20000125: GINI MERCator projection products

>From: Unidata User Support <address@hidden>
>Organization: Unidata Program Center
>Keywords: McIDAS ADDE GINI

Barb,

It was great to talk to you about GINI imagery issues at the AMS
meeting in Long Beach.  I want to take you up on your offer of
accessiblity when I have issues with the imagery navigation, but my
question is in the interpretation of values in the image headers.

As you may know, I have been developing McIDAS ADDE AGET and ADIR
servers for the GINI imagery data in NOAAPORT.  I was, in fact, using
these servers at the AMS show when you dropped by.  I have been having
problems using the header information in the Puerto Rico and Hawaiian
MERCator images in McIDAS.  In particular, I can't directly assign
values I unpack from the Puerto Rico and Hawaii sectors to McIDAS
navigation entries.

As a review, the McIDAS MERCator navigation module, nvxmerc.dlm, requires
input of 10 parameters:

 1) Name of the projection:: 'MERC'
 2) image line of the equator
 3) image element of the equator
 4) standard latitude [packed integer in DDDMMSS]
 5) spacing at the standard latitude [m]
 6) normal longitude [packed integer in DDDMMSS]
 7) radius of the planet [m]
 8) eccentricity of the planet, * 1000000
 9) coordinate type, >=0 -> planetodetic, < 0 planetocentric
10) longitude convention, >=0 -> west positive; < 0 west negative

My interpretation of these McIDAS navigation is:

 1) 'MERC'
 2) line of equator:: calculated from the image
 3) element of equator:: calculated from image: basically the mid point of
    the sector
 4) std. lat.:: this should be the latitude at which the projection
    cylinder intersects the earth; the value for this in the Hawaii
    and Puerto Rico MERCator sectors is 20.0000 N.
 5) spacing at std. lat.:: this is listed in Table 4.8 of the Interface
    Control Document (ICD) for AWIPS-National Environmental Satellite,
    Data, and Information Service (NESDIS) publication AA0130008 CH-2
    August 1, 1999 as 1.0000 km for VIS, 4.000 km for the IRs, and 8.000 km
    for Water Vapor at 20N for both the Puerto Rico and Hawaii MERCator sectors
 6) normal longitude:: this is image dependent; it is basically the
    middle longitude of the sectors
 7) Earth radius:: the ICD states that the model used in creating the
    sectors uses a spherical earth that has a radius of 6371.2 km
 8) eccentricity:: 0 for a spherical earth
 9) coord. type:: 0 - planetodetic
10) long. convention:: McIDAS convention is positive west

The problem I am having is that the standard latitude is unpacked from
the images as 20 N.  In order to make the McIDAS navigation work,
however, I have to tell it that the std. lat. is 2 N (i.e. the value in
the header divided by ten).

The second problem I have is the ICD declares the product resolution to
be 1.0000 km for VIS, 4.0000 km for IRs, and 8.0000 km for Water Vapor
at 20 N for both MERCator sectors.  In order to get McIDAS navigation
code to work, I have to set the spacing at the std. lat.  to be a
function of this resolution (which is octet 42 in the GINI header)
and the std. lat.:

space = imgres / cos(lat)

So, you might guess that I am somewhat frustrated that I can't gleen
correct values to use in McIDAS navigation from the values I see
in the GINI MERCator sector headers.

My questions to you are:

o am I crazy (a distinct possibility)
o is the "Latin" (octets 39-41) value in the GINI header incorrect
o is the McIDAS navigation module hosed

I guess that the easiest approach to getting my questions answered
would be for me to a quick overview of what and how these sectors are
produced.  A look at the code that is doing the job should tell me
what, if anything, is wrong with the values I see in the images.

Any help you or anyone else in your group could provide would be most
appreciated.

Tom Yoksas