areatonetcdf for Meteosat-8

Tom W.,

Sorry to have missed you this year at the MUG meeting...hope you are doing well.

A question....we are about to start processing Meteosat-8 here, and Doris' 
group is interested in viewing the data in AWIPS, as well as McIDAS.  I'm told 
that they are using your area-to-netcdf program for Meteosat-7 data.  Do you 
know if it works with Meteosat-8 AREAs, as well?

I'm still a little confused as to why Doris is using your program rather than 
the ADDE conversion code in IMGCOPY,  but it may have something to do with 
different "flavors" of netCDF.

Any ideas?

Thanks,

Bryan Batson
281-853-3238


From address@hidden  Wed Nov 10 08:10:26 2004

Hi Bryan!

Hey, sorry I missed you, too...and Doris.  Especially since it was her "Last 
Hurrah" before getting hitched...

By "Meteosat-8" do you mean "MSG"?  If you do, then the answer is that the 
current release does not have the navigation code for that; however, I can 
make a new release if it's needed since I've just released the Java library 
MSG navigation (although there is some controversy about this since the 
EUMETSAT ADDE server puts different data into the nav block than the SSEC 
server apparently).  For the AWIPS output, though, the navigation may not be 
needed, since the input AREA is a re-mapped image into an AWIPS projection.

The reason they're using areatonetcdf is because IMGCOPY does not yet write an 
AWIPS formatted netcdf file.  There is a server developed at Huntsville that 
does (I think they use it for MODIS?).  If you want to go that route, you 
might contact "them" (not knowing who "them" is, but I do know that Russ 
Dengel has been working on-and-off on this for our ADDE server suite).

So...that's the scooper -- hope all's well down your way!

tom

From address@hidden  Wed Nov 10 09:12:35 2004

Thanks Tom.  There still seems to be some confusion here (at least on my
part) about the difference between what IMGCOPY does and what your code
does.  Regardless, I think we would like to get ahold of the
MSG-friendly version of your code.

Thanks,

Bryan

From address@hidden  Wed Nov 10 09:15:56 2004

Bryan:

The AWIPS NetCDF files must be in a particular format (attributes, variables, 
etc) in order for AWIPS to display the data.  AFAIK, the current ADDE netcdf 
writer in MCIDAS-X does not produce netcdf files that are in that format.

I'll update the areatonetcdf this week and let you know -- might not be til 
early next week, though...

tom

From address@hidden  Wed Nov 10 09:18:28 2004

OK. No sweat.   Is it your impression that your areatonetCDF code
generates AWIPS format data?  I ask this because one of Doris' folks
said this: 

The netCDF images are created
with a java script (AreaToNetCDF), which I think was written by Tom
Whittaker at SSEC.  Also, the netCDF files that are written out are not
necessarily "AWIPS friendly".  We had to do some navigational
manipulating
on AWIPS to get them to display.

Thanks,

Bryan

From address@hidden  Wed Nov 10 09:52:14 2004

Bryan:

First off, I use "java" not "java script" (I wish the person at Netscape, Inc 
who coined "java script" would have chosen another name).

Second, areatonetcdf output is strictly for the remapped AREA files created 
from McIDAS.  Originally, there were only 2-3 projections that were "canned" 
in AWIPS and we got the remap "template" AREAs from the NWS for our work. 
They are correct that if you're using some other projection, then you have to 
manipulate the AWIPS settings...but it has nothing to do with areatonetcdf per 
se, it has to do with the remapping in McIDAS.

the reason for all this is apparently the navigation info that one can put 
into netcdf files for AWIPS was ignored for the standard projections (which 
was what was used when we wrote this)...i don't know if it's still ignored or 
not, but the "bottom line" is there are no community conventions for storing 
images in netcdf (unlike gridded data), and so we could not, for example, put 
the GOES O&A data into the file and hope it would mean anything.

If new conventions have been adopted by FSL for AWIPS image NetCDF files, then 
perhaps the new ADDE write server Russ is working on (or the one done a HSV) 
can make use of that information -- I just don't know (the areatonetcdf was 
created over 5 years ago and not really touched since then except to add 
updated navigation modules).

tom