Rich Signell wrote:
jJohn, But zeta *does* have a time dimension (it's the free surface, which in general changes with time). But for this specialized application we were trying to put in zeros for all three dimensions (lon, lat, time).
yes, the bug is that we arent correctly handling the time dimension for ocean sigma, so if you remove it you get a temporary workaround.
-Rich On Wed, May 13, 2009 at 10:21 AM, John Caron <address@hidden> wrote:anyway, a temporary fix may be: <variable name="zeta" shape="ny nx" type="float"> but i dont know the effect of the zero eta coords. i will fix the code after i check why my regression tests didnt fail.