Prioritized NorESM2.5 twin simulations - July 2024 #7
Replies: 8 comments 18 replies
-
|
Code version for simulation ? |
Beta Was this translation helpful? Give feedback.
-
|
@oyvindseland - I need to create a new alpha version - noresm2_5_alpha03_v2. However I need to test this on betzy before I create it - and betzy is still not functional. As soon as betzy is working - I'll start the testing and let everyone know. |
Beta Was this translation helpful? Give feedback.
-
|
OK: I see Betzy is up but with quite limited access to the compute cores |
Beta Was this translation helpful? Give feedback.
-
|
How should we give the trigrid as a shortname in case name? |
Beta Was this translation helpful? Give feedback.
-
|
I have created a new alias ne30_f09_tn14 - but it will only be available in noresm2_3_alpha03_v2. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry, I wrote a 2 degree definition. Yes so ne30_f09_tn14 |
Beta Was this translation helpful? Give feedback.
-
|
@oyvindseland @adagj @matsbn - where should the cases for these simulations be create? |
Beta Was this translation helpful? Give feedback.
-
|
We already discussed this during the core meeting, and agreed that the cases should not be stored in a common directory. However, the case folders need to be archived together with the hist files and all output in the common directory /projects/NS9560K/noresm3/cases/ which I will make now |
Beta Was this translation helpful? Give feedback.
-
|
@oyvindseland @mvertens Can you please test if you have write access |
Beta Was this translation helpful? Give feedback.
-
|
I think a common case directory for the test simulations can be a bit crowded so I suggest to do the simulations on /cluster/projects/nn9560k/$user/ On nird however I think we should have the archive on a common directory. /projects/NS9560K/noresm2.5/cases/ ? |
Beta Was this translation helpful? Give feedback.
-
|
I'm referring to the dualgrid/trigrid fully coupled twin simulations that I will be doing. It specifies mam4 - but at this point I'm wondering if it should be oslo-aero - since our previous dualgrid/trigrid comparison was also oslo-aero but for an amip configuration. |
Beta Was this translation helpful? Give feedback.
-
|
I think that it is best to focus on oslo-aero simulations since the impact may be different on mam4 and analysing maa4 should not be our main target but for this I hope that Ada and Michael also add to the discussion. |
Beta Was this translation helpful? Give feedback.
-
|
I agree. The reason was that oslo aero was not possible to run with the most recent radiation code, at least that is what we thought. Since the default CAM configuration so far contains the "old" radiation code, I see no reason to use MAM4 as default. I can update the description of the twin simulations accordingly. However, the most important aspect of the twin simulations is to only change one thing at the time. |
Beta Was this translation helpful? Give feedback.
-
|
Let's use oslo aero for all the twin simulations. But will simulation 1 in set 1 and simulation 1 in set 3 be identical? Then we only need one of them |
Beta Was this translation helpful? Give feedback.
-
|
That may even be true for simulation 1 in set 3. Unless set 3 is the only one with DMS from ocean? |
Beta Was this translation helpful? Give feedback.
-
|
Separating out the discussion on the placement of the experiments. |
Beta Was this translation helpful? Give feedback.
-
|
I suggest to start a new discussion for this topic |
Beta Was this translation helpful? Give feedback.
-
|
I agree with this. |
Beta Was this translation helpful? Give feedback.
-
|
I think the reason to use MAM4 was that for the latest CAM version only MAM4 was working. But now that the new radiation is not any more an argument, I dont see why the MAM4 should be used throughout, if OSLO-Aer could be used instead. Are there other details of CAM which are newer when including MAM4? |
Beta Was this translation helpful? Give feedback.
-
|
I agree. The reason was that oslo aero was not possible to run with the most recent radiation code, at least that is what we thought. Since the default CAM configuration so far contains the "old" radiation code, I see no reason to use MAM4 as default. I can update the description of the twin simulations accordingly. However, the most important aspect of the twin simulations is to only change one thing at the time. |
Beta Was this translation helpful? Give feedback.
-
|
So for the dualgrid/trigrid comparison - both will use oslo-aero in a fully coupled configuration using the same compset. The only thing that will change is the grid. |
Beta Was this translation helpful? Give feedback.
-
|
I suggest in the future that we make a separate discussion for each topic under discussion. This thread has become messy. As an example, for this particular discussion thread, I thing it would be better to have separate discussion threads for: case name (which short name to use), common case folder or not, where to store the output. While if Oslo-Aero should be used as default or not belong in this thread. |
Beta Was this translation helpful? Give feedback.
-
|
The update on Betzy have run into trouble and may not be complete before end of next week. https://opslog.sigma2.no/ The capacity today are only 20 % so if it will be hard to get the job running this week if the situation does not improve soon. If the August update runs into similar type of problems I guess we should consider if we can do some development work on Fram |
Beta Was this translation helpful? Give feedback.
-
1st set: Trigrid vs. Dual grid fully coupled configurations
Simulation: Trigrid
Description: Test the impact of running the land component on FV grid instead of SE grid. This simulation will run atmosphere (SE) and land (FV) on different grids
Set up: N1850, fully coupled with Oslo-Aero, CAM on SE grid, CLM on FV grid, BLOM (hybrid) and CICE on tnx14
Duration: 10 years
Responsible: Mariana
Simulation: Dual grid
Description: Test the impact of running the land component on FV grid instead of SE grid. This simulation will run atmosphere and land on the same spectral element grid
Set up: N1850, fully coupled with Oslo-Aero, CAM and CLM on SE grid, CAM6 radiation, BLOM (hybrid) and CICE on tnx14
Duration: 10 years
Responsible: Mariana
2nd set: hybrid vs. isopycnic vertical coordinate in BLOM - fully coupled configurations
Simulation: Hybrid BLOM
Description: Test the impact of the new vertical grids in BLOM. This simulation will run BLOM with the hybrid coordinate grid
Set up: N1850, fully coupled with Oslo-Aero CAM on SE grid, CLM on FV grid, CAM6 radiation, BLOM (hybrid) and CICE on tnx14
Duration: 10 years
Responsible: Ada
Simulation : Isopycnic BLOM
Description: Test the impact of different vertical ocean grids. This simulation will run BLOM with the isopycnic vertical grid (used in CMIP6)
Set up: N1850, fully coupled with Oslo-Aero, CAM on SE grid, CLM on FV grid, CAM6 radiation, BLOM (isopycnic) and CICE on tnx14
Duration: 10 years
Responsible: Ada
3rd set: NorESM configuration with Oslo-Aero vs. Oslo-Aero-COARE - fully coupled configurations
Simulation: Oslo-Aero
Description: Test full NorESM configuration with trigrid, hybrid BLOM, OSLO-AERO
Set up: N1850, fully coupled with Oslo-Aero, CAM6 radiation, dms from ocean, CAM on SE grid, CLM on FV grid, BLOM (hybrid) and CICE on tnx14
Duration: 10 years
Responsible: Øyvind
Simulation: Oslo-Aero-COARE
Description: Test full NorESM configuration with trigrid, hybrid BLOM, and OSLO-AERO, and COARE flux parameterization
Set up: N1850, fully coupled with Oslo-Aero, CAM6 radiation, dms from ocean, COARE flux param., CAM on SE grid, CLM on FV grid, BLOM (hybrid) and CICE on tnx14
Duration: 10 years
Responsible: Øyvind
Beta Was this translation helpful? Give feedback.
All reactions