phoemplocal

    This program finds lunar-Lambert or Minnaert photometric functions
    to approximate a more realistic but more complex Hapke model. The fit
    is performed at a single geometry rather than for a range of phase 
    angles.  The user specifies the phase angle and the incidence and emission 
    angles of the mean ground plane (datum), as well as the RMS (root mean
    squared) slope relative to the dataum.  Artificial data are then created, 
    with slopes drawn from an isotropic Gaussian distribution relative to the 
    datum.  The simpler model is fit at these orientations to the Hapke model 
    by adjusting its one parameter and its overall brightness so that the 
    sum-squared-residual between the two is minimized.  Both the parameter 
    (which, for both types of simple model, mainly controls limb darkening) 
    and the brightness (normalized as an empirical phase curve) are reported.

    This program should serve as a clear example for developing a new program 
    to fit the parameter of an empirical photometric function to part of an
    actual image. The differences can be summarized as follows:
         1) Fit is done only once (not in a table vs. phase angle)
            since the input image has a fixed geometry.
         2) The data to be fit *to* are obtained from an ISIS cube plane
            rather than modeled in in the code.
         3) Similarly, the incidence, emission, and phase angles are
            obtained from backplanes of the same cube and maintained
            in buffers corresponding point-for-point with the image.
         4) Both image data and angles should be converted to double
            precision before being used.
         5) Optionally, a region-of-interest mask could be obtained
            from a backplane and only those pixels in the ROI used
            in the fit.  This would be in addition to the incidence
            and emission angle tests already done in the code.