I may not have a CF compliant thinking process here, but I see little, if any, difference in the need of an offset reference attribute verses the need of a units attribute. The units attribute is often a multiplier between equivalent descriptions, while the reference is an additive difference between equivalent descriptions.
Maybe the existence of the needed metadata attribute is the only thing needed to be encoded into the standard name, like:
>>
On the other hand, we could also take advantage of the situation by
introducing some new standard names, more general, like
"sea_surface_height_above_reference_surface" and
"sea_surface_height_above_reference_profile", that would include all sea
surface height kinds of parameters, and all the reference datums?
>>
Which could imply the existence of the attribute "reference_surface" or
"reference_profile"
or more simply "reference".
If a person wanted to specify his sea surface height without a reference
datum I suppose that they could use a name like "sea_surface_height"
with no reference, but they might as well also leave out the units
attribute for feet or meters, for all the good such a piece of
incomplete data would be. (OK, a bit harsh, but true.)
Tom Gross
Roy Lowry wrote:
>There is an interesting general point raised here by the statement 'a standard name for each reference datum would be a
>little bit excessive'. If the specific reference datum isn't held in the standard name then the information needs to be in another standardised location in the CF attributes or the data are not fully decribed by the metadata. I remember the issue of datums came up before with regard to CF, but can't remember if it was ever resolved.
>
>Cheers, Roy.
>
>
>
>
>>>>Olivier Lauret <olauret at cls.fr> 8/22/2006 12:26 pm >>>
>>>>
>>>>
>Dear Tom,
>
>The fact is that MSL, MLLW, NAVD88.. especially concern tide gauges,
>coastal oceanography - among the applications, and do not really apply to
>satellite altimetry. Of course I only focused on altimeter data, that
>usually refer to, let's say, 4 different reference surfaces: the reference
>ellipsoid (it can be WGS84 or any other), a mean along-track profile or a
>mean sea surface, and (indirectly) the geoid.
>But I agree with you: a standard name for each reference datum would be a
>little bit excessive.
>At the same time "sea_surface_height_above_sea_level",
>"sea_surface_height_above_geoid" were already availble in the standard
>names table, and none of them could "include" or be compared to altimeter
>sea_surface_height_above_reference_ellipsoid, because beyond the reference
>datum, the meaning is different in terms of physical processes to be
>considered. Then it raises the question of how we define the reference
>surfaces.
>
>One one hand, if the whole entire sea surface height parameters (in situ,
>tide gauges..) could refer to {"sea_surface_height_above_sea_level",
>"sea_surface_height_above_geoid"}, except satellite altimetry, maybe it is
>possible for us to introduce the missing altimeter dedicated standard names?
>
>On the other hand, we could also take advantage of the situation by
>introducing some new standard names, more general, like
>"sea_surface_height_above_reference_surface" and
>"sea_surface_height_above_reference_profile", that would include all sea
>surface height kinds of parameters, and all the reference datums?
>For "sea_surface_height_above_reference_surface", the reference surface
>could be a mean sea surface, a reference ellipsoid, or the geoid, that
>could be specified as attributes.
>For "sea_surface_height_above_reference_profile", the reference profile
>could be a mean sea level, a mean profile etc, defined as attributes as well.
>
>I don't know. That depends on the fellings of the ocean/climate users
>community, but the discussion is open.
>
>Best regards
>
>
>Olivier Lauret
>
>CLS Space Oceanography Division
>Data Information and Diffusion
>8-10 rue Hermes,
>31526 Ramonville-St-Agne, Cedex, France
>
>Email: olivier.lauret at cls.fr
>Tel: (+33) (0) 561 39 48 51
>Fax:(+33) (0) 561 39 48 51
>Internet:
>http://www.cls.fr
>http://www.aviso.oceanobs.com/
>
>
>_______________________________________________
>CF-metadata mailing list
>CF-metadata at cgd.ucar.edu
>http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
>
>
>
Received on Tue Aug 22 2006 - 11:34:16 BST
