[production/RRFS.v1]saSAS sigmab initialization modification during the first timestep

[JiliDong-NOAA]

This PR follows Jongil's suggestion and aims to reduce the large convective reflectivity caused by saSAS adjustment in the first timestep during a warm start. The issue is likely related to the inconsistency when DA updates the moisture at t but not the moisture from the previous timestep (t-36s). The moisture from the previous timestep is needed for initializing sigmab (updraft area fraction) when calculating qadv (q advection or tendency term).

The PR forces qadv to zero in the first timestep when a namelist parameter sigmab_coldstart is set to .true. It also reduces the lower limit of sigmab from 0.01 to 0.0 in the first timestep.

sigmab_modification.pptx

[JiliDong-NOAA]

Could we add @JongilHan66 as the reviewer? Thanks!

[lisa-bengtsson]

@JiliDong-NOAA I think these changes look good - do you plan on setting sigmab_coldstart to true only if you cycle with DA? If you restart from restart files, then q is updated at the previous timestep and read in to be used in physics for the current time-step. I would propose including this change in the ufs-weather-model and also alert the HAFS, GFS and GEFS teams to set this flag to true in cycled configurations.

[JiliDong-NOAA]

@JiliDong-NOAA I think these changes look good - do you plan on setting sigmab_coldstart to true only if you cycle with DA? If you restart from restart files, then q is updated at the previous timestep and read in to be used in physics for the current time-step. I would propose including this change in the ufs-weather-model and also alert the HAFS, GFS and GEFS teams to set this flag to true in cycled configurations.

You are right, Lisa. The idea is to set sigmab_coldstart to false by default and switch it to true only for cycled DA warm start. This should ensure the result remains unchanged for restart. I will submit a PR to FV3 later.

Thanks for reviewing this PR!

[grantfirl]

One comment line that was missed when other comment lines were removed.

[grantfirl]

There are no issues from a CCPP perspective, but how much more development should be expected in this branch? Plus, how much of this development needs to go to ufs/dev?