[CF-metadata] date and time

From: Jonathan Gregory <j.m.gregory>
Date: Thu, 30 Oct 2008 17:42:41 +0000

Dear John, Roy, Steve et al.

I don't think there is a problem with having more than one variable in a file
with the standard_name of time (or any other standard_name), but I expect that
you'd want some other metadata to discriminate among them, as Roy said. I also
see no problem with having other kinds of "time" standard name, though. We
already have forecast_reference_time. I suggested a definition of plain "time"
as being the temporal coordinate (or the best estimate of it - any measured
coordinate has some uncertainty of course) of the geophysical data. The raw
time that came out of some clock which might require correction could have some
other standard name to discriminate it from this best estimate, if that were
useful, I think.

I don't think we have a thorough answer to the issue of instrumental and raw
data, though. As with everything else, we have addressed it as and when it has
come up, and consequently we already have some names that relate to measure-
ments and how they are made, whereas the stdname table started out with only
geophysical quantities named in it - things which claim to be properties of the
world, rather than things which are part of the measurement process. Since we
are only addressing it piecemeal, we might be making a mess of it, but I
certainly don't know enough about this to comment. As Steve said wisely, you
can only standardize the things that you understand thoroughly!

Best wishes

Jonathan
Received on Thu Oct 30 2008 - 11:42:41 GMT