b.e23_alpha16b.BLT1850.ne30_t025.045

[gustavo-marques]

Description:
The goal of this run is to check how increasing the horizontal resolution of the ocean/sea ice model affects the overall solution and, in particular, the state of the Labrador Sea. The run is based on a case (037) which freezes after ~ 30 years. However, unlike in case 037 where liquid and frozen runoff are spread using a 1000 km radius, here we are using a radius of 250 km instead (as in case 046).

Same as 037 but using an eddy-permitting (1/4 deg) configuration for the ocean and sea ice models. Below is a brief summary of the ocean setup:

• IC’s from WOA18;
• Vertical grid is z* with 65 layers;
• Mixed-layer eddies parameterization (Fox-Kemper et al., 2011);
• No mesoscale parameterizations. Biharmonic dissipation is the maximum of i) dynamic and ii) static contributions.

Sea ice ICs: restart from a corresponding G case at 0029-01-01.

---

Case directory:
Locally (if still available):
/glade/p/cesmdata/cseg/runs/cesm2_0/b.e23_alpha16b.BLT1850.ne30_t025.045

---

Sandbox:
Locally (if still available):

/glade/work/gmarques/cesm.sandboxes/cesm2_3_alpha16b_t025 (same as Cecile's cesm2_3_alpha16b but adding model_grid alias="ne30pg3_t025")

On github: https://github.com/ESCOMP/cesm/tree/cesm2_3_alpha16b

---

Diagnostics:
Diags (if available)

https://webext.cgd.ucar.edu/BLT1850/b.e23_alpha16b.BLT1850.ne30_t025.045/

---

Output:
Output (if still available):
/glade/scratch/gmarques/archive/b.e23_alpha16b.BLT1850.ne30_t025.045

---

Contacts:
@gustavo-marques
@cecilehannay
@dabail10

---

Extra details:

./xmlchange MASK_MESH=/glade/work/gmarques/cesm/tx1_4/mesh/tx1_4_mesh_221216_cdf5.nc
./xmlchange OCN_DOMAIN_MESH=/glade/work/gmarques/cesm/tx1_4/mesh/tx1_4_mesh_221216_cdf5.nc
./xmlchange ICE_DOMAIN_MESH=/glade/work/gmarques/cesm/tx1_4/mesh/tx1_4_mesh_221216_cdf5.nc
./xmlchange ROF2OCN_ICE_RMAPNAME=/glade/work/gmarques/cesm/tx1_4/runoff_mapping/map_r05_to_tx1_4_nnsm_e250r250_230914.nc
./xmlchange ROF2OCN_LIQ_RMAPNAME=/glade/work/gmarques/cesm/tx1_4/runoff_mapping/map_r05_to_tx1_4_nnsm_e250r250_230914.nc
./xmlchange GLC2OCN_ICE_RMAPNAME=/glade/work/gmarques/cesm/tx1_4/runoff_mapping/map_gland4km_to_tx1_4_nnsm_e250r250_230914.nc
./xmlchange GLC2OCN_LIQ_RMAPNAME=/glade/work/gmarques/cesm/tx1_4/runoff_mapping/map_gland4km_to_tx1_4_nnsm_e250r250_230914.nc
./xmlchange NTASKS_OCN=2592
./xmlchange NTASKS_CPL=1800
./xmlchange NTASKS_ATM=1800
./xmlchange NTASKS_LND=724
./xmlchange NTASKS_ICE=1080
./xmlchange NTASKS_ROF=724
./xmlchange NTASKS_GLC=36
./xmlchange NTASKS_WAV=36
./xmlchange ROOTPE_OCN=1800
./xmlchange ROOTPE_ICE=724

in user_nl_cice:

