Checksum error and Iceberg born in land cell

[Soma Mishra]

I have compiled the model again with dynamic_nonsymmetric whatever you have suggested. Now, the warning is coming like:
Surface exchange size equals one.
diamonds, calve_icebergs: WARNING!!! Iceberg born in land cell 1137
556 -15.8750000000000 12.8898440186713
This are printing so many times so the PBS is going to crash. I have tried with one solution using this link: Checksum error following restart · Issue #296 · NOAA-GFDL/MOM6-examples whatever Niki Zadeh has suggested there for this same problem. Then again checksum error is coming.
What should I do in this regard?

 

[adcroft]

You might have two problems with the calving forcing field: i) it's non-zero over land, and ii) it's non-zero in the Tropics. Look at the forcing diagnostics in SIS2 too see if your forcing fields look right, and that calving and runoff line up with the ocean wet mask.

 

[Soma Mishra]

Dear Dr. Adcroft,
I have tried by the steps:
1. deleting the checksum

2. then changing the missing values according to:

cp ice_model.res.nc ice_model.res1.nc
#All ice restart variables should be 0 where they have a "missing value" (9.96E36)
for each var ( T_skin coszen rough_moist rough_heat rough_mom sym_str_s str_t str_d sym_v_ice_C sym_u_ice_C sal_ice enth_ice H_to_kg_m2 h_ice enth_snow h_snow h_pond flux_sw_nir_dif flux_sw_nir_dir flux_sw_vis_dif flux_sw_vis_dir p_surf calving_hflx runoff_hflx calving runoff fprec lprec flux_lw flux_salt flux_q flux_t flux_v flux_u )
ncap2 -s "where( $var > 1e36) $var =0.0" ice_model.res1.nc ice_model.res2.nc
mv -f ice_model.res2.nc ice_model.res1.nc
end
#except part_size that should be 1 for category 1 and 0 otherwise. But I don't know how to do that!
#The next best guess would be 1/6 for all categories so that they add up to 1:
ncap2 -s 'where( (part_size>1e36) ) part_size =1.0/6.0' ice_model.res1.nc ice_model.res2.nc
cp -f ice_model.res2.nc ice_model.res.nc

#Same story with calving.res.nc
cp -f calving.res.nc calving.res1.nc
ncap2 -s 'where( stored_ice>1e36 ) stored_ice =0.0' calving.res1.nc calving.res2.nc
mv -f calving.res2.nc calving.res1.nc
ncap2 -s 'where( stored_heat>1e36 ) stored_heat =0.0' calving.res1.nc calving.res2.nc
mv -f calving.res2.nc calving.res.nc
cd $work
endif

3. I have given 7 months run. Now the error is coming like:

WARNING from PE 667: Extreme surface sfc_state detected: i=1003 j= 503 lon= -49.375 lat= -0.250 x= -48.696 y= -0.250 D= 9.5000E+00 SSH= 1.1838E+00 SST= 5.5031E+01 SSS= 1.0000E-02 U-= 6.2047E-01 U+= 0.0000E+00 V-= 0.0000E+00 V+= 3.3339E-01


FATAL from PE 0: There were a total of 1 locations detected with extreme surface values!

Is this the right approach to proceed?

 

[Soma Mishra]

Dear Dr. Adcroft,
I am using JRA55-do as my forcing field. For licalvf and runoff I have done padding + regridding . I am not understanding

1. forcing diagnostics in SIS2

2. calving and runoff line up with the ocean wet mask

 

[adcroft]

Those changes you made were to the restart files. I was referring to the forcing. Icebergs are only created in a cell if there is a value of calving or frozen runoff in that cell that overflows a bucket. What you did was empty the buckets of ice in the restart, but there is possibly still a supply of ice into those buckets.You need to look at your forcing.

 

[adcroft]

The new error has excessive SST near the mouth of the Amazon river (SSS look fresh). Again, this could be coming from the forcing so I suggest diagnosing the forcing fields in both SIS2 and MOM6 to see what is driving the state to this extreme.