Ting finds:
In the current RRFS DA run, the full_ensemble=.false., which is true in EMC parallel run to do the "trajectory correction" , though it has been our interests to do further investigation on its benefits . But it is ok for RRFS to start from the safe full_ensemble=.false. Just hope this difference is noted.
I checked gsiparm.anl in FV3LAMDAX, full_ensemble=.false.
I also checked RRFS_prototype, this is not used in gsiparm.anl.
Ting finds: In the current RRFS DA run, the full_ensemble=.false., which is true in EMC parallel run to do the "trajectory correction" , though it has been our interests to do further investigation on its benefits . But it is ok for RRFS to start from the safe full_ensemble=.false. Just hope this difference is noted.
I checked gsiparm.anl in FV3LAMDAX, full_ensemble=.false. I also checked RRFS_prototype, this is not used in gsiparm.anl.