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

19990426: eta ?

Bob,
Nominally, the ETA is now run on the 32km grid 4 times per day,
rather thanprior to last June or so when it was a separate
model - 48km at 0,12Z and 29km at 3Z and 15Z.

I'm told that the 32km grid was a compromise for computer time/size.

Currently the model is run at 0Z, 3Z, 12Z and 18Z. 
The goal is to move the 3Z run to 6Z, but they have some processes
that currently depend on a 3Z run.

The 211 grid (Q) is the 80km interpolation from the 32km grid
to the AWIPS projection for NOAAport/FOS. AWIPS requires certain
grid projections to fit under their already remapped imagery.

The 212 grid R is the 40km interpolation, and the 215 grid U is
the 20km interpolation. The 20km interpolation is somewhat fictitious
since the model is only a 32km model...but they output primarily surface
and boundary layer parameters on that grid. The 20, 40, and 80km grids
are basically all the same lambert projection withsimply double
and quadruple the 80km grid respectively. The "off" time run(eg off 0Z
and 12Z) are only output through 30hr forecasts on NOAAport while
the 0Z and 12Z runs output through 48 hours.

NOAAport carries models on several projections for AWIPS since they
don't "compute" projections from the gridpoints, they just distribute
their imagery on the matching projections. This is why you will find 5 
projections of the same NGM and MRF runs on NOAAport. This sounds
pretty silly given bandwidth considerations- but thats the AWIPS
way.

For higher resolution ETA runs, I have setup an IDD feed from NCEP 
using data that they post for FTP on their servers. This includes
the ETA 212 40km grid 4x per day, and includes levels and
parameters not available on NOAAport. This feed is decribed under the
CONDUIT section on our WWW pages (and at:
http://www.unidata.ucar.edu/staff/chiz/ncep_idd/).

The 32km ETA is run on 50 vertical levels. Eventually, they want
to run ETA on a rapid update schedule similar to RUC.
When NCEP has computer crashes, they usually have to revert back
to a 48km or 80km ETA run, or have to scratch the run altogether if
they are very backlogged. NCEP still produces 20km, 40km and 80km grids,
however, the interpolations from the runtime domain to the distributed
grids can obviously be degraded if they have to use the 80km run
from Cray 5 when Cray 3 crashes. To keep up with daily runtime
information (which is useful for knowing what is happening with
the realtime data feed), consult the NOUS42 KWNO bulletin.

NWS has been approved to purchase a new computer- hopefully that will
ease some of the size/time restrictions that exist on the current
resources.

Steve CHiswell
Unidata User SUpport



>From: "Bob Creasey" <address@hidden>
>Organization: .
>Keywords: 199904262239.QAA27210

>I am trying to understand what we get with regard to the eta model data...now.
>
>HRS    ^[YZ].Q... KWB.   80km eta
>HRS    ^[YZ].R... KWBE   40km eta (3,18z pressure sfc)
>HRS    ^[YZ].R... KWBD   40km eta (00,12z sfc only)
>HRS    ^[YZ].U... KWB.   20km eta (3,18z sfc and pagl)
>
>Which model run generates the grids?  Does the 80km and 40km grid come from th
> e
>48km model run?  Does the 29km models run generate the 20km grid?  Are there
>more changes on the way with eta?
>
>As always...thanks for your time and help!!!!
>
>         ////
>        [^ ^] Bob
>        (\_/)
>
>
>
>
>-- 
> Bob Creasey                         (  +*))   (+ * )_ 
> Naval Postgraduate School          ((* - * )(++,, , )) 
> Dept of Meteorology Rm 254 MR/CY (( ,- *-  ,, -,- ), ))__
> 589 Dyer Rd.                    (_______________________))
> Monterey Ca 93943-5114              \   \     \\\\\\\    
> address@hidden               / \  /\_    \\\\\\   
> Work Phone: 1-831-656-3178           / / / \    \\\\\      
> DSN: 878-3178 FAX: 1-831-656-3061      
>