More miscellaneous runtime parameter changes for beta05#219
Conversation
- Added new topographic edits and updated channel list. - Enabled COMPUTE_MONIN_OBUKHOV = True to improve the OBL seasonal cycle by limiting the OBL depth to no deeper than the Monin-Obukhov depth. - Updated background mixing parameters: - KV = 1.0E-6 - KD = 1.0E-7 - PRANDTL_BKGND = 10 - Configured MLE backup settings: - Set MLE_FRONT_LENGTH_FROM_FILE = True - Updated MLE_FL_FILE to "mle-lf-clim-tx2_3v2_20250108.nc" - Enabled Leithy parameterization with: - USE_LEITHY = True - LEITH_BI_CONST = 90.0 - LEITHY_CK = 1.0 (full backscatter) - LEITH_AH = False
|
I will add the new files to input_data_list.yaml later today. |
|
I've added MINIMUM_OBL_DEPTH = 2.5 for cases using tx2_3v2. This solves the issue with the GNU debug test reported in NCAR/MOM6#331 |
|
Thanks, @gustavo-marques. I'll re-run the test suite. |
alperaltuntas
left a comment
There was a problem hiding this comment.
Dimensional scaling test for the L dimension fails when this PR gets merged. I am investigating.
|
@gustavo-marques Is this ready for testing now? |
|
Yes, thanks! |
|
With the latest beta05 update, we are getting negative Laplacian viscosities near coastlines. @iangrooms thinks this is a bug; fixing it will take a few days (perhaps more). To avoid additional delays in tag creation, we suggest going back to I am also reverting to using a constant frontal length scale for MLE. As Gokhan pointed out yesterday, the MLE overturning with the variable frontal scale is too strong, reverting the near-surface residual MOC in the Southern Ocean (see plot below). |
|
I updated the initial description by removing the following:
|

MINIMUM_OBL_DEPTH = 2.5.Fixes: #218, #213, #208, #220