Dear Olivier
I can't see the map (I expect the email system removed it) so I'm still
puzzled what you mean by
tendency_of_local_average_sea_level_change
or
local_tendency_of_global_average_sea_level_change
I appreciate that these definitions about sea level are subtle and complicated.
I wonder if maybe you mean the local rate of change in the thickness of the
ocean, from the sea surface to the sea floor? That is not what the altimeter
observes, which is the change in the sea surface wrt an Earth reference frame.
It is the quantity you get after correcting for GIA (on all timescales),
including the global-mean subsidence of the ocean floor. If that is what you
mean, it could be called tendency_of_ocean_thickness. "Ocean thickness" is
a surprising phrase, but it's consistent with the use of "thickness" in many
other names such as ocean_mixed_layer_thickness and sea_ice_thickness. We use
"depth" as a vertical coordinate, and "thickness" as a vertical interval.
I thought a bit more about the amplitude and phase. Earlier I suggested we
might consider describing these with cell_methods, but I'm not sure that really
suits the purpose, and your way is better. We could regard amplitude_of_X and
phase_of_X as new "transformations" in the sense of the standard name
guidelines
http://cf-pcmdi.llnl.gov/documents/cf-standard-names/guidelines,
where there is a table of them. As you suggest, they can be defined wrt fitting
a sinusoid.
However it would be better to be general about the period, or else we would
need different transformation phrases for annual, semiannual and any other
period. As you remark, this idea may find other uses in CF. A more general way
to do it would be to require the addition of a coordinate variable to specify
the frequency or period of interest. For that we would also need a new general
standard_name, for example harmonic_period or harmonic_frequency. What do
you think? That has the advantage that you could store the numbers for many
different periods in one data variable.
Best wishes
Jonathan
Received on Thu Jul 05 2012 - 04:30:58 BST
