Child pages
  • Changing Dates of a Run
Skip to end of metadata
Go to start of metadata

Either use the xmlchange function or open the env_run.xml and edit directly to alter:

 ./xmlchange RUN_STARTDATE=$run_start
 ./xmlchange STOP_OPTION=$stop_option
 ./xmlchange STOP_N=$stop_n

These are by default RUN_STARTDATE  = 2005-01-01, STOP_OPTION = ndays, and STOP_N = 5, the recommended "out-of-the-box" test set-up.

RUN_STARTDATE has format YYYY-MM-DD. *** Be sure to change the meteorological forcing files used and that emissions cover the dates specified***.

STOP_OPTION has many options including ndays, nmonths, or nyears (see env_run.xml for more options). STOP_N will stop after the specified number of STOP_OPTION increments. For example the default:

STOP_N = 5 

will stop simulations after 5 days. Depending on your set-up and computational resources, it is usually recommended to run 6 to 12 months at a time, after testing phase with 5-days is complete and the experimental runs are begun.

You will also need to point to initialization files by updating the path by adding the following to user_nl_cam:

 ncdata = '$path_to_init_file'

as well as the land initialization in user_nl_clm:

finidat = '$path_to_init_file'
use_init_interp = .true.

Publicly available initialization files from CESM2 are available, check out this page: Restart Files

Initialization files are written out from a previous CESM1 simulation and specify initial states of chemical fields, instead of spinning up from zero. Another option is to use the default initialization file, perform a one-year spin up and re-initialize with the output from that simulation to minimize the initial condition influence.

Some initialization files can be found on Cheyenne mass storage system (HPSS):

Every year from 1993-2015: /home/buchholz/CESM206_CAMchem_longrun_fmerra_branchbase

Every year from 1975-1993: /home/buchholz/CESM206_CAMchem_longrun_fmerra

  • No labels