minghua_zhang@stonybrook_edu
New Member
Hello,
I am trying to use the SCAM on Derecho to do some physics tests. I checked out the latest version of CESM2.2.2 (and tested CESM2.2.0) by using the compset FSCAM. The default SCAM simulation (ARM9707 case) ran, but the data in the output file (e.g., fscam.cam.h0.1997-06-18-84585.nc) are incorrect. For example, T, Q, CLOUD can all reach 10e+18 etc.
I checked into the model simulation and looked at the state variables in subroutines of physpkg.F90 or scmforecast.F90, they are all correct. The out error appears caused by the outfld in the cam_history.F90. It seems to be related with netcdf output on Derecho in the FSCAM mode. The output from the GCM run is fine.
Has anyone had this problem before? Any suggestion will be appreciated.
Thanks,
Minghua Zhang
I am trying to use the SCAM on Derecho to do some physics tests. I checked out the latest version of CESM2.2.2 (and tested CESM2.2.0) by using the compset FSCAM. The default SCAM simulation (ARM9707 case) ran, but the data in the output file (e.g., fscam.cam.h0.1997-06-18-84585.nc) are incorrect. For example, T, Q, CLOUD can all reach 10e+18 etc.
I checked into the model simulation and looked at the state variables in subroutines of physpkg.F90 or scmforecast.F90, they are all correct. The out error appears caused by the outfld in the cam_history.F90. It seems to be related with netcdf output on Derecho in the FSCAM mode. The output from the GCM run is fine.
Has anyone had this problem before? Any suggestion will be appreciated.
Thanks,
Minghua Zhang