Main menu

Navigation

E1850CN equilibrium simulations

5 posts / 0 new
Last post
eem409@...
E1850CN equilibrium simulations

Hello all,

I am currently using CESM to do a few equilibrium simulations. I am using the E1850CN compset straight from the book. Part of my experiment is looking at soil carbon in Alaska. After 100 years worth of simulation I noticed that the results that I had were not quite what is to be anticipated. I have attached a graph of what the SOILC distribution looks like from the model. From my understanding there should be ample soil carbon in Alaska. I was wondering if someone else could attach a similar image of their SOILC plot to the thread. That way I could know if this is simply my problem or a more global problem.

Thanks,
Erik

Attachment: 
slevis

Erik,

Dave Lawrence (dlawren@ucar.edu) noticed the same and attributed it to low productivity simulated by the CLM's Carbon-Nitrogen model. He and Sean Swenson (swensosc@ucar.edu) have been working on improvements to the CLM hydrology I think, that seem to help.

Pls write to Dave and/or Sean for more info. They may be away from email for a few days, so they may not respond right away.

Sam

Terrestrial Sciences Section CGD/NESL/NCAR

devarajun@...

Hello Erik

I am a novice user.I am trying to build the CESM1.0.4 E1850CN compset. But it is getting build. I am using the following steps to do it:

svn list https://svn-ccsm-release.cgd.ucar.edu/model_versions
svn co https://svn-ccsm-release.cgd.ucar.edu/model_versions/cesm1_0_4 cesm1_0_4

create_newcase -case ctrl -compset E1850CN -res f19_f19 -mach generic_linux_intel -scratchroot /home/devaraju/angshuman/cesm1_0_4/exec -din_loc_root_csmdata /home/devaraju/angshuman/cesm1_0_4/input -max_tasks_per_node 12

cd ctrl
./configure -case

./ctrl.generic_linux_intel.build

Now while building all the inputdatasets are downloaded automatically. to run I used :

qsub ctrl.generic_linux_intel.run

but it is not running. Is it the problem with inputdata sets as it must have the SOM related dataset.

There is no error message in the error file while in the output file following message is coming:

Thus no job control in this shell.
Welcome to n50.caos.iisc.ernet.in
---------------------------
Today Date: Thursday, 31 January 2013.
Today Time: 10:20:38 AM IST
---------------------------
NIS: mandhan.caos.iisc.ernet.in
---------------------------
My Diskspace -
---------------------------
PBS_MOMPORT=15003
CCSM_COMPSET=E_1850_CAM5 (E1850C5)
COMP_ATM=cam
COMP_LND=clm
COMP_ICE=cice
COMP_GLC=sglc
COMP_OCN=docn
COMP_CPL=cpl
CCSM_LCOMPSET=E_1850_CAM5
CCSM_SCOMPSET=E1850C5
COMP_RUN_BARRIERS=FALSE
COMP_INTERFACE=MCT
BUILD_COMPLETE=TRUE
SMP_BUILD=a0l0i0o0g0c0
SMP_VALUE=a0l0i0o0g0c0
CICE_DECOMPTYPE=cartesian
CICE_AUTO_DECOMP=true
-------------------------------------------------------------------------
CCSM BUILDNML SCRIPT STARTING
- To prestage restarts, untar a restart.tar file into /home/devaraju/angshuman/cesm1_0_4/exec/ctrl/run
- Create modelio namelist input files
CCSM BUILDNML SCRIPT HAS FINISHED SUCCESSFULLY
-------------------------------------------------------------------------
CCSM PRESTAGE SCRIPT STARTING
- CCSM input data directory, DIN_LOC_ROOT_CSMDATA, is /home/devaraju/angshuman/cesm1_0_4/input
- Case input data directory, DIN_LOC_ROOT, is /home/devaraju/angshuman/cesm1_0_4/input
- Checking the existence of input datasets in DIN_LOC_ROOT
CCSM PRESTAGE SCRIPT HAS FINISHED SUCCESSFULLY
Thu Jan 31 10:21:06 IST 2013 -- CSM EXECUTION BEGINS HERE
Thu Jan 31 10:22:21 IST 2013 -- CSM EXECUTION HAS FINISHED
Model did not complete - see /home/devaraju/angshuman/cesm1_0_4/exec/ctrl/run/cpl.log.130131-102039

I have checked the cpl.log file as well but could figure out the problem.

Many thanks
Angshuman

eem409 wrote:

Hello all,

I am currently using CESM to do a few equilibrium simulations. I am using the E1850CN compset straight from the book. Part of my experiment is looking at soil carbon in Alaska. After 100 years worth of simulation I noticed that the results that I had were not quite what is to be anticipated. I have attached a graph of what the SOILC distribution looks like from the model. From my understanding there should be ample soil carbon in Alaska. I was wondering if someone else could attach a similar image of their SOILC plot to the thread. That way I could know if this is simply my problem or a more global problem.

Thanks,
Erik

Dev

slevis

Will be great if Erik (eem409) can help.
Meanwhile, though, I am confused because you say "not getting built" and then "not running."
If the model is not getting built, then it's completely normal for it to not be running.

Sam Levis

Terrestrial Sciences Section CGD/NESL/NCAR

devarajun@...

Hello

Earlier I was using f19_f19 grid. When I changed the grid to f19_g16, its running. But the run is too slow.
I checked the ccsm_timing.$CASE.$datestamp which shows:

Overall Metrics:
Model Cost: 9440.31 pe-hrs/simulated_year (scale= 1.00)
Model Throughput: 0.15 simulated_years/day

I do not understand why it is this much slow.

The $CASE.$machine.run is :

#PBS -N som6
#PBS -q Q2
#PBS -l nodes=5:Q2_nodes:ppn=12

Please help me out.

Many thanks
Angshuman

Dev

Log in or register to post comments

Who's new

  • Nicholas.Davis@...
  • numarsanifa@...
  • bingdian_46@...
  • mxy2832029@...
  • nthg2000@...