Status of the CAM-SE CESM/DART assimilation:

DART supports assimilation using the CAM-SE model
in the DART Classic release.  The scripting supports
CESM release 1.2.1 and some support for release 1.5.

Contact us at dart@ucar.edu for more information.



Notes from the model developer:

Tasks from 2017-8-activities:

        CESM2 uses ne0 for var res CAM-SE assim setup.
           There are 4 predefined resolutions, like
           <sname>ne0np4_conus_x4v1_lowcon_ne0np4_conus_x4v1_lowcon</sname>
>       I can't ask for lon-lat and PHIS fields to be written from CAM-SE to the same file,
           or PHIS won't be usable by model_mod.  Can lon-lat be written to h1, or the last h#?
           Or what standard is being adopted by Karspeck/multi-component?
        Test xyz_location; find closest 2
           Much slower than the single-closest xyz_location.
           2nd test with more boxes even slower; handed off to Nancy.
           Need seems to be reduced; choosing correct compiler options makes speed
              comparable to FV core, and minimal compared to file I/O and archiving.
        Script to make cross sections of CAM-SE output?
           Interpolated to a lon,lat grid?
        Work on scripts for automating some of the processing of CAM-SE+DART output,
           so that there aren't so many manual steps in dealing with the cubed-sphere fields.

CAM-SE as of 2017-9 needs new/different initial files according to Lauritzen
because of a new vertical coordinate.

The last committed model_mod which handles CAM-SE is
   Working Copy Root Path: /glade/u/home/raeder/DART/Trunk
   Relative URL: ^/DART/trunk/models/cam
It's not RMAized or Manhattanized.
Now that the cam-fv has been rewritten (2017-10) this model_mod
will only be useful as a source of subroutines and how to use them.
The subroutines will probably need to be updated for RMA.
