[netCDFJava #QAX-163810]: Data Value Discrepancies Between Netcdf 4.2 and 4.3

Kevin,

When you say that the "altitude levels" are flipped, what exactly are you 
referring to?  I've opened up the file in ToolsUI version 4.3, and I'm looking 
at the coordinate axes.  NetCDF-Java recognizes a GeoZ coordinate with sigma 
units, midpoint values 50.0-1.0 in steps of one.  This is consistent with 
version 4.4.3 (current development branch) and 4.4.2 (current alpha release).  
Also, your initial e-mail refers to 'uComponentOfWind', while I'm seeing 
u_component_of_wind_height_above_ground u_component_of_wind_hybrid 
u_component_of_wind_tropopause.  Did you mean one of these three, or are these 
the "new" names that have been disambiguated?

Unfortunately, I don't have a branch of NetCDF-Java 4.2 to compare...it has 
passed out the back end of my git repository when I got a new machine, and I 
don't see a tag for it on Github anymore.  If necessary, you could upload your 
4.2 jarfiles to the same dropbox link I previously gave you for me to try out.

Of course, the behavior initially manifested when using NetCDF-Java to copy 
files, not necessarily view them, so I may not be reproducing the problem 
correctly.

-Lansing

Ticket Details
===================
Ticket ID: QAX-163810
Department: Support netCDF Java
Priority: High
Status: Open