Improves buoyancy gradient calculation in GM#633
Improves buoyancy gradient calculation in GM#633mark-petersen merged 2 commits intoMPAS-Dev:ocean/developfrom
Conversation
changed horizontal buoyancy gradient in GM routine to be at fixed depth not at a fixed model index.
|
@mark-petersen I noticed when I made changes to GM I didn't correctly compute the horizontal buoyancy gradient. It currently assumes layer thickness is constant horizontally which is a poor assumption in places. This PR fixes this, but please do check my math for the chain rule. This one will be pretty critical for water cycle, so if you can work on this as soon as you have a chance I'd really appreciate it. |
|
Passes MPAS-Ocean nightly regression suite on grizzly with gnu compiler. I'm also testing in E3SM, will post MPAS-Analysis diagnostics tomorrow after 10 years complete. |
|
Here are some testing results All JAS MLD -- first the G-case baseline (6030v3 grid, 60 layer, GM pre this PR)
This is what I saw with Jigsaw 60-30 + 64 layers pre this PR and #608
This is the same as previous, but with the changes in this PR + those in #608
Finally here is the same but 60 layers, the sensitivity between 60 and 64 layers is again pretty small (although still a bit larger than I remember for v1)
|
|
@maltrud just wanted to alert you to this PR, it will. likely impact your WC14 simulations |
|
links to full analysis Baseline case -- https://portal.nersc.gov/project/e3sm/lvroekel/20200625_Gbaseline_GMPAS-IAF.T62_oECv3_anvil_yr_1_5/ |
|
I realize I had an issue and plotted an inconsistent figure for the old case. Unfortunately the new one doesn't look as different from the 64 layer with these changes. There still seems to be some issue with the 64 layer case. I will continue the 64 layer case to test further. |
|
Test merge passed:
|
|
@mark-petersen is there anything I can do to help move this PR along? It's needed for upcoming water cycle testing |
|
Tested with most recent commit. Passes nightly regression suite for gnu, intel17, intel19 for debug and optimized. |
Update mpas-source: improve GM drhodx This PR brings in a new mpas-source submodule with changes only to the ocean core. Currently the GM horizontal buoyancy gradient is computed at a fixed vertical index. At a number of places in the ocean this is problematic where layer thickness changes quickly between cells (especially near the bottom for partial bottom cells). See MPAS-Dev/MPAS-Model#633 [non-BFB]
mark-petersen
left a comment
mark-petersen
left a comment
There was a problem hiding this comment.
Approved, from MPAS testing, visual inspection, and Luke's E3SM tests.
Update mpas-source: improve GM drhodx This PR brings in a new mpas-source submodule with changes only to the ocean core. Currently the GM horizontal buoyancy gradient is computed at a fixed vertical index. At a number of places in the ocean this is problematic where layer thickness changes quickly between cells (especially near the bottom for partial bottom cells). See MPAS-Dev/MPAS-Model#633 [non-BFB]
4 participants
Currently the GM horizontal buoyancy gradient is computed at a fixed vertical index. At a number of places in the ocean this is problematic where layer thickness changes quickly between cells (especially near the bottom for partial bottom cells).