[CF-metadata] geoid, sea surface, height, and standard names

Jonathan Gregory j.m.gregory at reading.ac.uk
Fri Mar 24 11:06:39 MDT 2017

Dear Karl

> >There is a surface z=0, with respect to which height and depth are measured
> >in ocean models. The surface z=0 is usually the geoid, so it's not the
> >same as mean sea level.  As we have discussed in other emails, in models which
> >conserve volume rather than mass, the geoid can't change, but in the real
> >world and in models which conserve mass, it can.  However, there's a choice to
> >be made: you can stick with the original geoid (for the original volume of
> >the ocean) or make it time-dependent. Either way, I think the statement is
> >correct, but we could omit it if you think it's unhelpful in the definition.
> O.K.  I understand what is meant now, but it leads to the rather
> confusing implication that the model's surface can be located at a
> *depth" that is not zero.  When I think of an ocean depth
> measurement, I usually think of it as measure of the distance to the
> the *surface*.   If this isn't the usual understanding, then it's
> o.k. with me to leave in the sentence about the ocean GCM.
> Otherwise, I think it might confuse folks, and we should delete it.

To be honest, I'm not sure what's done in all models. Steve could comment but
he's at sea. This remark has been in there for a long time; when it was first
written, lots of ocean climate models had rigid lids, and none had a free
surface, although even in terrain-following coordinates (e.g. see ocean sigma
coordinate in App D) there is a fixed datum (such as the geoid). The remark
was intended to make things clearer, but I'm happy for it to be removed if it
now causes confusion.

Best wishes


More information about the CF-metadata mailing list