[CF-metadata] ungridded data

Jonathan Gregory jonathan.gregory at metoffice.com
Tue Jun 25 09:55:09 MDT 2002


Dear Bob

Thanks for your further comments and definitions. Yes, I agree with you.
Structurally, a horizontal grid that is represented as 1D because lat and lon
are not independent looks the same as irregularly scattered data. But, as you
say,

> Conceptually, a horizontal grid represents a discretization and
> structuring/ordering of lat-lon space.

whereas irregularly scattered data do not have an natural order and are not
constructed in such a way as to discretize the entire horizontal area. In both
cases, there may or may not be boundary information.

You having put it that way, I would suggest again that the presence of the
grid_mapping attribute tells you which case you have. This attribute indicates
how the horizontal space has been discretised. It therefore also implies how
the boundary lines should be drawn, and the connectivity information of the
points. Would it be sufficient to rely on the application to know how to
understand the grid_mapping in order to produce this information?

Best wishes

Jonathan



More information about the CF-metadata mailing list