MERRA2 InputData

Input Data Information This is a page to keep information about input data (source, processing, gaps and known issues).

Conventional data pre-QC

PREPBUFR

• RADCOR=HAIMB_HRAD
• cdas0 (1979-2000)
  • data from Jack Woolen; recieved 2013 (1979-1997 reprocessed data, 1997-1998 pulled additional files from CFSRR)
    • updates to METEOSAT AMV (early reprocessed 82-87) JW
    • updates to ADPSFC (data types 181 and 183 separated before 1998) JW
    • 1982-2000 added Vaisala correction (from original files processed by Chris Redder, adpupa_fixvai)(DM)
    • exclusion of NPN profiles data (we have separate files for npn profiler from UCAR) (DM)
• cdas (2001-present)
  • data from Jack Woolen for original MERRA
    • Vaisala correction (Chris Redder)
    • NPN profiler excluded until 20080731 (profiler obs file) (except station in Pacific and Europe that have the same kx 223)
• PAOB data 19790101-20100817
• SATWND data in prepbufr until 20100701 (after in bufr format file)
• GMS WV satwinds (KX=250) in prepbufr files are passive (missing obs. error). This is necessary because GMS WV satwinds in prepbufr format do not have information needed to distinguish between cloud-top and deep layer WV winds, and we are not able to turn on only cloud-top winds as we do for the winds from the satwind dump bufr format file.

ASCAT (20090301-recent)

• from ASCATW (bufr files needs to be processed (~msienkie/ascat/prepdata; historical data were processed by DM, realtime-it will be part of preprocessing)
• gap in September 2010 was originally filled from ASCATT format by MS)
• adding data from prior 2009? (WM??)
• need to add real-time preprocessing!

AVHRR polar winds (1982-recent)

• source: CIMSS
• obs. errors table in prepobs_errtable.global added (copy of modis ir amv's obs errors)
• ascii to prepbufr (code YK,processed DM)
  • (82-Sep09): ftp://stratus.ssec.wisc.edu/pub/winds/histavhrr/newversion
  • (Oct09-Jun10): ftp://stratus.ssec.wisc.edu/pub/winds/archive
• no observation for 12z and 18z between 20091223-10100103
• bufr format (code WC, processed DM)
  • 20100701 we switch to bufr format for all AMV
  • (Jul10-recent): ftp://stratus.ssec.wisc.edu/pub/winds/archive
  • need to add real-time preprocessing!

SATWINDS (Atmospheric Motion Vectors)

• switch from prepbufr to bufr format on Jul1, 2010
• 1979 GOES - additional AMV from FGGE (First GARP Global Experiment), which ran from Dec. 1978 to Nov. 1979. There should be AMVs from GOES East and West, Meteosat-1, GMS-1, and GOES Indian Ocean. The winds were produced here at UW-Madison, using manually tracking (from Dave Santek,CIMSS).

ERS scatterometer

• ERS1 (19910805-19960525) lots of short gaps
  • exclude 06z Aug 24, 1992 (hurricane Andrew); in original MERRA this set of winds had negative impact
• ERS2 (20030813-20110329) new version of data
  • The re-processing of ers-2 runs from from 1997-2003. The three available formats of the re-processed data are:
    • ERS.ASPS.UWI (re-processed data in UWI format for compatability with the real-time product)
    • ERS.ASPS20.N (re-processed data in ASPS format: nominal resolution 50km)
    • ERS.ASPS20.H (re-processed data in ASPS format: high resolution 25km)
  • We are processing ERS.ASPS20.N. For this data, lets give a filename like: ers2_asps20n.YYYYMMDD.tHHz.bufr
  • For the period from 2003 through the end, we have the fast delivery product (WSC.UWI).

QuikSCAT Data

• data from JPL (19990719-20091122; MS added 2003,2004)

SSMI Wind Speed

• updated to Wentz Version 7 (19870709_00z-present)
  • F08 SSM/I Jul 1987 Dec 1991
  • F10 SSM/I Dec 1990 Nov 1997
  • F11 SSM/I Dec 1991 May 2000
  • F13 SSM/I May 1995 Nov 2009
  • F14 SSM/I May 1997 Aug 2008
  • F15 SSM/I Dec 1999 present (do not use after Aug 2006 for climate study)
  • F16 SSMIS Oct 2003 present
  • F17 SSMIS Dec 2006 present
  • F18 SSMIS Oct 2009 present

Modis

• data processed from CIMSS (YK); experimental data before NESDIS was processing data and they were part of SATWIND prepbufr
• 20020701-20040708
• 20040709-20040930

NPN Profiler

• data 19920510-20080731 from UCAR, gaps filled from original Jack Woolen's cdas files in period 20060101-20080731
• data after 20080731 from NCEP
• (19920510-20080731)(excluded from cdas and cdas0)

NO-preQC data

tcvitals

• (1980-present)
• data prior 2006 from Jack Woolen (added Nov2013)

SSMI Rain Rates

• GPROF data
• F08 - 19870709_00z-19900226_18z
• F10 - 19910105_00z-19971115_18z
• F11 - 19920101_00z-19961130_18z
• F13 - 19950503_00z-20090916_18z
• F14 - 19970508_00z-20080824_18z

TRMM TMI Rain Rates

• 19980101-present
• Separate streams for land and ocean

GPS Radio Occultation

• 20010501-20071214 (from Jack Woollen:gps data starting in 2002 is from the CHAMP instrument, reprocessed by the Boulder folks. Reprocessed COSMIC starts in around 2006)
• 20071215-present

Radiances

GOES Sounder

• 20010424-20071204
  • 5x5 averaged in goesnd.* files
    • old prepbufr file - but the input is no longer through prep stream
• 20070222-present
  • 1x1 single FOVs in *goesfv* files
• Control of usage set in satinfo_db
  • Switch occurs on 20070401_00z in MERRA2

SEVIRI

• 20120215-present

TOVS/ATOVS data

AMSU-A

• 19981101-present

AMSU-B

• 19981101-present

HIRS2

• 19781101-20060930

HIRS3

• 19980721-present

HIRS4

• 20090401-present

MSU

• 19781101-20061010

MHS

• 20051101-present

SSU

• 19781101-20060505

IASI

• 20080916_18z-present

Metop-AMSU-A

• 20080916_18z-present

AIRS

• 20070401_00z-present

Aqua-AMSU-A

• 20070401_00z-present

ATMS

• 20111116-present

CrIS

• 20120406-present

SSMI

• Wentz version 7(processing YK)
• F08 - 19870709_00z-19911231_18z
• F10 - 19901208_00z-19971114_18z
• F11 - 19911203_00z-20000516_18z
• F13 - 19950503_00z-20091104_18z
• F14 - 19971218_00z-20080808_18z
• F15 - 19991218_00z-20060826_18z

SSMI/S

• 20110228_00z-present

Ozone data

SBUV

• 19781031_00z-20040930_18z (Intentionally not needed (used) past 20040930)

MLS

• 20041001_00z-present

EOSMLSt

• 20040813_00z-present

Additional information

A.C.(internal link) [1]