heimbach@mit.edu 07-Nov-2002.
This verification experiment is almost identical to the
global_ocean.90x40x15 experiment, except that it uses
the exf (and cal) package instead of the 
external_fields_load routine.
To obtain identical results between the two experiments,
the following 2 modifications are necessary:
1. in external_fields_load.F:
   replace the line
      Imytm=int(myTime*rdt+0.5)
   by
      Imytm=int((myTime+1296000.)*rdt+0.5)
2. in exf_set_climsst.F:
   comment the 'quality' check, i.e. avoid the setting
   to freezing temperature if less than climtempfreeze.

================================

menemenlis@jpl.nasa.gov 05-Aug-2003

Input-field spatial interpolation has been added to pkg/exf.
It is enabled by defining CPP option USE_EXF_INTERPOLATION
in EXF_CPPOPTIONS.h or in ECCO_CPPOPTIONS.h
Both bi-linear and bi-cubic interpolation schemes are supported.

This package is a placeholder until a more general coupler
is made available by the ESMF project.  The output grid can be
arbitrary (cubed-sphere should be OK), but it is assumed that
the input grid is Cartesian with arbitrary latitude increments
in the y-direction and with equidistant longitude increments in
the x-direction.  The input grid must encompass the complete
output grid in the y-direction (i.e., extrapolation is not
supported).  It is also assumed that the grid is periodic in
the x-direction; to use the interpolation routine with a
non-periodic domain, make sure there is sufficient padding
at the edges, i.e., two points for bicubic and one for bilinear
interpolation.

The verification/global_with_exf experiment has been modified
to use the USE_EXF_INTERPOLATION option, the input grids
being defined by variables *_lon0, *_lon_inc, *_lat0, *_lat_inc
*_nlon and *_nlat in input/dat.exf and input/data.exf_clim

   *_lon0, *_lat0    :: lon and lat of sw corner of global input grid
   *_lon_inc         :: scalar x-grid increment
   *_lat_inc         :: vector y-grid increments
   *_nlon and *_nlat :: input x-grid and y-grid size

In this particular example the input and output grids are
the same, so the results of the verification experiment remain
unchanged from before.