How to run the regionally refined model (RRM)

When creating the regionally refined model (RRM) case, please refer to the steps in this article.

Step-by-step guide

1. Check-out ACME : see Development Quick Guide for details.
2. Create and build a new case from the scripts directory:

   1. $ ./create_newcase -res conusx4v1_conusx4v1 -case $CASENAME -compset $COMPSET -mach $MACH -project $PROJECT
      an example for L30: $ ./create_newcase -case f1850c5_conusx4v1 -project 150001 -compset F1850C5 -res conusx4v1_conusx4v1 -mach skybridge
      an example for L72: $ ./create_newcase -case f1850c5_conusx4v1_L72 -project 150001 -compset F1850C5 -res conusx4v1L72_conusx4v1L72 -mach skybridge

3. Edit env_mach_pes.xml as needed.  For Sandia's Redsky machine, using 200 cores and 1 thread for each component runs at 0.25 simulated years per day.
4. Edit env_run.xml as needed to specify the run-length. 
5. Set-up and build the case.
   $ ./cesm_setup   
   $ ./$CASENAME.build  
    (When building a case for the first time, the script will look for the boundary and initial condition files which are located on the ACME data server.  See ACME Input Data Repository.)
6. Edit the namelist files (i.e., user_nl_cam) for user-specified output settings.  The recommended settings (as of 02 Jul 2015 ) are to follow the ne120 tunings for v0.3 listed in ne120 user_nl_cam.   The user should change zmconv_tau from 3000 to 3600, an ne30 setting, when running RRM.
7. Run your $CASENAME
8. Report the machine upon which you successfully ran the RRM configuration.  (Configuration Management for a list of machines)