search

COSIMA / access-om3

ACCESS-OM3 global ocean-sea ice-wave coupled model
13 stars 7 forks source link

Differences between MOM6-CICE6 1deg_jra55do_iaf and 1deg_jra55do_ryf configurations #171

Open ezhilsabareesh8 opened 6 months ago

ezhilsabareesh8 commented 6 months ago

There is a significant difference between the MOM_input of the IAF and RYF configurations, which may be causing the error in RYF but not in IAF. The IAF MOM_input is outdated and needs to be updated. Refer here.

<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

variable | MOM_input_one_deg_RYF | MOM_input_one_deg_IAF -- | -- | -- adjust_net_srestore_to_zero |   | TRUE ah_vel_scale |   | 0 bbl_use_eos |   | TRUE bt_thick_scheme |   | FROM_BT_CONT cfc_bc_file |   | cfc_atm_20230310.nc coord_config |   | none debug |   | FALSE default_2018_answers |   | FALSE depth_scaled_khth |   | FALSE energysavedays |   | 1 eqn_of_state |   | WRIGHT fatal_unused_params | TRUE |   fix_ustar_gustless_bug |   | TRUE gill_equatorial_ld |   | TRUE grid_rotation_angle_bugs |   | FALSE hmix_min |   | 2 int_tide_decay_scale |   | 300.3003003003003 interp_type2 |   | LMD94 interpolate_res_fn |   | FALSE kappa_shear_all_layer_tke_bug |   | FALSE kappa_shear_iter_bug |   | FALSE kdml |   | 0 kh_vel_scale |   | 0 khth |   | 0 khth_max |   | 0 khtr_max |   | 0 mask_srestore_under_ice |   | FALSE max_ent_it |   | 20 max_rino_it |   | 25 maxtrunc |   | 0 min_salinity |   | 0 nihalo |   | 4 njhalo |   | 4 prandtl_turb |   | 1 remap_uv_using_old_alg |   | FALSE simple_tke_to_kd |   | TRUE smag_bi_const |   | 0.06 tolerance_ent |   | 1e-05 topo_file |   | topog.nc use_cfc_cap |   | FALSE use_contemp_abssal |   | FALSE use_gm_work_bug |   | FALSE use_land_mask_for_hvisc |   | TRUE use_psurf_in_eos |   | TRUE visc_res_scale_coef |   | 0.4 z_init_file_salt_var |   | salt z_init_remap_old_alg |   | FALSE

minghangli-uni commented 6 months ago

The parameters in MOM_input_one_deg_RYF were updated by removing those that were set to their default values. Essentially, the remaining parameters are the same as MOM_input_one_deg_IAF.

aekiss commented 6 months ago

1deg RYF MOM6-CICE6 was used as the development config to establish suitable parameters to then be copied to the other configs.

fatal_unused_params is probably the main difference, but when https://github.com/COSIMA/access-om3/issues/117 is done we'll have a better way to track and confirm this.

Finalisation of parameters will require longer test runs, to be done after we have decided on efficient core counts and improved model stability.

ezhilsabareesh8 commented 6 months ago

Thanks, @aekiss and @minghangli-uni. I've noticed that some parameters are not included in the RYF configuration. For example, hmix_min = 2 is missing, with the default value being zero.

I am trying to determine what differences are causing the following crash in RYF after 5 months, but not in IAF. This issue occurs while calculating the remaining fraction of a layer's momentum after a time-step of viscosity in this routine.

Image              PC                Routine            Line        Source
libpthread-2.28.s  000014ED69E89CF0  Unknown               Unknown  Unknown
access-om3-MOM6-C  00000000037CB97F  mom_vert_friction        2065  MOM_vert_friction.F90
access-om3-MOM6-C  00000000037C6E34  mom_vert_friction        1721  MOM_vert_friction.F90
access-om3-MOM6-C  0000000003597CC4  mom_dynamics_spli         581  MOM_dynamics_split_RK2.F90
access-om3-MOM6-C  0000000002E49A33  mom_mp_step_mom_d        1173  MOM.F90
access-om3-MOM6-C  0000000002E4058B  mom_mp_step_mom_          853  MOM.F90
access-om3-MOM6-C  0000000002E1496D  mom_ocean_model_n         633  mom_ocean_model_nuopc.F90
access-om3-MOM6-C  0000000002D3505D  mom_cap_mod_mp_mo        1759  mom_cap.F90
aekiss commented 6 months ago

We decided to use HMIX_MIN=2: https://forum.access-hive.org.au/t/namelist-configuration-discussion-meeting/1917/9 - this was mistakenly missing in RYF due to https://github.com/COSIMA/access-om3/issues/121

To distinguish whether the RYF forcing or the RYF parameters are to blame, have you tried running RYF with the IAF parameters?