2025-10

Here's how to derive new experiment group(s) based on a complete control group experiment in MPAS.

  1. Even before the control group experiments, prepare a directory with a skeleton model. This could probably include:

    • The model
    • WRF physics
    • Grid file and static file.
    • Possibly scripts for conveniently submitting as a job
    • You may want to delete the namelists, but keep the streams and the streamlists.
  2. Copy the skeleton directory over. It’s probably a good idea to use cp -L (dereferencing the links) because with this you get the actual model binaries copied. So it should be faster if you’re running multiple experiments at the same time. (*not benchmarked)
  3. Copy over the namelist of the control group. Make appropriate edits.
  4. Link/Copy over the init of the control group.
  5. Link/Copy over the sfc_update of the control group
  6. Link/Copy over the LBCs of the control group.
Basically, what you would do with init_atmosphere mode 7, 8 and 9. You don't need to use WPS on the downloaded data again.
  1. Copy any other input streams (e.g. FDDA).