Model Output Data for GAPP from NCAR

1
Model Output Data for GAPP from NCAR

• Steven Worley
• and
• Chi-Fan Shih
• NCAR
• Scientific Computing Division
• 31 August 2004

2
Outline

• NCAR archive of NARR
• Other model output data
• Access to NARR for GAPP PIs
• NLDAS 51-year hydrologic reanalysis (option)

3
NCAR archive of NARR

• Now
• Downloading from NCEP
• Analyses, accumulations (from 0-3hr forecasts) ,
  and averages (from 0-3hr forecasts).
• On board, 1990 2003, 2.2 TB
• daily tar files, 465 MB/day
• Current structure, bulk files, not optimal for
  usage
• Will organize collections for science needs
• Suggestions UPA, BL, Soil, surface?
• Need to investigate subset products mentioned by
  K. Mitchell
• Climatologies swapping custody with JOSS

4

• NARR facts
• NCEP Environmental Modeling Center (EMC)
• 1979-2003
• Continued daily runs as Regional Climate Data
  Assimilation System (R-CDAS)
• Eta model frozen in 2003
• 32km, 45 levels
• Processed output for users
• 32km horizontal, 179 fields, 3-hourly (8x/day)
• Standard data on 29 vertical lvls (25mb for
  1000-700mb)
• Many single lvl and vertical integrated fields
• Climatologies, basis period (1979-2001)
• Monthly (170 fields)
• Monthly means at each synoptic time (170 fields)
• Monthly means daily (45 fields)
• Monthly means day 3-hourly (17 fields)

5
Other model output data

• Mesoscale
• NCEP Eta 212 grids, 40km, 26 lvls, 1995-2004
• Analysis, initialization and forecasts (60hr)
• NOAA FSL MAPS, 40km
• For GCIP, selected (3 or 4) vertical lvls,
  1996-2003
• Hybrid isentropic-sigma coordinate
• 40 lvls, 1997-2000
• 50 lvls, 08/2004-cont.
• Global Operational and Reanalyses
• NCEP/NCAR Reanalysis I, 2.5, 17 lvls, 1948-2004
• NCEP/DOE Reanalysis II, 2.5, 17 lvls, 1979-2004
• NCEP FNL, 1, 26 lvls, 1999-2004
• ERA-40 (more following)

6

• ERA-40
• Data for 46-years (1957-2002)
• Model T159, 60 levels, 6-hrly, 35 TB
• Analyses 2.5 and 1.1º
• 23 pressure lvls., surface, vertical integrals
• Monthly means
• Forecasts
• Access
• Restricted to non-commercial research in U.S.

7
Access to NARR for GAPP PIs

• Coordinate a plan with GAPP DMC and JOSS
• Portal to in situ and model data
• No duplication of effort, one-stop discovery
  interface
• Modes of Access
• Direct computation at NCAR via MSS
• Direct file download from online service
• Data selection by user request

8

• Online service
• Order 2TB for most needed products
• More if required
• Order 1TB for custom requests
• TBD
• What products?
• Should standard subsets be created?
• Should the NARR-GAPP model data have limited
  access?
• Registration, authentication, etc.
• Software for GRIB formatted NARR
• NCARs NCL (graphics, analysis, format conversion
  to netCDF)
• GrADS

9

• Online service, continued
• Interface to request subsets and products
• Processed in delayed mode
• Individual network downloads
• Bulk data options
• Variety of magnetic tapes
• (DLT, SDLT, LTO, DDS3, Exabyte, ...)
• CDROMs
• DVDs
• Portable hard drives
• Handle on case by case basis

10
NLDAS 51-year hydrologic reanalysis (option)

• Yun Fan et al.
• 51-year, 1948-1998, land surface hydrology
  reanalysis
• 1/8
• Forcing NN reanalysis and CPC daily precip.
• Output
• Energy and water (mass) balance
• Snow cover and depth
• 4-layer soil temperature and moisture
• Continuation to 2002 likely.
• General Interest?

11
Conclusions

• Will have NARR at NCAR
• Propose to prepare GAPP specific service
• Much data online
• Augment with other model data, ERA-40
• Tightly couple with JOSS and in situ data
• May offer NLDAS 51-year hydrology reanalysis