GEOS-5

MERRA2 Preparations: Difference between revisions

← Older edit
Wrmccart (talk | contribs)
96 edits
No edit summary
Dmerkova (talk | contribs)
135 edits
 
(51 intermediate revisions by 4 users not shown)
Line 1: Line 1:
This is a page to keep tabs on MERRA2 tasks that need to be performed.
This is a page to keep tabs on MERRA2 tasks that need to be performed.


==Files to be updated==
==GSI Updates==
* gmao_global_satinfo.rc
===Files to be updated/In Progress===
* gmao_global_scaninfo.rc
* GSI_GridComp.rc.tmpl (WM)
* gmao_global_pcpinfo.rc
**Initial version available on discover:
* gsi.rc.tmpl
***/gpfsm/dnb31/wrmccart/merra2/gsi/GSI_GridComp.rc.tmpl.merra2
* GSI_GridComp.rc.tmpl
**Issues/Questions:
***airs/mhs .blk suffix & consistency w/ obsys.rc
***Which data stream for EOS/Aqua AMSU-A
***TMI rainrate bufr: separate ocean/land sources or combined file?
***SSMI rain rates - unsure if we assimilated all SSMI's in MERRA
***Ensure prepbufr configured correctly
***Needs file structure for AVHRR winds - will need to be consistent w/ obsys.rc
***Entries needed for MLS - T & O3?
***Entry (entries?) needed for TOMS?
***Not updated for cbkg != NONE
* gsi.rc.tmpl (WM)
**Initial version available on discover at:
***/gpfsm/dnb31/wrmccart/merra2/gsi/gsi.rc.tmpl.merra2
**Issues/Questions:
***All changes above w/ GSI_GridComp.rc.tmpl need to be consistent with and probably to this file
***Reorganized based on Obs type (like GSI_GridComp.rc.tmpl) to ensure dvals/dthins are consistent
***Ozone not organized - defer to Kris/Larry to tidy this up/make dthin&dvals consistent - '''DONE!'''
***Single FOV GOES Sounder only exists from 2007-onward.  GOES-08,10 need only sndr_gXX entry, GOES-13,14,15 need only sndrDN_gXX (N=1,4) entries, GOES-11,12 need both
***GOES-14 included w/ numbers fudged from GOES-13 in case future implementation is desired
* all *.arc
* all *.arc
* satbang/satbias restart
* obsys.rc (DM: $NOBACKUP/merra2/run/obsys.rc))
* obsys.rc
**GPS Startdate need to be confirmed/adjusted.  We have data that preceeds the obsys.rc start date
* sac.nl.tmpl
**MetOp-A observations need to be adjusted. 
* gmao_global_ozinfo.rc -> punt to Pawsonland?
***IASI files start rougly Sept 2008, but the AMSUA/MHS/HIRS streams may start earlier
* update sat-db:
**ozone - as soon as TOMS data are moved to the proper place - edit obsys.rc (Kris)
*** edit ozinfo-db 'active.tbl' to facilitate a switch from SBUV to MLS O3 in 2004 (Kris) - '''DONE''', file sent to RT
**sat
*** GOES Sounders need to be consistent w/ end of 5x5 prep stream (''gXX_prep'' entries in gsi.rc) and 1x1 (''gXX'' entries)
* update conv-db (presently being considered to control conventional obs)-added to tag (JG), will be used as additional tool to stream obs (diverting METEOSAT * AMV's , ERS data for Andrew huricane etc) (DM)
 
