search

ufs-community / ufs-weather-model

UFS Weather Model
Other
140 stars 247 forks source link

RRFS and WRF-Hydro Coupling: Noah-MP #1643

Open jkbk2004 opened 1 year ago

jkbk2004 commented 1 year ago

Description

Improving land-surface flux partitioning in operational short range forecasts through integration of NOAA weather and water models Coupling Details:

  1. The RRFS passes the following information to WRF-Hydro. Infiltration/runoff, soil drainage, soil temperature, fractional soil moisture, liquid fractional soil moisture.
  2. The WRF-Hydro passes back fractional soil moisture, liquid fractional soil moisture.

Solution

  1. Rebase code to the latest ufs-community/ufs-weather-model and create draft PRs.
  2. Set up a regional RRFS based baseline case with Noah-MP LSM
jkbk2004 commented 1 year ago

@danrosen25 -DCCPP_SUITES option for noahmp are FV3_GFS_cpld_rasmgshocnsstnoahmp_ugwp, FV3_GFS_v16_noahmp, and FV3_RAP_noah_sfcdiff_cires_ugwp. and for hafs case, FV3_HAFS_v1_thompson_noahmp. @barlage @RatkoVasic-NOAA any comment might be useful here about these suite options for a rrfs test run.

jkbk2004 commented 1 year ago

@grantfirl @ChunxiZhang-NOAA @dustinswales In this case, just setting up a rrfs baseline with FV3_GFS_v16_noahmp might be a good starting point, I guess. Can I ask your opinion ?

RatkoVasic-NOAA commented 1 year ago

For the regression tests, I suggest to start from 'regional_control" test in /~tests/rt.conf That is high resolution domain (3km) with all parameters used in our current parallel ("LAM", using FV3_GFS_v15_thompson_mynn_lam3km SDF). Just domain is small enough to run it in small wallclock time - it covers some of California and Nevada and little bit of Pacific Ocean. Soon (hopefully in couple of months) we should move to RRFS parallel's parameters, FV3_HRRR SDF. And that would be our new CONTROL run for the regional tests.