Skip to content
Snippets Groups Projects
Normal view History Permalink
README 2.5 KiB
Newer Older
Guillaume Samson's avatar
Initial commit
Guillaume Samson committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
WARNING: For now, the test case ICE_ADV2D only works if the logical "ll_neg" is set to FALSE in the routine icedyn_adv_umx.F90
                  (it is still unclear why)
-------
Purpose
-------
This demonstration case can serve different purposes:

1) The main one: study of the advection of a patch of ice in a bi-periodic channel with a slab ocean (i.e. 1 ocean layer)
              ==> set ln_icethd=false in namelist_ice_cfg

2) Secondary: study of ice thermodynamics in the same basin
              ==> set ln_icedyn=false in namelist_ice_cfg
              ==> eventually change usrdef_sbc.F90 in MY_SRC to fit your needs

-----------
Experiments
-----------
Two experiments can be configured:

a) A simple channel at 3km horizontal resolution (slightly variable => +- 10%)
              ==> in MY_SRC/usrdef_hgr.F90, uncomment the part "variable scale factors" and comment part "constant scale factors"

b) A simple channel at 3km horizontal resolution (constant scale factors)
              ==> in MY_SRC/usrdef_hgr.F90, comment the part "variable scale factors" and uncomment part "constant scale factors"

----------
How to run
----------

a) Compile and run the model once to get a mesh_mask.nc file with the following command:
Guillaume S's avatar
Revert "Merge branch '22-rheology_diags' into 'main'"  
Guillaume S committed
31 32 
../../../makenemo -r ICE_ADV2D -n ICE_ADV2D -m X64_ADA -j 4
mpirun ./nemo -np 1
Guillaume Samson's avatar
Initial commit
Guillaume Samson committed
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 

b) Create the initial condition file for sea-ice (initice.nc) by running this python script: 
python ./make_INITICE.py

c) Run the model a second time
mpirun ./nemo -np 4

---------------
What to look at
---------------
In case of purpose 1, One can test
---
a) the advection scheme: Ultimate-Macho (ln_adv_UMx=T) versus Prather (ln_adv_Pra=T)
   for a square (ice concentration) or a gaussian (ice volume)
      with either a constant velocity (ln_dynADV2D=T)
               or a constant ice-atm. stress, thus velocity is calculated by rheology (ln_dynRHGADV=T)
      with 1 or 5 ice categories (jpl=1 or 5).
         (note that ln_dynADV2D=T only works with jpl=1)

In case of purpose 2, one can test conservation of properties:
---
b) ice should not change at all if surface fluxes = 0 and SST = freezing temperature

--------------------------------------
Interpretation of the results (remarks)
--------------------------------------
- Prather conserves the max values but also creates side lobes
- UM does not conserve the max but does not create side lobes
- The "unmoving" ice (if any) is due to ice rheology which states that ice mass below a certain thresold (1kg/m2) is considered to move at the ocean velocity (thus 0 m/s)