[CF-metadata] standard name proposals

From: Jonathan Gregory <j.m.gregory>
Date: Thu, 7 Aug 2003 20:05:42 +0100

> Having packed data in the range 1024-1028 and using an add_offset of -1000
> was John's suggestion. His point (I think) is that applications using this
> data expect the unpacked values to be in the range 24-28.

OK. If the software processing the data really wants to have values with 1000
subtracted, I agree that this is not a packing issue but a genuinely different
quantity from density. We'll have to decide a good name. They are usually
called sigma-t and sigma-theta I think but that is really rather cryptic for a
non-oceanographer. Jonathan
Received on Thu Aug 07 2003 - 13:05:42 BST