CRAFT Configuration


Dear Tim,

I wanted to follow up on your message earlier this morning concerning the
strategy for bringing in the 88D wide band data.  (For those seeing this
for the first time, I repeat Tim's message below.)
I need to double check after last week's email exchange that ONENET
will connect to the 8-port RIDDS hub and bring the uncompressed/unaltered base
data stream to Norman and any LDM processing will be done in Norman.  I want
to be able to tell the Southern Region folks what the connection involves and
that you are not looking to install LDM PCs in the forecast offices at this
time.

Thank you.
Tim

As I noted earlier today, the above assumption is correct.  However, I hope
we haven't lost our flexibility in this regard.  The strategy described
above would prevent us from compressing the data at the radar site, and
thus would preclude us from using much cheaper and lower bandwidth links,
as discussed in our meeting two weeks ago.  In other words, we'd be running
a full T1 into every radar link when only a 56K line might be required.  I
don't know whether the Regents are going to go for this...they're certainly
willing to pay the T1 costs if necessary, but if we put a PC running LDM
at each radar site, we would have a more effective, sensible system because
we could compress the data before its shipped out.  My understanding is that
this particular solution is not attractive because it would require each
local NWS office to manage yet another computer.  But Tim, I'm wondering if
we could present a counter-argument about the benefits each office would be getting from the data/analyses/forecasts, not to mention the use of another PC
as a web browser, etc.  I realize these aren't easy issues, and that space
is another consideration.  But I would like to suggest that we contact the
WFOs in question to get their response.  I just can't see the sense in
shipping huge volumes of data when it's not necessary.  Indeed, if CRAFT is
to be a viable blueprint for the future, the proposed strategy would certainly
be identified as its weakest element, particularly in the eyes of funding
agencies from whom we'll be seeking support for research in distributed
computing, data compression, etc.  Thanks for considering my thoughts!

Sincerely,

Kelvin




--------------------------------------------------------------------------
Prof. Kelvin K. Droegemeier, Director        E-mail: kkd@xxxxxx
Center for Analysis and Prediction of        Phone:  405-325-0453
 Storms                                     FAX:    405-325-7614
University of Oklahoma                       WWW:    http://www.caps.ou.edu
Sarkeys Energy Center, Rm 1110                       http://hubcaps.ou.edu
100 East Boyd Street
Norman, OK  73019  USA