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

20031007: NDFD feed

>From:  Jim Steenburgh <address@hidden>
>Organization:  University of Utah
>Keywords:  200310072033.h97KXCk1007985 NDFD data volumes

Hi Jim,

>    Just following up on NDFD stuff from the Usercomm meeting.  Tom's 
>info was pretty much on the mark.  The NDFD is updated hourly, although 
>many FOs only send in revised grids twice a day, so there's quite a bit 
>of repetition in the grids.

Sounds good.

>    Folks at the NWS tell me that the "subsectors" available on tgftp 
>are just that...sub sectors of the full CONUS.  Thus, if we send the 
>CONUS, we're pretty much covered.

I had a feeling that was the situation after looking at the volume
in the conus versus the sum of the volumes in the 16 sectors.

>Although many FOs send in their 
>forecasts on a grid that's higher res than 5 km...this information 
>doesn't get sent out, except after back interpolation to the 5-km grid.

Fred Branski (Bransky?) left a voicemail on Linda's phone that
said that the volume would quadruple in the next year as the resolution
is increased from 5 km to 2.5 km.   He also noted that the number of
fields (elements) would be increased from 14 to 30 by 2005 resulting
in an additional 3X increase.  He then mused about documents floating
around that had the numbe of elements increasing to up to 90, so
the volume could increase by and additional factor of 3 at some point
in the future.

>    I can't remember what Tom cited for data volume, but a quick look at 
>ftp://tgftp.nws.noaa.gov/SL.us008001/ST.expr/DF.gr2/DC.ndfd/AR.conus/
>suggests about 50 MB/h or ~1.2 GB/day.

The upper limit I came up with after looking at the conus and sector
directories was 200 MB/hour.  I arrived at this by taking the largest
field from one sector (about 1 MB) and multiplying it by 11 (the
number of fields that I saw) and then by 16 and then adding the 40 MB
seen for the conus.

>     Thus, I have two questions for Tom...are these numbers accurate?  

Yes, the appear to be.

>Am I missing anything?

I don't think so, but Linda is going to get Fred B. to FAX a document
that supposedly spells things out in a little more detail.

>For Chiz, can we add this sort of volume to 
>CONDUIT, or would something need to go?  Can it be put on a stream other 
>than CONDUIT?
>
>Thanks,

No worries.

Tom
--
+-----------------------------------------------------------------------------+
* Tom Yoksas                                             UCAR Unidata Program *
* (303) 497-8642 (last resort)                                  P.O. Box 3000 *
* address@hidden                                   Boulder, CO 80307 *
* Unidata WWW Service                             http://www.unidata.ucar.edu/*
+-----------------------------------------------------------------------------+