Comparing the nemo namelists
Here I will report on the differences between the namelists and then we can discuss what to do. cc: @kkeller @nrocha
1. Comparing IR with Development Resolution
I compared experiment/control-30y-1990-intermediate-configuration with experiment/control-30y-1990-development-configuration.
The relevant differences that I found are:
| namelist key | IR | dev |
|---|---|---|
| namsbc | ||
| ln_ssr | .false. | .true. |
| namlbc | ||
| ln_shlat2d | .false. | .true. |
| namtra_ldf | ||
| rn_Ud | 0.011 | 0.027 |
| namdyn_ldf | ||
| rn_Uv | 0.0838 | 0.1895 |
rn_Ud and rn_Uv are scale dependent parameters so it makes sense they are different. The other differences that I didn't list here are all related to differences in the forcing files and run parameters like the timestep. We will need to change the ln_ssr for the dev resolution and we need to discuss what to do with shlat2d.
IR-DEV_namelist_cfg_diff.xlsx
2. Comparing debug with development resolution
I compared experiment/control-30y-1990-debug-configuration with experiment/control-30y-1990-development-configuration. There are way too many differences in the namelist to list them here. I also don't know which of the differences are relevant at all. So, I've attached the xls file with all the differences.
LR-DEV_namelist_cfg_diff.xlsx
3. Comparing coupled namelist with ocean only namelist for eORCA12
I compared experiment/ocean-spinup-nemo-2020-eorca12-5y with experiment/control-30y-1990-development-configuration.
| namelist key | IR | dev |
|---|---|---|
| namsbc | ||
| ln_blk | .true. | .false. |
| ln_ssr | .false. | .true. |
| ln_wave | .false. | .true. |
| ln_sdw | .false. | .true. |
| ln_tauw | .false. | .true. |
| ln_stcor | .false. | .true. |
The only differences seem to be in the surface boundary condition module. That is expected. OceOnly-Coupled_eO12_namelist_cfg_diff.xlsx
4. Comparing the eORCA12 and eORCA025 ocean only namelists.
I compared experiment/ocean-spinup-nemo-2020-eorca12-5y with experiment/ocean-spinup-nemo-1990-eorca025-5y.
The differences are the same as for the coupled model.
PS: I used my compare_namelist tool from the experiment-scripts repo to compare the namelists.