc385f9

jimmielin marked this pull request as ready for review September 30, 2026 19:53

peverwhee mentioned this pull request Sep 30, 2026

sima0_30_001: Moving-mountain GWD: tilt/precip source options and directional stress diagnostics ESCOMP/atmospheric_physics#448

Merged

cacraigucar added this to CAM Development Oct 1, 2026

jimmielin self-requested a review October 1, 2026 21:12

cacraigucar added the Must have CESM3 label Oct 1, 2026

cacraigucar self-assigned this Oct 1, 2026

jimmielin assigned jimmielin and unassigned jimmielin Oct 1, 2026

Update atmos_phys external to atmos_phys0_30_001

76c0e28

jimmielin approved these changes Oct 2, 2026

View reviewed changes

Merge branch 'cam_development' into movmtn-mmtau-diagnostics

ef028c6

jimmielin changed the title ~~Moving-mountain GWD: MMTAUE/W/S/N diagnostics and atmos_phys bump for tilt/precip source options~~ cam6_4_206: Moving-mountain GWD: MMTAUE/W/S/N diagnostics and atmos_phys bump for tilt/precip source options Oct 3, 2026

Finalize ChangeLog for cam6_4_206.

d82b442

jimmielin merged commit 6966333 into ESCOMP:cam_development Oct 4, 2026
2 checks passed

github-project-automation Bot moved this to Tag in CAM Development Oct 4, 2026

cacraigucar assigned jimmielin and unassigned cacraigucar Oct 6, 2026

Navigation Menu

Repository navigation

cam6_4_206: Moving-mountain GWD: MMTAUE/W/S/N diagnostics and atmos_phys bump for tilt/precip source options - #1684

cam6_4_206: Moving-mountain GWD: MMTAUE/W/S/N diagnostics and atmos_phys bump for tilt/precip source options#1684
jimmielin merged 7 commits into
ESCOMP:cam_developmentfrom
JulioTBacmeister:movmtn-mmtau-diagnostics

JulioTBacmeister commented Sep 24, 2026

Uh oh!

jimmielin left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

JulioTBacmeister commented Sep 24, 2026

Summary

AI involvement disclosure

Test plan

Uh oh!

jimmielin left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants