Fix for EnKF cycling#251
Conversation
chunhuazhou
requested review from
CurtisAlexander-NOAA,
JeffBeck-NOAA,
TrevorAlcott-NOAA,
christinaholtNOAA,
guoqing-noaa,
hu5970,
jaymes-kenyon and
terraladwig
as code owners
|
@hu5970 @guoqing-noaa Can you please take a look at this PR and let me know if there is any modifications needed? Thank you! |
|
@chunhuazhou Did you answered my question on extend boundary condition check? |
Where is it? I don't see any questions. |
| <and> | ||
| <datadep age="00:00:05:00"><cyclestr>&FG_ROOT;/@Y@m@d@H{{ slash_ensmem_subdir }}/ics/gfs_data.tile7.halo0.nc</cyclestr></datadep> | ||
| <datadep age="00:00:05:00"><cyclestr offset="-{{ extrn_mdl_ics_offset_hrs }}:00:00">&FG_ROOT;/@Y@m@d@H{{ slash_ensmem_subdir }}/lbcs/gfs_bndy.tile7.{{ "%03d" % boundary_len_hrs }}.nc</cyclestr></datadep> | ||
| <or> |
There was a problem hiding this comment.
@hu5970 Is this the section that you are concerned about? The reason I changed it is that when I initialize the cycle at 00Z and have the ics and lbcs tasks done for 00Z.But because I am having extrn_mdl_ics_offset_hrs=6 for GDASENKF, it can't find lbcs from 18Z and therefore won't start prep_cyc. By changing it to look for lbcs in the past few hours, it will kick off the task since I have lbcs at 00Z.
| {%- for h in range(0, extrn_mdl_ics_offset_hrs+1) %} | ||
| <datadep age="00:00:05:00"><cyclestr offset="-{{ h }}:00:00">&FG_ROOT;/@Y@m@d@H{{ slash_ensmem_subdir }}/lbcs/gfs_bndy.tile7.{{ "%03d" % boundary_len_hrs }}.nc</cyclestr></datadep> | ||
| {%- endfor %} | ||
| </or> |
There was a problem hiding this comment.
Can you explain why we need this extension of checking for boundary files in retro?
There was a problem hiding this comment.
I guess I wasn't so sure why it was set to find the lbcs from extrn_mdl_ics_offset_hrs ago. So to keep its original setting and expand it for checking lbcs for the start-up cycle, I extended this check for the time window (- extrn_mdl_ics_offset_hrs, 0). May we can talk more about this tomorrow at the 10am meeting.
…ent density in hybrid/envar tasks (NOAA-GSL#251) * Add the variable controlling SDL/VDL to CONUS 3 km config file * workflow changes to run FED EnVar assimilation in th hybrid_radarref task * quick edit to script to add model fed to NetCDFs in preprocessing * FED assimilation needs SST in the anavinfo * Set GSI namelist setting to use model FED to calculate the innovation when FED is assimilated in EnVar to true * specify correct chunk size in ensemble members' FED * add dependency for hybrid task when GLM FED preprocessing is on * First draft of changes needed to move ensemble FED calculation to ensemble workflow * add FED workflow generation variables to sample config files * one line typo * one-letter typo * Corrections to save_restart script for ens when adding model FED * acts of desperation to convince GSI to read modified NetCDFs * remove old commented code that will not be used * Changes for SDL/VDL FED option * roll back change that was just used for testing * Change naming of EMC tiles to PROD tiles * change EMC tile naming to PROD * Change GLM EMC tile names to PROD in sample configs * the name of the GLM preprocessing task changed slightly? * save restart task lua
2 participants
DESCRIPTION OF CHANGES:
Fixing some task/data dependencies for EnKF cycling.
TESTS CONDUCTED:
Tests done on Hera.
ISSUE (optional):
CONTRIBUTORS (optional):
If others have contributed to this work aside from the PR author, list them here