Longwave down sent from the atmosphere model is negative or zero

[agustin@unex_es]

I just compile cesm2-2-0 in my computer in Spain. I am using a BHIST compset ( --compset BHIST --res f09_g17 --run-unsupported --mach login1 --compiler intel )
and the initial file is: b.e21.BHIST.f09_g17.CMIP6-historical.011.

At step nstep: 343 the run crash giving the error:

lnd_import ERROR: Longwave down sent from the atmosphere model is negative or zero.

Any help is welcome



Agustin

 

[nusbaume]

Hi Agustin,

That error usually occurs if the air temperature field in CAM has unrealistic values somewhere. Did you make any modifications to the model code or settings beyond what you listed for create_newcase? If not then it is possible that the initial conditions you chose may be causing some sort of instability in the model somewhere. What happens if you try the simulation as a startup, or with a different set of initial conditions?

Another option would be to re-build and run the model with DEBUG set to TRUE in env_build.xml. This will turn on the debug compiler flags which should catch if a value has turned into a NaN or infinity.

Finally, you can try outputting CAM's air temperature field every few time steps by adding the following lines to user_nl_cam:

avgflag_pertape = 'A','I'
nhtfrq = 0,1
mfilt = 1,10
ndens = 2,2
fincl2 = 'T'

That should result in cam "h1" files being written every ten time-steps that contain the values for the air temperature in CAM. You can then examine those files to see if the air temperature is going bad somewhere (although please note that this will slow the model simulation down somewhat).

Hope that helps, and good luck with the tests!

Jesse

 

[agustin@unex_es]

Thanks Jessi for your answer, I have changed the initial conditions file (from b.e21.BHIST.f09_g17.CMIP6-historical.011 to b.e21.BHIST.f09_g17.CMIP6-historical.001) and the model run for 2 years. Now it crash again but the error is different: zap_snow_temperature: temperature out of bounds!. I had this same problem some time ago. I'm affraid I have some problems with CESM2 model in Spain that makes the temperature in the atmosphere go to low. Thanks again, Merry Christmas and a happy new year

Agustin