Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

0.05 deg run stalls out 'Calculating orbital parameters'

amans

Aman Shrestha
New Member
What version of the code are you using?
alpha-ctsm5.2.mksrf.23_ctsm5.1.dev171


Have you made any changes to files in the source tree?
None


Describe every step you took leading up to the problem:
  1. ./create_newcase --case /glade/work/amans/cases_0.05/Spinup_000_I2000Clm51BgcCrop_MRB_0.05_MOSART_0.125 --compset IHistClm51BgcCrop --res f09_g16 --run-unsupported
  2. Added surface, landuse, paramfile, and initial condition file to user_nl_clm in case dir: /glade/work/amans/cases_0.05/Spinup_000_I2000Clm51BgcCrop_MRB_0.05_MOSART_0.125
    Code:
    fsurdat='/glade/work/amans/data/MRB-CONUS_0.05/merged_ctsm5.3_w_FMAX_irrig/surfdata_480x770-MRB_0.05_hist_1979_78pfts_c250324.nc'
    paramfile='/glade/work/amans/data/inputs/params/calibrated_Aman/d_max_watsat/ctsm51_params_Mengqi_d_max_15_watasat_sf_1.4.c250324.nc'
    flanduse_timeseries='/glade/work/amans/data/MRB-CONUS_0.05/ctsm5.3_fsurf/landuse.timeseries_480x770-MRB_0.05_hist_1979-2022_78pfts_c250319.nc'
    finidat='/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/initdata_map/clmi.I2000Clm50BgcCrop.2011-01-01.1.9x2.5_gx1v7_gl4_simyr2000_c240223.nc'
  3. Added CONUS404 forcing at my surface resolution and clipped to surface extent to user_nl_datm_streams.
  4. Xmlchanged land domain mesh, atm domain mesh, land and atm NX NY, ROF grid NX and NY. Xmlchanged DATM yr align and start to my forcing start year.
  5. Xmlchanged JOB_WALLCLOCK_TIME=03:00:00,RESUBMIT=9,STOP_N=1,STOP_OPTION=nyears
  6. ./case.build and ./case.submit


Describe your problem or question:
I am trying to run a 0.05 deg resolution regional CLM-only simulation using transient landuse. Previously I had no problem running I2000 static landuse cases using the same surface creation methods and my paramfile. This specific case is supposed to be for spinup for my actual simulation. However, even one-year of run stalls out at 3 hours. My previous I2000 runs were achieving ~1 sim-year/1 hour. I am using 15 nodes so this previous thread does not help. The lnd.log file at /glade/derecho/scratch/amans/Spinup_000_I2000Clm51BgcCrop_MRB_0.05_MOSART_0.125/run has repetitive Computed Orbital Parameters' blocks.
(shr_orb_params) ------ Computed Orbital Parameters ------
(shr_orb_params) Eccentricity = 1.671178E-02
(shr_orb_params) Obliquity (deg) = 2.344237E+01
(shr_orb_params) Obliquity (rad) = 4.091465E-01
(shr_orb_params) Long of perh(deg) = 1.025529E+02
(shr_orb_params) Long of perh(rad) = 4.931478E+00
(shr_orb_params) Long at v.e.(rad) = -3.253449E-02
(shr_orb_params) -----------------------------------------
Obviously, I could increase the wallclock time. But is there any other way to speed things up? Also, why are orbital parameter calculations occurring for IHist compset but not for I2000 compsets?
 

slevis

Moderator
Staff member
To make sure I understand correctly:
- The I2000 case runs at 0.05 deg resolution
- The transient case does not

The fact that things hang could suggest a memory issue. In that case, you may need to run on more than 15 nodes.

If it's not a memory issue, my first troubleshooting suggestion is to try the default transient case for f09, without all the changes that you made. Then maybe try another default resolution, such as f19. When both cases have worked, compare their case directories to see everything that changes for a case to go from f09 to f19. This may give you insight into anything that you may have missed when going to 0.05.

You could follow a similar procedure by comparing the f09 transient to an f09 I2000 case directories. This will show you everything that changes to go from a time-slice to a transient simulation. Then you could generate a new I2000 case at 0.05 deg (which you said works) and then manually make the modifications needed to make it transient.

I hope this helps.
 

amans

Aman Shrestha
New Member
Will try that. Thank you slevis.
On a side note, I am curious about the orbital parameter calculations and have not seen it previously in I2000 cases. What does it do and what is it related to in the IHist cases?
 

slevis

Moderator
Staff member
Earth's orbital parameters affect climate. For example the glacial-interglacial cycles are attributed to Earth's orbital parameters. As an Earth System Model, the CESM calculates Earth's orbital parameters to calculate the solar radiation delivered to Earth at any given time. If you are interested in climate topics, I highly recommend reading more about that.
 
Top