20010405: sporadic problem creating 500 mb U,V grids

>From: "Paul L. Sirvatka" <address@hidden>
>Organization: COD
>Keywords: 200104060312.f363CNL24685 McIDAS-X RAOBCON

Paul,

Sorry I couldn't get to this before today.

>sorry to do this but I keep finding problems.

No problem.  If there really is a problem in code, I want to know about
it so it can get fixed.

>I have the following in a script which works sometimes.
>
> REM SFC-500 MB Shear; Total Totals
>
>REM Create grids 1-4 in gridfile 22
>IGU DEL 22
>RAOBCON STREAML 500 NA OUT=SAVE MYDATA/GRIDS.22
>RAOBCON STREAML SFC NA OUT=SAVE MYDATA/GRIDS.22
>
>WAIT 4
>
>
>REM Create Grid 5 which is U 500 - U sfc
>GRDCOPY MYDATA/GRIDS.22 MYDATA/GRIDS.22 DGRID=5 G1='PARAM U; GRID 3'
>G2='PARAM U
>;GRID 1' MATH='G2-G1' NEWPAR=U KTS SHR
>
>REM Create Grid 6 which is V 500 - V sfc
>GRDCOPY MYDATA/GRIDS.22 MYDATA/GRIDS.22 DGRID=6 G1='PARAM V; GRID 4'
>G2='PARAM V
>;GRID 2' MATH='G2-G1' NEWPAR=V KTS SHR
>
>REM Sfc - 500MB Wind shear wind barb
>GRDDISP MYDATA/GRIDS.22 10 MAP=#UBASEMAP LEV=SHR PAR=WINDB OUT=PLOT
>COLOR=4 LSIZ
>E=8 PIN=2 2
>
>REM Contour high shears ( >50kts )
>TE SHR "50 70 90 110
>GRDDISP MYDATA/GRIDS.22 10 MAP=OLAY G1='PARAM U; GRID 3' G2='PARAM U;GRID
>1' G3=
>'PARAM V; GRID 4' G4='PARAM V;GRID 2' NEWPAR=SHR KTS SHR
>MATH='SQRT((G2-G1)**2+(
>G4-G3)**2)' FORMAT=I3 COLOR=3 CINT=SHR DASH=ALL

OK, I put this into a BATCH file on my local machine and ran it for
today's 0 and 12 Z data.  The output looked reasonable.

>This did not work at 0Z on April 6. When I plotted the grid that had the
>500 mb U component, I had values over 300 kts! It seems that 
>
>RAOBCON STREAML 500 NA OUT=SAVE MYDATA/GRIDS.22
>
>gives me problems.
>
>Any ideas on this one?

I tested this out by modifying the RAOBCON invocation lines at the
beginning to:

RAOBCON STREAML 500 NA 0 2001096 OUT=SAVE MYDATA/GRIDS.22
RAOBCON STREAML SFC NA 0 2001096 OUT=SAVE MYDATA/GRIDS.22

And pointed to ADDE.UCAR.EDU.  Since the output from this also looked good
I am left with the impression that the problem my relate back to the dataset
you were pointing to when you ran the BATCH file.  Perhaps there is some
sort of a decoding problem (which may actually be a data ingestion problem,
but who knows) on the ADDE server you were pointing at when you ran the
BATCH file?

If you see this happen again, please do me a favor and rerun the script
while pointing (DATALOC ADD RTPTSRC, that is) at ADDE.UCAR.EDU.  If
the script runs OK while pointed at that server, then the problem must
be in the data ingestion/decoding side on the other server.

Along the same lines, but more closely related to a previous exchange we
had, I want to pass along that I learned how to correctly specify 
selection information in the definition of TT (Totals Totals) in the
RAOBCON context file(s).  As you will remember, we couldn't figure
out how to do this when you were first trying to plot TT.  The correct
syntax for specifying TT in a context file is:

TT   UNIT=C \
     FORMAT=F5.1 \
     MATH='P1+P2-2*P3' \
     IRAB='P1=T P=850' 'P2=TD P=850' 'P3=T P=500'\
     RAOB='P1=T P=850' 'P2=TD P=850' 'P3=T P=500'

The selection information (P=850 and P=500 in this example) are added to
the parameter definitions as keyword sequences that are not separated
by commas.

Since neither of us could figure this syntax out, it was decided that
the TotalTotals index example should be included in the global context
file RAOBCON.CORE so that others could see how to specify the selection
clauses.

I suggest, however, that the name of the parameter be changed from TT
to TOTTOT or something else that looks more like totals totals, Do you
have a strong preference?  Hopefully you will look at this today
because today up until sometime tomorrow is the window I have for
getting this added to SSEC distributed version of RAOBCON for 7.8.
Just to make sure that something gets in, I will add it as TT for now
with an eye towards changing it later (if you chime in with a different
name).

Tom

>From address@hidden Sun Apr  8 14:39:05 2001
>Subject: Re: 20010405: sporadic problem creating 500 mb U,V grids 

Thanks...

For the probelms I will let you know when I see it again and will do the
appropriate tests.

TOTTOT is fine or maybe TTOTALS...whatever works for me.

TTYL (Not TOTTOTYL):-)

Paul