search

NOAA-EMC / RDASApp

Regional DAS
GNU Lesser General Public License v2.1
1 stars 7 forks source link

Explore and develop jedi conversion of model fields to latlon grids on regional domains #98

Open TingLei-NOAA opened 6 days ago

TingLei-NOAA commented 6 days ago

Recent work have found the function of conversion/interpolation of model grids to latlon grids on regional domains need further exploring and enhancements. For fv3jedi, started from convertstate_gfs_c2ll.yaml (suggested by @andytangborn ), it was found , quote from email with Andy,

1)the dimension of the output grids can't be larger than the input geometry (for the app would go through reading the grid data step for the output geometry)
2) the output grid is assumed global 
From the above point, the results will be in a global domain with too coase resolutions for my case.

For mpas-jedi, I had created an interpolator app as described https://github.com/JCSDA-internal/mpas-jedi/issues/948, in which the target latlon domains are still on the global domain. Further developments are also needed. This issue will facilitate and document collaborated efforts on this.

guoqing-noaa commented 6 days ago

@TingLei-NOAA Could you talk more about this?

Do you mean to develop a tool similar to the MPASSIT program created by LarissaReames-NOAA at NSSL?

https://github.com/LarissaReames-NOAA/MPASSIT

TingLei-NOAA commented 6 days ago

@guoqing-noaa What i am talking about is the tools in jedi as fv3jedi_convertstate.x and the one https://github.com/TingLei-daprediction/mpas-jedi/pull/1 for mpasjedi , which need some enhancement to work for regional cases. (conceptually, the target grids could be "arbitrary" only if they are known to the app.
Thanks for pointing to the [MPASSIT] work. It seems it is specifically for wrf grids for being now, not exactly what we want described in this issue. But of course, it would be a good reference we can go to and the developers' expertise we could count on in those applications/developments.

guoqing-noaa commented 6 days ago

@TingLei-NOAA Thanks for a quick tagup with me on this issue and I now understand more. For reference, on the MPAS side, this will provide a one-way nest capability (similar to ndown for WRF).

TingLei-NOAA commented 6 days ago

@guoqing-noaa Thanks a lot for your sharing of the information and your idea through this online/offline discussions. If it is ok, would you please share the link to the MPAS slides you mentioned? Thanks.

guoqing-noaa commented 6 days ago

@TingLei-NOAA Sorry I forgot what "MPAS slides" you refer to.

Do you mean this MPAS practice guide? https://www2.mmm.ucar.edu/projects/mpas/tutorial/Boulder2023/ or run MPAS on Hera/Jet

TingLei-NOAA commented 6 days ago

@guoqing-noaa Thanks a lot. Either is working for me! Together, that is double-great! Have a wonderful time!