[ferret_users] difficulty comparing variables because of time axis mismatches

[Keith Lindsay <klindsay@xxxxxxxx>]

Hi,

I am sometimes unable to get ferret to compute TEMP[d=2]-TEMP[d=1] where 
the datasets have different time axes. In some instances I am able to 
get ferret to compute the difference by applying the transformation 
gt=time@asn to one or both of the quantities. But this doesn't always 
work for me, particularly in the presence of variables with non-COARDS 
dimensions.

I've placed example problematic files A.nc and B.nc on ftp.cgd.ucar.edu 
in /pub/klindsay/ferret.

Executing "set data A.nc ; set data B.nc" yields the output
 *** NOTE: Unspecified or unsupported ordering of axes in variable MOC
 *** NOTE: The default ordering will be used
 *** NOTE: Axis TIME marked as TI
 *** NOTE: but it has different directions in grids. Axes in grids may 
be inconsistent.
 *** NOTE: Unspecified or unsupported ordering of axes in variable MOC
 *** NOTE: The default ordering will be used
 *** NOTE: Axis TIME1 marked as TI
 *** NOTE: but it has different directions in grids. Axes in grids may 
be inconsistent.

The command "stats TEMP[d=2]-TEMP[d=1]" yields
 *** NOTE: Ambiguous coordinates on E axis: TEMP[D=2]-TEMP[D=1]
 **ERROR: illegal limits: TEMP does not exist at E=01-JAN-0052 00:00:00
          Axis extremes are E=01-JAN-0052 00:00:00

I am unable to compute the difference using transformations on the time 
dimension that I can think of.
I'm looking for suggestions to enable me to coax ferret into computing 
the difference.
I am using v6.84 of ferret.

Thanks,
Keith Lindsay