20020212: reading/writing McIDAS area files for use in GEMPAK

>From: David Ovens <address@hidden>
>Organization: University of Washington
>Keywords: 200202121913.g1CJDNx21764 McIDAS AREA

David,

Chiz already responded to you on this from a GEMPAK perspective, but
I figured I would throw in my 2 cents worth on the McIDAS side.

>I am hoping to create a bogus McIdas area file -- basically a GOES-10
>image rotated up so Seattle is about at the center of the image -- so
>that I can get a different satellite projection in GEMPAK (if you want
>an example of what my output would look like, see 
> http://www.atmos.washington.edu/~ovens/loops/wxloop.cgi?joes_okc_h500+all
>).  Anyhow, I think that all I need to do is figure out how to read an
>existing McIdas file (GOES-10) and write it back out changing only the
>4th word in the navigation section from "102042" to "474500" or
>something like that.  Is there FORTRAN (preferably), C, or PERL code
>out there that could help me with this?

This can easily be done in McIDAS.  You can use the LWU command to list
out and change values in any file.  So, for instance, let's suppose
that the image you want to play around with is in file AREA1234.
Use LWU from a McIDAS session to list out the first 64 words of the
file:

LWU LIST AREA1234 0 63

Now, suppose you wanted to change the value of word 123 to 474500:

LWU POKE AREA1234 474500 123

Remember that the index into the file that LWU will use is zero based.

As a side comment, the GVAR navigation model is so complex that I don't
think that changing the value of one word in the NAV block will do the
rotation that you want.  But hey, I have never tried this so go for
it!

After changing a NAV value, you will need to redisplay the image
and then draw a map on it to see the effect:

IMGDISP MYDATA/IMAGES.1234 REFRESH='EG;MAP'

If this fails, it means that you can not simply change one NAV parameter
in the GVAR navigation and be successful.

One navigation that does lend itself to simple rotations in McIDAS is
that for METEOSAT.  The subsatellite longitude is called out explicitly
in METEOSAT imagery.  Changing it effectively rotates the image to
the new longitude.  What is not called out, however, is the subsatellite
latitude as that is assumed to be the equator.  The only problem for
you using a METEOSAT image is that GEMPAK does not support METEOSAT
navigation.

>By the way, I have looked at maknav and imgremap in McIdas-X and I
>understand the LAMB, MERC, PS, RADR, RECT, and MOLL projections, it's
>possible that SIN, TANC, or NOAA does what I want, but I don't
>understand those.  Basically, I didn't see how I could do a new
>satellite projection if I did not already have a file with the
>projection that I want to create.

IMGREMAP can be used to write image data into a new AREA file
specifying the projection that you want.  The projections
supported by IMGERMAP for new output files are: 

     =LAMB slat1 slat2 slon | Lambert Conformal, standard latitude
                              and standard longitude (slat1 def=30,
                              slat2 def=50, slon def=center longitude)
     =MERC slat | Mercator projection and standard latitude (slat def=0)
     =MOLL slon | Mollweide projection and standard longitude (slon def=0)
     =PS slat slon | Polar Stereographic projection, standard latitude
                     and standard longitude
                     (slat def=60, slon def=center longitude)
     =RADAR rot | Radar projection and rotation angle (rot def=0)
     =RECT | Rectilinear projection
     =SIN | Sinusoidal Equal Area projection
     =TANC slat slon | Tangent Cone projection, standard latitude and

If you want the navigation of the resultant image to be that for GVAR,
then your only option is to remap into an existing image.

I am curious to hear the results of your changing NAV parameters in
a GVAR image!

>Thanks for any help or suggestions.
>
>David
>David Ovens            e-mail: address@hidden
>(206) 685-8108          plan: Real-time MM5 forecasting for Pacific Northwest
>Research Meteorologist
>Dept of Atmospheric Sciences, Box 351640
>University of Washington 
>Seattle, WA  98195

Tom Yoksas