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

[GEMPAK #XFV-201111]: dcgrib2 precision



Hi Brian,

I've been playing around with this and while I'm not certain I have a useful 
solution or workaround, I can say that DCGRIB2 is decoding the LAPS grib2 files 
correctly.

Rather than plot the function PVOR(THTA,WND) I've plotted the layer difference 
LDF(THTA) for 300:200. In the first image attached (GEMPAK_thta.gif), contoured 
on top of this layer difference is THTA @ 300 (black) and THTA @ 200 (white).  
It's clear to see that the discontinuities in the layer difference field appear 
because of the value of THTA at these two levels.  

Compare to an image of 300:200mb difference for THTA produced by the IDV 
(IDV_thta.gif) using the LAPS grib2 file and you see the same artifacts (you'll 
have to ignore/excuse the white block artifacts - I'm not sure why these 
appear).

So it appears DCGRIB2 is doing it's job correctly, but that answer doesn't help 
you much since the layer difference for THTA doesn't look natural plotted in 
either GEMPAK or IDV.

What I can suggest is you play around with different values for IJSKIP, FINT 
and CONTUR which should help you lessen the visual impact of these "noodles".

What software package are you using which produced PVOR plots (and I should ask 
THTA difference plots) without these artifacts? 

Michael James
Unidata



> Hello!
> 
> I'm using GDCNTR with these parameters:
> 
> GDATTIM  = last
> GLEVEL   = 300:200
> GVCORD   = pres
> GFUNC    = pvor(thta,obs)
> GDFILE   = 081432000.gem
> CINT     = 2
> LINE     = 3
> MAP      = 1
> MSCALE   =
> TITLE    = 1
> DEVICE   = xw
> SATFIL   =
> RADFIL   =
> IMCBAR   =
> PROJ     =
> GAREA    = grid
> IJSKIP   =
> CLEAR    = yes
> PANEL    = 0
> TEXT     = 1
> SCALE    = 999
> LATLON   =
> HILO     =
> HLSYM    =
> CLRBAR   =
> CONTUR   = 0
> SKIP     =
> FINT     = 2
> FLINE    = 30-8
> CTYPE    = f
> LUTFIL   =
> STNPLT   =
> 
> and the plot is showing what I can only describe as "noodles" of potential 
> vorticity.
> Another plotting program used the grib data file directly and didn't produce 
> this
> effect.  Does dcgrib2 maintain the precision of the original grib file?
> 
> The grid is a LAPS grid nest, and I put the grib file and my dcgrib2 decoded 
> .gem
> file here in case you want to take a look:
> 
> http://www-frd.fsl.noaa.gov/mab/jamison/tmp/
> 
> 
> Please let me know, thanks!
> 
> Brian Jamison
> 
> 
> 
> 


Ticket Details
===================
Ticket ID: XFV-201111
Department: Support GEMPAK
Priority: Normal
Status: Open

Attachment: GEMPAK_thta.gif
Description: GIF image

Attachment: IDV_thta.gif
Description: GIF image