Thermo Error

[agustin@unex_es]

I have again an: ERROR: ice: Vertical thermo error problem. I am running compset: BHISTcmip6 --res f09_g17
I had this problem some time ago running the compset BSSP585cmip6 for a simulation of the XXI century. In that ocassion I solved the problem changing dt_count=48 in the pop namelist, setting clubb_gamma_coef = 0.30800001 in the atm namelist and changing the sentence "ferr > ferrmax" to " ferr > 1.1_dbl_kind*ferrmax" in the module ice_therm_vertical.F90 as sugested by Keith Lindsay and Michael Mills. But in this case all of these changes have failed and the model crash after 4 years 11 months of simulation. Though the model crash with a error in the CICE model, the error arise in the atmosphere because there appear to be too low temperatures:
zTsn: -100.647254043646
Tmin: -100.000000000000
Tmax: 2.696343900877508E-003
zqsn: -180443145.817767
zap_snow_temperature: temperature out of bounds!
istep1, my_task, i, j, k: 24720630 44 8 13
2
zTsn: -101.488219603677
Tmin: -100.000000000000
Tmax: 2.705124078207166E-003
zqsn: -181030727.697692
zap_snow_temperature: temperature out of bounds!
istep1, my_task, i, j, k: 24720630 44 8 13
3
zTsn: -101.393066112180
Tmin: -100.000000000000
Tmax: 2.704130619534671E-003
zqsn: -180964244.038821
Negative conc. in ch4tran. c,j,deficit (mol): 167053 1
2.409208359789274E-003

Starting thermo T < Tmin, layer 1
zTin = -100.100470277430
Tmin = -100.000000000000
istep1, my_task, i, j: 24720643 44 8 13
istep1, my_task, iblk = 24720643 44 2
category n = 2
Global block: 903
Global i and j: 67 348
Lat, Lon: 65.1595262563077 23.6830892204639
aice: 0.999999767194440
n: 2 aicen: 5.910550284331862E-003
ice: Vertical thermo error
ERROR: ice: Vertical thermo error

I have running with success a simulation similar to this with the compset --compset BHIST --res f09_g17 using CESM2.1.0 instead of CESM2.1.1 for the period 1980-2010.

Any help is wellcome.

Best regards

Agustin

 

[dbailey]

I think this might be an initialization issue. Not sure how CAM is producing temperatures like that. Here are some tips for debugging thermodynamic errors like this:

https://xenforo.cgd.ucar.edu/cesm/threads/faq-cice-thermodynamic-convergence-errors.4202/

 

[agustin@unex_es]

I think this might be an initialization issue. Not sure how CAM is producing temperatures like that. Here are some tips for debugging thermodynamic errors like this:

https://xenforo.cgd.ucar.edu/cesm/threads/faq-cice-thermodynamic-convergence-errors.4202/

Thank very much por you answer, I have seen the atm.log correspondig with this run and I saw that the minimum temperature went from 163 K to 106 K from step 87503 to step 80515:
.............
nstep, te 87503 0.26947064168226132E+10 0.26947070150460548E+10 0.33234323068923705E-04 0.98288046241679112E+05

240000 19841229
Total Mass= 982.880515394785 (mb), Dry Mass= 982.879999995254 (mb
)
Total Precipitable Water = 5.255875194731038E-003 (kg/m**2)
PS max = 1037.37006724897 min = 532.488734998079
U max = 191.259580937480 min = -43.6944485722194
V max = 44.3765245190135 min = -44.7249808773762
T max = 299.865799043514 min = 163.925132763825
W (mb/day) max = 1573.21220558405 min = -986.583403269169
Average Height (geopotential units) = 2678.65919360881
PRECC max = 0.000000000000000E+000 min = 0.000000000000000E+000
PRECL max = 33.1861203928287 min = 1.716389005750208E-004
Total precp= 1.07760148712647 CON= 0.000000000000000E+000 LS=
1.07760148712647
....
....
nstep, te 87515 0.26946086099859304E+10 0.26946092472893353E+10 0.35405411939030553E-04 0.98288045625232087E+05

60000 19841230
Total Mass= 982.880529988953 (mb), Dry Mass= 982.879999995254 (mb
)
Total Precipitable Water = 5.404701728500677E-003 (kg/m**2)
PS max = 1037.13699615100 min = 533.297100458602
U max = 191.779597687563 min = -45.1408047614560
V max = 43.2701921730363 min = -51.1236928583403
T max = 304.023161963603 min = 106.118330638998
W (mb/day) max = 1505.33904698085 min = -890.607087545187
Average Height (geopotential units) = 2678.63562868852
PRECC max = 0.000000000000000E+000 min = 0.000000000000000E+000
PRECL max = 37.4350351450922 min = 3.948918462407720E-004
Total precp= 1.09767064290164 CON= 0.000000000000000E+000 LS=
1.09767064290164

--------------------------
So in my opinion the problem seems to be in the atmosphere more than the ice or ocean. I have run a simulation similar to this (no cmip6) with compset --compset BHIST --res f09_g17 (CESM2.1.0) from 1980 to 2015, without any problem at all.

Thanks again