From gokhan@ncar.ucar.edu Thu Aug 29 17:56:40 2002
Date: Thu, 29 Aug 2002 11:46:42 -0600 (MDT)
From: Gokhan Danabasoglu <gokhan@ncar.ucar.edu>



     caveat emptor:

  - The ice fraction data sets are filtered to have a minimum
    ice fraction of 1%. This may eliminate any legitimate values.

  - The data sets contain "filled" ice fraction values over
    land on the given 1x1 grid mask. This filling appears
    to be problematic. When interpolated on to the model
    grid, one needs to carefully check the ice fraction data
    sets for the presence of ice in unusual places
    (e.g. Mediterranian).