f_fresh = 'mxxxx'
f_fresh_ai = 'mxxxx'
ice_ic = "/glade/p/cesmdata/cseg/inputdata/cesm2_init/gmom.e23_b16.GJRAv4.TL319_t025_zstar_N65.t025_baseline.003/0029-01-01/gmom.e23_b16.GJRAv4.TL319_t025_zstar_N65.t025_baseline.003.cice.r.0029-01-01-00000.nc"

grid_file = "/glade/work/gmarques/cesm/tx1_4/seaice_files/horiz_grid_20221219.ieeer8"
grid_format = "bin"
grid_type = "tripole"
kcatbound = 0
kmt_file = "/glade/work/gmarques/cesm/tx1_4/seaice_files/topography_20221219.ieeei4"
ktherm = 2
update_ocn_f = .true.
frzpnd = "cesm"
prescribed_ice_mode = .false.
ns_boundary_type = "tripole"

[klindsay28]

FYI, in the initial submission of this experiment, the MOM6 output files have ne30_t232.037 in their filenames instead of ne30_t025.045.

[gustavo-marques]

Thanks for letting us know, Keith! I've fixed the diag_table and will rename the files once the current submission is completed.

[gustavo-marques]

This simulation has completed 20 years (finally).
@cecilehannay, @adamrher, @PeterHjortLauritzen: please run your diagnostics on this run. In particular, it would be good to see if we are also getting the E. Pacific dry bias in this simulation.
Thanks!

[klindsay28]

Something went wrong with the rho2 stream. There is a single mom6.h.rho2 file in $DOUT_S_ROOT/ocn/hist, and it doesn't have any data variables in it. Also, the time_bounds values are equal to 9.96920996838687e+36. This is causing my analysis scripts to fail.

Looking at the st_archive log files, it looks like every submission rewrote the file b.e23_alpha16b.BLT1850.ne30_t025.045.mom6.h.rho2.0001-01.nc.

I don't see a problem in the diag_table for the rho2 stream, the rho2 stream entries look like the entries for the other streams, but maybe I'm missing something.

Very odd.

[dabail10]

Sea ice diagnostics:

https://webext.cgd.ucar.edu/BLT1850/b.e23_alpha16b.BLT1850.ne30_t025.045/ice/b.e23_alpha16b.BLT1850.ne30_t025.045-obs/yrs1-20/

[gustavo-marques]

Thanks for reporting this, @klindsay28.
Although the rho2 stream is defined in the diag_table, MOM_input is missing the following:

NUM_DIAG_COORDS = 2
DIAG_COORDS = "z Z ZSTAR", "rho2 RHO2 RHO"
DIAG_COORD_DEF_RHO2 = "FILE:ocean_rho2_190917.nc,interfaces=rho2"

I will create an issue in MOM_interface reporting this. Apologies for the inconvenience in your analysis scripts.

[klindsay28]

It looks like this case has RESTORE_SALINITY = True, which complicates interpretation of the results in the Lab Sea. I found this because I noticed that global SSH is identically zero and I dug in to why this is so.

[cecilehannay]

ADF diags for high res ocean are here:
https://webext.cgd.ucar.edu/BLT1850/b.e23_alpha16b.BLT1850.ne30_t025.045/atm/
(@adamrher, @PeterHjortLauritzen, @JulioTBacmeister)

[adamrher]

As stated at the top of this thread, 37 is the baseline for this run (45), however 46 is the better comparison b/c it also reduces the liquid runoff radius.

SST bias (from CAM):

DJF Precip:

DJF Stddev Precip:

[gustavo-marques]

It looks like this case has RESTORE_SALINITY = True, which complicates interpretation of the results in the Lab Sea. I found this because I noticed that global SSH is identically zero and I dug in to why this is so.

Oh no! Well, we might have to invoke that quote from Francis Bretherton: "There are no mistakes in modeling, only unintended experiments".
Thanks for finding and reporting this issue!

[gustavo-marques]

Salinity restoring is turned off starting at year 22. I will keep this run going until Cheyenne is retired.
Thanks, Adam and Cecile for the diagnostics!