20021107: AWIPS Grid 220.

>From: "Stonie R. Cooper" <address@hidden>
>Organization: Planetary Data, Incorporated
>Keywords: 200211071511.gA7FBXX23023

>Steve,
>
>I just noticed this again, with 5.6.h, the carry over of a misprojection of 
>AWIPS Grid 220.  There is only one product I know of in this projection - sea 
>ice - but it displays a square over Australia instead of a PS on the south 
>pole.  I had forgotten about it, so haven't dug into it . . . let me know if 
>you have a quick fix, otherwise I'll dig into next week.
>-- 
>Stonie R. Cooper
>Planetary Data, Incorporated
>ph. (402) 782-6611
>


Stonie,

The A220 grid is supposed to have corners (as defined in 
$GEMTBL/grid/grdnav.tbl and ncep's grib document) of:
-36.899;139.806   -31.850;-31.899

I wrote dcgrib/dcgrib2 to always use the GDS if present, otherwise to look
up the navigation in the grdnav.tbl file. Looking at the gdinfo
output, it appears that the corners are being computed as:
   -36.87;139.81 24.47;142.77

so, the GEMPAK grid file is being created with the wrong navigation.
The Upper Right (UR) corner point is incorrect. Probably a small bug in the
southern hemisphere computation of gds.latur/gds.lonur in gbpolr.c
or the angle/poinint of projection flags are not correctly set from the grib.

I'll have to look at it for 5.6.i. You might try nagrib with
cpyfil=#220.


Steve Chiswell