[CF-metadata] Timeseries of station data

From: Jonathan Gregory <j.m.gregory>
Date: Mon, 28 Jul 2003 16:56:55 +0100

Dear Juerg

> Thanks for the clarification. In other words the application always has
> to check the variable type first in order to correctly interpret the type
> of the dimensions.
Yes. This is because netCDF doesn't have string arrays.

> But this still leaves me with the question of how a general application
> can determine that <station> is a spatiotemporal dimension/coordinate,
> whereas the <parcel>-dimension of the example in section 6.1 is not a
> spatiotemporal dimension, but a kind of "iteration" dimension.
I think they're the same kind of dimension, whether you call them spatiotemporal
or indices. They both "address" a set of spatiotemporal locations. It would not
be appropriate to label the coordinate variable with an axis attribute as it
does not apply to a continuous x y z or t axis.

> Wouldn't a scalar coordinate variable as shown below be
> helpful?
A scalar coordinate variable (what we have been discussing in a separate thread)
is for size-one dimensions.

Cheers

Jonathan
Received on Mon Jul 28 2003 - 09:56:55 BST