Help for TES

TES uses the TES algorithm to calculate the emissivity and temperature
from an input dataset. The input dataset must be in radiance units of
Watts per square meter per steradian per micrometer.  This is different
from versions of the program in effect prior to November, 2004.  As a
consequence, input pixels should be real (floating point) values.

The output has one more band than the input. This extra band contains
the tes temperature (the last band), the other bands contain the tes 
emissivity.

Testing
This program was tested on an artificial 2x2x6 band radiance dataset. 

RESTRICTIONS;
1. Maximum size of one band of the input is 10,000 values.
2. Maximum number of bands in the input is 20.
3. If the input radiance is LE zero (mW/m*m/um/sr) the pixel is considered bad
   and not processed. Bad pixel emissivities are set to -1 and -2 respectively.
4. If the emissivity calculated from the regression is LE zero or GE 1.0, the
   pixel is considered bad and set to -3 and -4 respectively. 

HISTORY
Written by Simon J. Hook in FORTRAN.
27 Oct 2004  ...rea...     Modified to accept FULLWORD and REAL input pixels,
                           and remove the restriction of 20,000 for maximum
                           input pixel value.
                           HOT keyword added.
 5 Nov 2004  ...rea...     Modified to treat both input and output pixels as
                           unscaled values.  Output pixels are now real (rather
                           than scaled integer) values.