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

[McIDAS #MUB-377647]: Converting MciDAS to NetCDF navagation issue



Hi Dave,

re:
> Thanks for taking a look at my problem.

No worries.

re:
> The reason I am looking into
> the McIDAS AREA vs.NetCDF issue,

I have been interpreting your emails to mean that you are investigating
the difference in displays between images converted to netCDF by two
different means: McIDAS-X and AreaToNetcdf.jar.  Unless I have missed
something (entirely possible, and I apologize if I have), I have not
seen a (graphical) comparison between the image displayed directly
from an AREA file and the image from the netCDF file created by
McIDAS-X.

Questions:

- have I missed something?

- have you made this comparison?

My working hypothesis is that the image converted to netCDF done by
McIDAS-X will be identical to that in the original AREA.  If this is
truly the case, then my suggestion would be to use netCDF versions of
the images you are interested in that are created by McIDAS-X.

re:
> is that I did a case study of
> hurricane eye/CDO using AREA files for a couple storm and now want to
> expand it to many storms.  Since my software (NCL) will not read AREA
> formatted files directly, I would like to use the CLASS NetCDF versions
> directly.  However, in checking for consistency between the data
> formats I came across this location discrepancy.  I am now wondering
> which navigation is correct before I choose a file format.

To determine which netCDF file has the correct navigation, I think that
a comparison between the display of the image from the original AREA
file and the image in the netCDF file created by McIDAS-X is needed.
This comparison can be done quantitatively using the IDV.  I will try
to get to this test sometime today and send you my results.

re:
> -I have put images and animations of the 2 versions of CLASS files onto
> the web page I created.  I have plotted out the data on a Satellite
> projection as well as a CED.  (You can see the small NW shift in both).
> 
> http://www.atmos.albany.edu/facstaff/vollaro/mcidas.html

OK.  What I do not see is a display created directly from the image
in the AREA file.

re:
> Q1. You indicate that the Navigation is specified in the lat/lon
> arrays?  How do I retrieve this info?  Using a ncdump -v lat file.nc
> I do not see any navigation info in this array
> 
> ie float lat(yc, xc) ;
> lat:long_name = "lat" ;
> lat:units = "degrees_north" ;

I just did an 'ncdump' of each of the two netCDF files you provided
and one I created using McIDAS-X directly.  The netCDF file that I
created using McIDAS-X was effectively identical (meaning that the
data and lat,lon values were the same; while things like the netCDF
creation date/time were not).  This strongly indicated that the netCDF
files available in NOAA/CLASS were produced by McIDAS-X.  This was not
a surprise since NOAA/SATEPS folks are members of the McIDAS User's
Group (MUG) that UW/SSEC operates.

re:
> However, I do see something in the startLine/Elem variables(
> ie   int startLine ;
> startLine:long_name = "starting image line (in satellite coordinates)" ;
> 
> If the Line/Element variables indicate the navigation, then you are
> correct that the conversion from Area to NetCDF does not change the
> navigation.  A ncdump of this file indicates:
> 
> int start_line ;
> start_line:units = "satellite coordinates" ;
> start_line:long_name = "image starting line" ;
> 
> The fact that the lat/lon ranges are slightly  different leads me to believe
> that it is not a plotting issue. When dumping the corner lat/lons  I get:
> AREA.nc                      NETCDF     diff
> 45.3134                    45.24942    .064   (1,1)LAT
> 89.9080                    89.8561      .0519  (1,1)LON
> .....
> 14.32536                 14.28732     .038   (674,1076)LAT
> 61.9835                   61.9640       .0195  (674,1076)LON
> 
> The startLine  and Elemant are the same for each version

I completely agree that the image represented in the netCDF created
by McIDAS-X and the one created by AreaToNetcdf.jar are different.
The question on the table is which one is correct.  I believe that
this can be determined by comparing the image in the original AREA
and the one in the McIDAS-X-created netCDF.  Again, I will do a
quantitative comparison using the IDV sometime today and let you
know my results.

Aside:

The other difference I see in the GIFs of netCDFs you provided is a
slight, yet noticeable difference in data values.  This is possibly
an indication that there is some interpolation being used in the
AreaToNetcdf.jar-created netCDF.

re:
> Thanks again for all of your help...

No worries.  This issue is, at least, interesting :-)

More later...

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: MUB-377647
Department: Support McIDAS
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.