If I remember correctly, when we discussed 'measured units' a while back (e.g., counts), the prevailing opinion was that this was not a useful thing for CF to worry about, because there was no a priori known conversion to 'CF units'. (I welcome any correction to that understanding, it may come up again soon.)
In this case, I appreciate that there is at least a broad understanding of the conversion from logarithmic units back to the CF units. Can we go to the trouble of making this conversion deterministic and computable, i.e., have an algorithm by which the computer reading the file can automatically determine the transformation required to restore CF units? And should the units that are specified be the units that will return once the inverse log function is applied? Or should the units be ... something else? John On Sep 24, 2010, at 12:01, Jonathan Gregory wrote: > Dear Steve > >> So far my reading has led me to the conclusion that the only possibility is >> to describe the variable in the long name to be the logarithm of the >> measured data. Is that the accepted solution? > > I believe that to be the case, yes. I think that is OK because I would argue > that the log of a quantity is not the same quantity, from a geophysical or > physical point of view. A log doesn't have units, for instance (except > dimensionless ones such as decibels). If the log was taken in order to pack > the data, that ought to be recognised as such by adding a new convention to > section 8 of the CF standard. > > Best wishes > > Jonathan > _______________________________________________ > CF-metadata mailing list > [email protected] > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata -------------- I have my new work email address: [email protected] -------------- John Graybeal <mailto:[email protected]> phone: 858-534-2162 System Development Manager Ocean Observatories Initiative Cyberinfrastructure Project: http://ci.oceanobservatories.org Marine Metadata Interoperability Project: http://marinemetadata.org _______________________________________________ CF-metadata mailing list [email protected] http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
