[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20040105: question - HDS or CONDUIT


The HDS feed does contain both AVN/GFS and UKMET models
on the global thinned grids. The patterns are:
GFS:    HDS     ^H.[I-P]... KWBC
UKMET:  HDS     ^H.[I-P]... EGRR

The I through P designations are the 8 sections of the globe
which can be stitched together for the global view (as
is done for GEMPAK, McIDAS and the NetCDF decoder). These are 6 hourly
to 72 hour forecast. The regional GFS grids go out to 120 hours.
I do provide these patterns in the GEMPAK decoder templates
in $NAWIPS/ldm/etc/templates/pqact.gempak_decoders.in
if you want to look at current patterns.

The CONDUIT feed contains the GFS grid 003 through 180 hours, 
and grid 002 from 192 to 384 hours at 3 hourly forecast steps.
The CONDUIT grids are larger (since they are not thinned),
contain more parameters, and the added number of time
steps output make this a very large data set (about 8GB per day
for the 4 runs vs 800Mb for the thinned GFS on HDS).

Most meso scale models would probably benefit from the added parameters 
in the CONDUIT stream (such as radiation fluxes, etc not part
of the package sent on NOAAPORT/HDS). If you are not needing that type
of input, the thinned grids from NOAAPORT (eg the source of HDS)
provide the necessary synoptic fields, though not as far into
the future (the N. Hemisphere GFS grid does go to 120 hours on NOAAPORT/HDS
if you need longer periods to analyze waves).

Also, another option for global data is the NOGAPS feed on IDD from FNMOC.

Steve Chiswell
Unidata User Support

>From: Paul Ruscher <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200401051531.i05FVgp2027262

>Hi - I've been wondering what the various advantages or disadvantages 
>are of receiving NCEP model data via CONDUIT rather than HDS?  Is there 
>an FAQ somewhere that compares them?
>I like the idea of having this source, since we are still a tier 2 site 
>on IDD, but maybe there is no advantage?
>I have not taught synoptic in a few years and so have not been paying 
>attention to our receipt of the global models, and just realized that 
>we do not receive the 003 grid global model (1x1 deg) on our LDM feed, 
>so I'm trying to get our staff up to speed so that we have the global 
>model once again.  I've found the old pattern actions for the "thin" 
>grid but I expect that the old AVN thin grids are no longer being 
>transmitted via LDM.  The only information I can find about grid 003 is 
>through the CONDUIT stuff on the web.  Any enlightenment would be 
>appreciated, and a happy new year to all!
>TIA...Paul Ruscher