cam3.1 slab ocean model

[wei_dong888@163_com]

Has anybody experience in running CAM with slab ocean model(T42)?

I created the boundary dataset as described in user_guide.but,when i run it ,has the error.the run log report as thus (the last fraction):

+1443 SOMINI: Read som data for dates 816 43200 and 916 0
+1444 INITEXT: ocnfrac= 0.000000000000000000E+00
+1445 nstep, te 0 3340167301.58742952 0.000000000000000000E+00 0.000000000000000000E+00 98466.2417923963367
+1446 icefrac shcoverage= 16.2616 million km^2
+1447 icefrac nhcoverage= 5.7440 million km^2
+1448 icevol shcoverage= 1.6262 10^13m^3
+1449 icevol nhcoverage= 1.1488 10^13m^3
+1450 snowvol shcoverage= 0.2420 10^13m^3
+1451 snowvol nhcoverage= 0.0031 10^13m^3
+1452 QNEG4 WARNING from TPHYSAC , lchnk = 3; Max possible LH flx exceeded at
1 points. Worst excess = -3.0040E-06 at i = 1
+1453 QNEG4 WARNING from TPHYSAC , lchnk = 3; Max possible LH flx exceeded at
2 points. Worst excess = -1.1538E-05 at i = 121
+1454 QNEG4 WARNING from TPHYSAC , lchnk = 3; Max possible LH flx exceeded at
1 points. Worst excess = -3.6576E-06 at i = 113
+1455 QNEG4 WARNING from TPHYSAC , lchnk = 3; Max possible LH flx exceeded at
1 points. Worst excess = -3.2043E-06 at i = 114
+1456 QNEG4 WARNING from TPHYSAC , lchnk = 3; Max possible LH flx exceeded at
2 points
------------------------------------------------------------------------
I think it may be something wrong in making the SST file for SOM. I read the "README" file in the "cam1modelsatmcamtoolsdefinemld" .it said :

"definemld1x1 -m mldfile.nc -v
(build 1x1 netcdf file)

definemldbdy -s 10 -v -i mldfile.nc -o pcmdi_sst_cam_bcT42_clim.032000.nc
(put that data on a T42 grid)"

in the setp 2,i could not find the sst file "pcmdi_sst_cam_bcT42_clim.032000.nc",so i use the SST file of model climate SST" sst_HadOIBl_bc_64x128_clim_c020411.nc" .Is it something wrong with this? or other?

finally,could the model research group provide the SST file for SOM run?

Thanks very much!

 

[eaton]

I don't know the details of running SOM, but I can tell you that the QNEG4 warnings you are seeing at initialization do not necessarily indicate a problem. I see these messages for non-SOM runs as well.

 

[bundy]

Did you use defineqflux to make an sst file?

Step (2) of the CAM user's guide section 2.3.2 Using the Slab Ocean Model (SOM) describes using defineqflux after using the definemld tools.
http://www.ccsm.ucar.edu/models/atm-cam/docs/usersguide/

./defineqflux -f firstfile_fromcontrolrun -l lastfile_fromcontrolrun -s sst.nc -v

This creates an sst.nc file using the data from your control run.

 

[scapps@uci_edu]

Were you able to resolve this? I am running into some problems building the SOM boundary dataset and wondered how you did it.


Thank you.

 

[you@geosci_usyd_edu_au]

I haven't seen questions about preparing the SOM-required three additional variables, ICEFRAC, SICTHK and TSOCN for the initial data cami. SOM-enabled CAM3 users may have no problem for that. Anyhow, when SOM is called to build in CAM3, these three variables are asked from cami. I would appreciate anyone can validate/comment if I did correctly for preparing these three variables because my CAM3 run stopped with a number of errors which might be associated with my SOM data input:

1) I get 12 monthly ICEFRAC from my control run cam history file (30th model year) and then use NCO tool ncea to have the annual mean ICEFRAC (I thought cami should have all annual mean variables);
2) In the similar way, I get annual mean TSOCN but from control run's TS since cam history file does not have a sst;
3) For SICTHK, since scripts in defineqflux have already got the monthly SICTHK, I simply use NCO tool ncra to get the annual mean SICTHK for cami.

I note that the SICTHK in sst file (or from defineqflux scripts) is simply a copy of prescribed ice_cov with both negative and positive values. The annual mean SICTHK is zero in the northern hemisphere after using ncra. The control run's ICEFRAC is all positive from 0-0.9. With these SOM inputs, my CAM3 run has complain but still stopped with errors such as 1) significant conservations error energy after zm_evap count, 2) significant conservations error water after zm_evap count, 3) findsp not converging at point i, k and 4) ENDRUN:FINDSP
QNEG3 from convect_deep/Q:m= 1 lat/lchnk=193 Min. mixing ratio violated at.....

 

[thrasher@pmc_ucsc_edu]

Were you ever able to resolve your conservation and findsp errors? I'm getting the same ones and can't seem to get past this point...

I note that the SICTHK in sst file (or from defineqflux scripts) is simply a copy of prescribed ice_cov with both negative and positive values. The annual mean SICTHK is zero in the northern hemisphere after using ncra. The control run's ICEFRAC is all positive from 0-0.9. With these SOM inputs, my CAM3 run has complain but still stopped with errors such as 1) significant conservations error energy after zm_evap count, 2) significant conservations error water after zm_evap count, 3) findsp not converging at point i, k and 4) ENDRUN:FINDSP
QNEG3 from convect_deep/Q:m= 1 lat/lchnk=193 Min. mixing ratio violated at.....

 

[rneale]

Did you go through the process of running CAM with fixed SSTs and calculating the QFLUXES before performing a SOM run? Also which code base are you using

 

[jianma@hawaii_edu]

Hi,

I am experiencing exactly the same error with OP. Even the problematic numbers and grid points are the same. I followed every step in making the SST file. It is strange that I can run the model with similar files and codes in T31 but just fail in T42.

Any solution yet?

Many thanks,
Tony

 

[jianma@hawaii_edu]

Did you go through the process of running CAM with fixed SSTs and calculating the QFLUXES before performing a SOM run? Also which code base are you using

Hi Richard,

Yes. I run through the process of running CAM with fixed SSTs and calculating the QFLUXES as I did in T31.

I am using CAM 3.1. The problem seems to happen in .../models/atm/cam/src/physics/cam1/tphysac.F90 when calculating the flux. I also have the warning:

atm states and fluxes
zlvl = 59.5578481243232645
potT = 271.403227672318565
Tair = 270.824453735447264
Qa = 0.287684085279874989E-02
swvdr = 0.000000000000000000E+00
swvdf = 0.000000000000000000E+00
swidr = 0.000000000000000000E+00
swidf = 0.000000000000000000E+00
Flw = 207.245659811670464
Fsnow = 0.491989229963775231E-56

ocn states and fluxes
frzmlt = -128635.595608370772
-99.1127728919929609 1200.00000000000000 -683.375112268299745 -0.701567995783501419E-04 -91.1804466231532302 118935.327470391552
sublimate away all sea ice
something is probably seriously wrong

For more info, pls visit my bluefire directory:

/ptmp/jianma/CAM/cam1/run_control/log.txt

Many thanks,
Tony