Hi all,
I favor Derrick's view and arguments. It would be nice however to
address John's goals. Can someone think of another way that the
existence of these measurements might be "discovered" and their
potential uses might be made clear?
Karl
[email protected] wrote:
Hi John,
Sounds like a tricky problem, but probably one worth solving. My initial reaction to your question
" Is there a standard technique that would let me convert any existing standard name to a 'raw units' standard name?"
is that this could potentially lead down a difficult road as the mapping is not one to one. A satellite may measure radiance across several wavelength bands which are combined together with independent parameters in an algorithm to estimate SST. Similarly, the same volts measurement from a CTD temperature sensor must be combined with volts measurements from an oxygen sensor (and algorithm parameters) to estimate dissolved oxygen, but independently, the first volts measurement, plus different algorithm/parameters can yield water temperature.
I'm no CF guru, but it seems like the full complexity of mapping raw data
(possibly many) + parameters + algorithm = measured value within the CF
conventions might be tough. I know SensorML is designed around this idea
though, so perhaps there are ideas that can be borrowed from that framework????
I realize this isn't overly helpful but I suppose my point is that during the
conversation that might hash out a solution for this within CF, my vote would
be for creating standard names for the variable that is actually measured (e.g.
volts) rather than semantically try to link engineering unit measurements to
physical unit measurements since that relationship is not always
straightforward.
I look forward to hearing more from the group.
Best regards,
Derrick
----- Original Message -----
From: John Graybeal <[email protected]>
Date: Thursday, March 5, 2009 6:44 pm
Subject: [CF-metadata] standard names for variables in 'raw engineering'
units
Here's a question about observational variables. Many of our
sensors
measure variables that are in the standard names list, but the data
is
not yet converted to science units, for example temperature or
oxygen
concentration.
These measurements are typically represented by a unit like
'volts' (or even 'counts') -- definitely not directly convertible
to
the canonical units! Is there a standard technique that would let
me
convert any existing standard name to a 'raw units' standard name?
If not, to start the conversation perhaps I can suggest one of the
following to indicate such metrics?
suffix: _in_engineering_units
suffix: _in_raw_units
prefix: raw_measure_of_
This capabilitiy will be important for finding data that
potentially
can be converted to the desired values, but has not yet been
converted. (Some of our collected parameters are not the primary
purpose of the experiment, and they remain in raw units unless
someone
needs them converted.)
John
--------------
John Graybeal <mailto:[email protected]> -- 831-775-1956
Monterey Bay Aquarium Research Institute
Marine Metadata Interoperability Project: http:// marinemetadata.org
_______________________________________________
CF-metadata mailing list
[email protected]
http:// mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
_______________________________________________
CF-metadata mailing list
[email protected]
http:// mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
