JCSDA development without the adjoint changes #807
Conversation
danholdaway
requested review from
bensonr,
rem1776,
thomas-robinson and
wrongkindofdoctor
as code owners
|
@bensonr hoping we can get this merged more quickly without the adjoint changes. Since it seems GMAO can't even build with the latest FMS it will be a long time before they can test those adjoint changes. |
danholdaway
changed the title
|
@danholdaway - with this new PR, do we still need #787 to remain open? |
I would suggest closing that PR. The conversation involving GMAO folks could move to an issue perhaps. |
|
@danholdaway (and others involved), I would like to say a few things:
Therefore,
Is incorrect. I would appreciate if would please: Also note that:
Thanks! |
|
@sanAkel I was referring to the issues Matt was reporting. But it seems he got it building so that's great. In any case it would be a while before the adjoint could be tested with the branch we were trying to merge so we closed that and took out the change. |
@danholdaway Thanks for the reply. I have nothing to do/say with the
Whatever floats your boat and makes it go fast! 😉 |
| character(len=*), intent(out) :: grid_file | ||
| character(len=*), intent(in) :: mosaic_file | ||
| type(domain2D), intent(in) :: domain | ||
| integer, intent(in), optional :: tile_count | ||
| character(len=*), intent(in), optional :: custom_path |
There was a problem hiding this comment.
Please add a doxygen description of this new variable per the style guide functions section
"Inline doxygen descriptions for all arguments, except the result variable."
There was a problem hiding this comment.
Will do. We'll have another PR coming up so will add it then.
Description
This a series of changes implemented by the JCSDA to enable data assimilation applications (without the adjoint part).
The main changes are to allow for paths to the namelist files, field_table, and the INPUT directory. The solution used follows the one suggested provided to @aerorahul by @bensonr.
Other change include allowing more tracers for EMC's CMAQ assimilation and changes to MPP adjoint for data assimilation with MOM6
How Has This Been Tested?
We have run the the unit tests of the JEDI system
Checklist:
make distcheckpasses