Real Time NOMADS <ftp2u> Client

(The NCEP component)

Some Nomads documentation is at:

http://nomad3.ncep.noaa.gov and http://nomad5.ncep.noaa.gov.
http://nomad5.ncep.noaa.gov/DOC/nomads_note1.html   -- for example, This is a HOWTO that includes GDS/DODS access. The file you are looking at is

http://nomad5.ncep.noaa.gov/DOC/nomads_note2.html

Here is the location of some general information:
http://wwwt.emc.ncep.noaa.gov/gmb/wd23ja/index1.html
under "presentations" are listed some links to conference reports on Real Time Nomads.
The NCDC archive pages for Nomads is at
http://www.ncdc.noaa.gov/oa/climate/nomads/nomads.html

Here is how some of NOMADS works (from an e-mail reply):
The ncep_data table
http://nomad3.ncep.noaa.gov/ncep_data (or change the 3 to a 5 when 3 is not available - like today)
does have a lot of datasets on it.  The ones toward the bottom of the table are from the real time operational models.  Two good choices for initial conditions are the entries NAM(Eta)
and GFS (rotating or archive of 2 months).  Over the CONUS the
NAM (Eta) EDAS at 12 km interpolated to the #218 lon/lat lambert conformal GRIB grid
is available but you mention the GFS.  The GFS files have a naming convention:
gfs.tNNz.pgrbfMM where NN=00, 06, 12, or 18Z cycle (model is run 4x/day) and
MM=00, 03, 06, .... out to 180 forecast times (hours).  In the <ftp2u> section you can (for any dataset)
check manually the GRIB files you want or enter, for example, in the "select files" box "gfs.t00z.*" to get
all the 00Z cycle files checked. You then scroll down and check the levels needed and the variables you want, like  UGRD, VGRD (the wind) , TMP (temperature) and RH, 3-D fields and PRMSL, the surface pressure (2D) or anything else you need.  You can also check the subset box if a subset of the global data is needed, usually so for mesoscale applications. Enter the lon/lat corners of the area you need.  If you have an ftp server you can enter its IP address and other info and the repackaged GRIB file
will be sent there or you can use our ftp server and use standard (IE) web services to get the data from us.  The URL produced by this process, when successful, is available (copy-paste it) to use again in a unix script as the quoted object of an interactive web download program like wget or wwwgrab for example, and you can re-issue the URL with any changes to get other data on other days.  One can also insert such a script in cron for automatic data flow.

see also
http://www.cpc.ncep.noaa.gov/products/wesley/fast_downloading_grib.html
for fast ftp and how to use the <http> button in the ncep_data data table.