search

NOAA-EMC / GDASApp

Global Data Assimilation System Application
GNU Lesser General Public License v2.1
15 stars 31 forks source link

Use _anl rather than _ges dimensions for increments in FV3 increment converter YAML #1013

Closed DavidNew-NOAA closed 6 months ago

DavidNew-NOAA commented 6 months ago

npx_ges should be npx_anl for the JEDI and FV3 increments. Same for npy and npz. See https://github.com/NOAA-EMC/global-workflow/pull/2420#issuecomment-2027575028

CoryMartin-NOAA commented 6 months ago

@DavidNew-NOAA where will you account for the interpolation of the increment when it is added to the background?

In current GDAS operations, the analysis resolution (i.e. that which the increment is computed on) is the ensemble resolution which is half the background resolution. That is why the .ensres files exist. The benefit of JEDI writing the increment on the input resolution is that you can just do analysis = guess + increment, but if you write out the native increment, you have to interpolate it to the guess resolution first.

Does that make sense?