===Files Updated/hopefully completed===
* gmao_global_satinfo.rc (WM) '''DONE'''
** Template for use w/ satinfo.db: /gpfsm/dnb31/wrmccart/merra2/gsi/gmao_global_satinfo.rc.merra2
* gmao_global_scaninfo.rc (WM) '''DONE'''
** Updated at: /gpfsm/dnb31/wrmccart/merra2/gsi/gmao_global_scaninfo.rc.merra2
* <s>gmao_global_pcpinfo.rc</s> - no update needed
* gmao_global_ozinfo.rc - '''DONE!'''
** Available on discover at: /gpfsm/dnb31/wrmccart/merra2/gsi/gmao_global_ozinfo.rc.merra2
* satbang/satbias restart (WM)'''DONE'''
** Available at:
***/gpfsm/dnb31/wrmccart/merra2/gsi/satbang_rst.merra2-scout
***/gpfsm/dnb31/wrmccart/merra2/gsi/satbias_rst.merra2-scout
**Should be acceptable for scout runs.  Sources of error values/bias coefficients/etc [https://gmao.gsfc.nasa.gov/intranet/personnel/wmccarty/20130708-MERRA2-satfile-sources.xlsx available here].
* sac.nl.tmpl '''DONE'''
** Available at: /gpfsm/dnb31/wrmccart/merra2/gsi/sac.nl.tmpl.merra2


==Observations to be Acquired and/or Processed==
==Observations to be Acquired and/or Processed==
* SSMI Radiances
* SSMI Radiances v7 (YK processing data-in progress, DM-testing)
* SSMI v7 products (RR/wind speed)
* SSMI v7 wind speed (processed YK, tested DM) '''DONE!'''
* AVHRR AMVs
* ERS2 scat. (DM-testing) '''DONE!'''
** Issue at hand for recent (post-2010) data stream.  Infrastructure developed, but base observations may not actually exist (observations used for development may be from an antiquated algorithm).
* AVHRR AMVs  
** <s>Issue at hand for recent (post-2010) data stream.  Infrastructure developed, but base observations may not actually exist (observations used for development may be from an antiquated algorithm).</s>
*** <s> WM - Awaiting further guidance from D. Santek</s>
*** WM - Will proceed forward - working with Rob to determine best way to acquire data
*** bufr files (2010-Oct2013) DONE (DM)
** prepbufr AVHRR winds(1982-2010) (reprocessing files for 1982-1990 due change in input data, DONE (DM)
* Consolidated prepbufr for 79-97 - ncep2013 (DM)
* GPS - preCOSMIc era ? (DM) (from Jack Woollen:gps data starting in 2002 is from the CHAMP instrument, reprocessed by the Boulder folks. Reprocessed COSMIC starts in around 2006)
* IASI (2008-now) (DONE)
* ASCAT (2009-now) DONE
* GOES 1x1 (2007-now) (Done)
* (Question from Mike B.) VAD data prior 2001? (DM) getting data from Jack Woolen (Rob)
**getting new prepbufr data from J. Woolen. DONE
* AIRCRFT data prior cdas prepbufr (2001) can we get reprocessed data? (DM) yes, DONE
* Preprocessed TOMS data need to be moved to a proper place on archive (Kris & Rob)(DONE)
* Get and process OMI and MLS v2.2 for 2012 - present (Kris). The same for MLS Temperature (Larry C.?) DONE
* Set up automatic processing of new MLS T, MLS O3 and OMI data
* tcvitals files (filling gap for 2005 and get some preop 2004 data?) (DM) -Jack Woolen provided 2000-2006 tcvitals (Rob has access to aquired data) DONE , getting additional data for 1980-1999.


==Issues to address==
==Issues to address==
* Infrastructure to bring in "spun-up" satbias/satbang files.  Specific issues:
* Infrastructure to bring in "spun-up" satbias/satbang files.  Specific issues: (WM)
** satbang coefficients & tlapmean values - spin up from scout diags
** Restart satbias/satbang files created for scout.  Question remaining is how will values be adjusted for full M2 (if at all):
** satbias coefficients - substitute in from "spun-up" scout satbias_rst (some delta-time in the future)
*** satbang coefficients & tlapmean values - spin up from scout diags
*** satbias coefficients - substitute in from "spun-up" scout satbias_rst (some delta-time in the future)
* (Question from Austin) How much of the data currently used as 'passive' do we need? E.g. SBUV N16 has been 'passive' in FP for some time; should it be just turned off in MERRA-2? It's bad data
Retrieved from "https://geos5.org/wiki/index.php?title=MERRA2_Preparations"