20020421: display of NEXRAD Composite Product loops

>From: Jim Koermer <address@hidden>
>Organization: Plymouth State
>Keywords: 200204210100.g3L10Za22943 FNEXRAD NEXRAD composite N1P NTP

Jim,

re: new N1P and NTP composites

>I came across them when I was trying to track down an increase in disk
>storage. No problem.

OK.

re: Do you have any suggestions on composites or frequency?

>Actually, the N1P data didn't look too bad with an animation at the
>6-minute intervals. It was only the NTP data that didn't show much at
>the high time resolution.

It pretty much shouldn't show much at high temporal updates since its
signature is the integration of precip over time.  The problem with
the storm total product is that the accumulation time is different
for every radar.  This means that the display is not a consistent one.

>I agree that NCR, NET, and NVL are the best
>candidates remaining for composites. It might be interesting to see the
>248NM base reflectivity (N0Z) as a composite to see if it fills better
>than the N0R product for a larger scale composite.

Since it is now really easy to generate composites to evaluate, I
will get together with Chiz to compare the N0Z composite with the
N0R one.  We have already looked at NETs and NVLs; they are not
exciting, but some may find them useful.

>I'd also like to see a BREF enhancement table that is much like the
>BREF24, but assigns no colors whatsoever for values less than 10 dbz.
>This would eliminate alot of ground clutter on the displayed composite.

You can easily generate this yourself in McIDAS.  I would:

o bring up McIDAS-X session
o load an N0R Level III product over a station that has ground clutter
o from the command mode, list the enhancement used:

  EU TABLE

o change the lowest level(s) to suit your preference

The trick part is when to use the other enhancement.  When the radar
is operating in clear air mode, half of the color table is used by
echos of less than 0 dBZ.

>I used this adaption of the BREF24 color scheme in my WXP-based single
>site NEXRAD data generator.

Regardless of the mode the radar is operating in?

>I have also noticed the the color bar on my McIDAS generated NEXRAD
>composite images is not always there, yet it is supposed to be there. If
>a do a loop, it is present on most images, but usually it gets dropped
>off one image or another. Is this a bug in McIDAS?

No, it is not a bug.

I am betting that the BAR for one or more frames is not getting the BAR
and/or MAP before the script that is doing the display exits.  When one
tells IMGDISP (imgdisp.k) to not only put up the image but also draw
the MAP and put up a BAR, you have to wait until those are done before
exiting.  The reason for this is that the actions in the REFRESH=
clause are processed asynchronously.  IMGDISP sends the commands off to
be processed and then moves on.  One of the ramifications of this is that
one has to be very explicit in the IMGDISP command to insure that the
items listed in the REFRESH= clause work on the intended frames.  A
command like:

IMGDISP NEXRCOMP/1KN0R-NAT STA=TWX MAG=-5 EU=BREF24 ALL=1 10 SF=YES 
REFRESH='MAP;BAR'

Will invariably not put the MAP or BAR on the correct frames.  A more explicit
version of the same command:

IMGDISP NEXRCOMP/1KN0R-NAT STA=TWX MAG=-5 EU=BREF24 ALL=1 10 SF=YES 
REFRESH='MAP GRA=(GRA);BAR GRA=(GRA)'

will as long as the session in which the IMGDISP invocation is running
last long enough for all of the commands to finish running.

If the image(s) that are not getting the BAR are at the end of the
loop, then adding a simple WAIT will insure that the processing in the
REFRESH= clause gets processed:

IMGDISP NEXRCOMP/1KN0R-NAT STA=TWX MAG=-5 EU=BREF24 REFRESH='MAP GRA=(GRA);BAR 
GRA=(GRA)'
WAIT 3

would work, but would cause more delay than you may want.  You could
replace this with serial invocations of the display, MAP and BAR:

IMGDISP NEXRCOMP/1KN0R-NAT STA=TWX MAG=-5 EU=BREF24 SF=YES
MAP
BAR

When the commands are run like this, they are invoked serially.  The
SF=YES tells IMGDISP to flip to the frame after the display has been
made.  This will insure that MAP will draw on the frame that the plot
was made in.

Depending on how you are running the commands to put up the loop, there
are different approaches to take to ensure that the MAP and BAR get
correctly displayed.  I could best advise you if I know which approach
you were taking.

Tom

>From address@hidden Sun Apr 21 13:55:33 2002
>Subject: Re: 20020421: display of NEXRAD Composite Product loops

re: chop off reflectivities < 10 dBZ no matter what

>Probably, I don't find the clear air mode reflectivities all that
>useful, so I don't mind chopping off values below 10 dbz, especially for
>a national display.

re: missing BAR on products loaded with IMGDISP

>I was using the following line in the script:

>/home/mcidas/bin/imgdisp.k NEXRCOMP/"$1" $2 "$3" "$4" REFRESH='EG;MAP H;BAR'
>and I was only waiting for 1 second. I put the map.k and bar.k serially
>as you have above and this did indeed take care of the occasional
>missing bar. Thanks. 
 
Jim
--
James P. Koermer             E-Mail: address@hidden
Professor of Meteorology     Office Phone: (603)535-2574
Natural Science Department   Office Fax: (603)535-2723
Plymouth State College       WWW: http://vortex.plymouth.edu/
Plymouth, NH 03264