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

[IDD #ZSJ-835116]: missing gfs grids on NGRID



Hi Gilbert,

re:
> Well, now that I am Mr. Satellite Receiver Dude(tm), here's a question:
> What would it take to receive, and decode into useful form (into McIDAS,
> GEMPAK, etc), GOES-12/13 imagery direct from the birds? I do have a spare
> dish to use. And if NOAAport goes down, I'd like to have my own ground
> station as a backup. I would receive only from G12 (GOES EAST), of course.
> My guess is that it takes major computing power beyond what I have got,
> since the bird spews out IMMENSE amounts of data. I think NESDIS has some
> huge servers to handle it. But just in case, I'm curious. After all, the
> http://www.rap.ucar.edu/weather/satellite page displays the info from your
> ground station out there. So does NASA's infamous 1 km image page that's
> been running for years now. And is an ADDE server to boot!

We license and run the SDI software from SSEC (not cheap) for the GOES family
of satellites.  The component of hardware that one has to have is:

- high quality satellite dish of sufficient size (we use 3.8 m dishes)

- appropriate LNB (different from the one for NOAAPort)

- high quality coax from the LNB to the receiver

- satellite receiver (also not cheap)

- SDI card or new SSEC receiver unit (~$20K for the receiver unit)
  http://www.ssec.wisc.edu/mcidas

  SSEC used to keep prices online, but I couldn't find them a couple of
  minutes ago.

- SDI software (available through the McIDAS User Group)

re: SSEC's NOAAPort ingest is not getting all of NGRID for some reason

> Sounds like noise

Yup.

> or the machine can't keep up?

Nope.  The two NOAAPort ingesters that SSEC runs are overkill if anything.

> There's a lot of data
> flowing across when the models hit the feed. Now that I can see it, it's
> impressive. 

We were able to ingest all of NOAAport on two very old, dual PIII machines with
no problem (dedicated to NOAAPort ingest as are the SSEC machines), so the
modern, fast, loaded with RAM machines that SSEC has should be idling all
of the time.

> Those 11 mb/sec spikes at the peak are interesting, and I know
> that they can do 40 mb/sec if the NWS could afford it!

Right. The current bandwidth use is 10 Mbps.  Expansion is possible, but at a
cost.

re: LSU machines ingesting
> Which led me to my comment about the LDM working properly. If it was, it
> should have switched to LSU to get the missing data. Unless I am missing
> something here, probably am.

The problem was that our cluster accumulators were not requesting NGRID
from the LSU machine!  Hard to get data unless you request it :-O

re: redundant feed request from NSF/ATM turned off some time ago; noise?

> Gotcha. One thing I have learned about noise: A lot of it can be
> eliminated with the 20K or 15K noise professional quality LNBs.
> Broadcasters do not use anything with a noise rating higher than 20 K, and
> a frequency drift of +/-5 khz. A broadcast engineer has used the Norsat
> 3120 for several years with flawless reception, except when the dish gets
> hit by lightning. I don't know why THAT would cause a problem... ;-)

The other problem is terrestrial inteference.  One never can forget about TI!

re:
> I am still seeing "GB -1" errors every now and then...and nothing else.
> Very weird. But all the data is coming in.

I see this on all of the NOAAPort ingest machines that I have login access
to. It is a GEMPAK error that _may_ have to do with model products being
received that are not in GEMPAK's grib tables.  _IF_ (I say IF since I don't
know yet) this is the case, I will need to update the GEMPAK tables with
the appropriate information and make it available to all sites running
our NOAAPort code.  I am thinking that it may be related to new products
since I have not seen the new Wave Watch III grids come across the IDD/NGRID
feed, and the products were supposed to start flowing on the 29th.

re: automatic failover does, in fact, work properly

> That's good, although I'm still wondering why it didn't switch to LSU
> during those times. The "bizarre" light is blinking.

Operator error (me or Chiz at some point in the past) :-)

re: Please let me know if you have questions on anything above.

> "If?" :-)

How silly of me :-))

Cheers,

Tom
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: ZSJ-835116
Department: Support IDD
Priority: Normal
Status: Closed