need help to run cam3.5 on redhat

[meir02ster@gmail_com]

Hi, all,

I am trying to run cam3.5 on redhat with pgi90 7.0, netcdf 3.6.1 (or netcdf 4) and mpi 1.0.5. It failed while it was running. It complains about the dimension id or name during the checking process in atm/control/readinitial.F90. I am not sure if this is caused by the in compatibility of netcdf library. I thought it's not since it has passed the compilation stage and it was failed while running.

Follwing is the endding part of the log file. I hope someone can help me diagnose the problem. I really appreciate it. Thank you.

Rui

************************************************
Advected constituent list:
1 Q Specific humidity wet
2 CLDLIQ Grid box averaged liquid condensate amount wet
3 CLDICE Grid box averaged ice condensate amount wet
(GETFIL): attempting to find local file cami_0000-09-01_64x128_L26_c030918.nc
(GETFIL): using
/home/meir02/cam3.5/inputdata/atm/cam2/inic/gaus/cami_0000-09-01_64x128_L26_c030918.nc
(GETFIL): attempting to find local file
topo-from-cami_0000-09-01_64x128_L26_c030918.nc
(GETFIL): using
/home/meir02/cam3.5/inputdata/atm/cam/topo/topo-from-cami_0000-09-01_64x128_L26_c030918.nc
(GETFIL): attempting to find local file landfrac_64x128_gx1v4_c060922.nc
(GETFIL): using
/home/meir02/cam3.5/inputdata/atm/cam/landfrac/landfrac_64x128_gx1v4_c060922.nc
landfrac_file_init: LANDFRAC dataset is:
/home/meir02/cam3.5/inputdata/atm/cam/landfrac/landfrac_64x128_gx1v4_c060922.nc
control/readinitial.F90 89

Invalid dimension id or name
ENDRUN:HANDLE_NCERR
[0] MPI Abort by user Aborting program !
[0] Aborting program!

 

[meir02ster@gmail_com]

I figured out the problem is caused by that CPP_Token Staggered is defined since the default configuration is using dynamical core finite volume and resolution 144*96.

One way to overcome this error is that I can stick with the current resolution and dynamical, and change the resolution and the varaible names in those boundary data accordingly such as: lat and lon, lon->slon, lat->slat, velocities: U->US, V->VS.

Another way is to change the dynamical core to Eulerian spectral, and spatial resolution to T42, 128*64. How can I make those changes in the configure file? I know little about perl programming since the configuration file is written using perl.

I will keep on trying to work on it but I am not sure how long it will take.
if someone can help me out here. That will be more efficient and I will appreciate that.

Thanks.

Rui

 

[hannay]

To run at T42 you should configure with:
configure -dyn eul -res 64x128

 

[meir02ster@gmail_com]

Thank you very much, Cecile. I figured that out. Now the error comes from inconsistency between land fraction and ocean mask. I am still diagnosing the issue. It must be some inconsistency from the inputdata. Hope I can make it run well soon.

 

[meir02ster@gmail_com]

Now I am able to run cam3.5 with the default settings in terms of the spatial resolution 144*96. However, I want to run it at T42 resolution. It seems there is always inconsistency among the inputdata I found for T42 regarding the land fraction, ocean mask and ice mask. Has anyone run it at T42 successfully before and can you point out the path of inputdata in the repository for me?