This is to document the email conversions between @daviddowellNOAA @TrevorAlcott-NOAA @hu5970 and me:
Per @daviddowellNOAA
Previously, MPAS initialization from RAP and RRFS rotated latlon grids didn't work.
Our temporary solution was to first use wgrib2 to interpolate to a Lambert conformal
grid, then run the MPAS initialization.
Soon, we won't need that temporary, intermediate interpolation.
Ming is currently testing code that allows us to initialize MPAS directly from RAP rotated lat lon.
For MPAS initialization from RAP 130 data, the initialization code worked as is (i.e., no wgrib2 step needed)
because the 130 grid is already Lambert conformal.
Per @TrevorAlcott-NOAA
Larissa already updated the MPAS init code to allow initialization from full-domain RAP.
Here's her response from when I brought up the issue:
"I’ve finally found all the bugs and have a working RAP grid implementation for the init_atmosphere core.
You can find it (on top of our NSSL_develop branch) here:
https://github.com/LarissaReames-NOAA/MPAS-Model/pull/new/NSSL_develop_RAPinit
"
The remaining issue is with RRFS North America grids, which are also rotated lat-lon but
a different type of rotated lat-lon (grid definition template #1) than RAP (grid definition template #32769).
So remapping full-domain RRFS to a compatible grid type is still required.
Per @hu5970
David and I are testing Larissa's code for initializing MPAS from RAP rotated latlon.
We should be able to report the results soon.
This is to document the email conversions between @daviddowellNOAA @TrevorAlcott-NOAA @hu5970 and me:
Per @daviddowellNOAA
Per @TrevorAlcott-NOAA
Per @hu5970