CLM45 with transient CO2 concentration problem

[zichongchen22@gmail_com]

Hi, With CESM1.2.2, I used the compset I20TRCRUCLM45BGC, and follow the instructions http://www.cesm.ucar.edu/models/cesm1.2/clm/models/lnd/clm/doc/UsersGuide/x12702.html, try to run the CLM45 with transient CO2 conditions. I set a startup run.But the model crashed and my cesm.log file shows as below:(shr_sys_abort) ERROR: (shr_stream_getCalendar)  ERROR: nf90_open file /atm/datm7/CO2/fco2_datm_1765-2007_c100614.nc(shr_sys_abort) WARNING: calling shr_mpi_abort() and stopping. Firstly I think it might be the NETCDF probelm,but I changed a few netcdf versions it still shows that the model crashed. and my module list shows as below:Currently Loaded Modulefiles:  1) local                                                   11) gcc/4.8.2  2) vars                                                    12) intel/2013/sp1_update3  3) user                                                    13) impi/5.0.0.028/intel  4) moab                                                    14) pnetcdf/1.5.0  5) suacct                                                  15) hdf5/hdf5-1.8.13-intel-2013-sp1-update3-impi-5.0.0.028  6) base                                                    16) netcdf/4.3.2-intel-sp1-update3-parallel  7) netcdf/fortran-4.2.intel                                17) netcdf/fortran-4.4-intel-sp1-update3-parallel  8) mpc/1.0.2                                               18) xerces/3.1.1_intel  9) mpfr/3.1.2                                              19) esmf/6.3.0rp1 10) gmp/5.1.3                                               20) cesm/1.2.2Anyone have ever run into this problem before, please help me figure it out. Thanks. Zichong

 

[jedwards]

it looks like your DIN_LOC_ROOT is not properly defined - does this file exist?   /atm/datm7/CO2/fco2_datm_1765-2007_c100614.ncif not then the problem doesn't have anything to do with which netcdf you are using.

 

[ihmn_zgz@gmail_com]

Hello, I'm having exactly the same problem you had. What was your solution? I checked my inputdata directory and there was no /atm/datm7/CO2, so I manually created it and downloaded it there. But I still get the same error "ERROR: nf90_open file /atm/datm7/CO2/fco2_datm_1765-2007_c100614.nc". I also checked DIN_LOC_ROOT in my case directory with: > ./xmlquery DIN_LOC_ROOT
env_run.xml: DIN_LOC_ROOT = /home/hermoso0/scratch/inputdata> ls /home/hermoso0/scratch/inputdata/atm/datm7/CO2/
fco2_datm_1765-2007_c100614.ncSo I don't understand what it is wrong. Any clue? Is somehow looking for /atm/datm7/CO2/fco2_datm_1765-2007_c100614.nc in RUNDIR instead of DIN_LOC_ROOT? Your help will be welcome.

 

[ihmn_zgz@gmail_com]

I ended up finding the solution for myself. As usual, I will post what was wrong just in case somebody else finds the same trouble.
Notice that the error says it can't open the file /atm/datm7/CO2/fco2_datm_1765-2007_c100614.nc, it does not write the full path from the inputdata directory. In the instructions "Running stand-alone CLM with transient historical CO2 concentration" we are said to create a 'datm.streams.txt.co2tseries' file with the cat command, copying the text that appears in the instructions, that specifies the access to this fco2_datm_1765-2007_c100614.nc. The path is specified with this: $CSMDATA/atm/datm7/CO2I did not realize that when copying this text directly with the cat command, the resulting text appears like /atm/datm7/CO2, because there is no variable $CSMDATA. Therefore, it looks for the file in the wrong place. To solve this, I directly wrote my inputdata path "homehermoso0scratchinputdata" (which is always the same) instead of $CSMDATA. Another solution would have been to write $CSMDATA. I am not sure why we use $CSMDATA instead of $DIN_LOC_ROOT, though. Is it the same thing?Anyway, I hope this is useful other users as dumb as I am.