[CF-metadata] positive attribute expansion of use and reserved values

Kehoe, Kenneth E. kkehoe at ou.edu
Thu Sep 28 09:46:37 MDT 2017


Jonathan and Jim,

Thanks for your replies. As Jonathan stated I was only trying to expand the current use of an existing reserved attribute so the CF standard would not need to reserve further attributes. I understand and agree that putting information in two locations is a bad idea (positive attribute and standard_name), but my problem is that we can not reasonably expect our users to have software to look up positive direction in an external standard name table. We need the basic information to be contained in the netCDF file. I don’t see us parsing the standard_name description or looking up information in an ancillary column in the standard_name table as that would require work beyond what my users are willing to do. We currently put direction in multiple locations (long_name, comment, variable name) and I was trying to find a solution to have one place and allow it to be machine readable. For now I’ll continue to encourage the direction information to be placed in a comment attribute and continue to look for a more machine readable way to parse the direction information.

Thanks,

Ken


On Sep 28, 2017, at 7:48 AM, Jonathan Gregory <j.m.gregory at reading.ac.uk<mailto:j.m.gregory at reading.ac.uk>> wrote:

Dear Jim

Yes, that's right, the positive attribute is currently supported only for
vertical coordinate variables. Following the COARDS rules, it can be used
to identify a variable as a vertical coordinate variable. If other kinds of
coordinate variable than vertical were allowed to have a positive attribute,
it's likely that existing software would be confused. However, that's not
the issue I was writing about.

If I understood Ken's email correctly, he was asking about the possibility
of using the positive attribute on data variables as well, instead of or as
well as the sign convention indicated in the standard name. As I wrote, I
think this is problematic, though I do understand the reason for making the
suggestion.

Best wishes

Jonathan

----- Forwarded message from Jim Biard <jbiard at cicsnc.org<mailto:jbiard at cicsnc.org>> -----

Date: Thu, 28 Sep 2017 08:44:52 -0400
From: Jim Biard <jbiard at cicsnc.org<mailto:jbiard at cicsnc.org>>
To: cf-metadata at cgd.ucar.edu<mailto:cf-metadata at cgd.ucar.edu>
Subject: Re: [CF-metadata] positive attribute expansion of use and reserved
values
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:52.0)
Gecko/20100101 Thunderbird/52.3.0

Jonathan,

I am confused by your answer to Kenneth. As you know, we already
have a CF attribute named positive that derives from COARDS. CF 1.7
states in Section 1.3

  Vertical coordinates with units of pressure may also be identified
  by the*|units|*attribute. Other vertical coordinates must use the
  attribute*|positive|*which determines whether the direction of
  increasing coordinate value is up or down.

I don't see Kenneth asking for anything new other than an expanded
vocabulary for the attribute, but in your answer I get the
impression that you feel he is asking for something fundamentally
different from what we currently have. Are you suggesting that we
should deprecate the use of the 'positive' attribute? Your response
doesn't seem to reflect the scope of his request.

Grace and peace,

Jim

On 9/27/17 7:45 PM, Jonathan Gregory wrote:
Dear Ken

Thanks for your email. I sympathise with the problem, but I don't agree with
the proposal. Actually similar suggestions have been made before.

It's an important principle with the standard names that they always indicate
their sign convention. This is so that, if a standard name is provided, the
sign convention is unavoidably specified; if the sign convention were not in
the standard name, but in another attribute e.g. positive, it is certain that
it would be omitted sometimes by mistake. If the sign convention were specified
in the standard name *and* another attribute, it is certain that they would
sometimes be inconsistent by mistake. Either mistake would make the data less
usable. You mention cases where the wrong standard name is given. I agree, that
makes the data less usable too, but I'd say the solution to that is to fix the
data, when the problem has been identified; we should not have to modify the
convention in a way which would make it generally more error-prone. You also
mention cases where you don't have a standard name but you do need a direction.
Presumably you must have some other information in that case about what the
quantity is - which attribute are you using? If it's the long name, you could
put the sign convention in there, for example, as for standard names. New
standard names can also be requested for instrumental quantities, and the
direction doesn't have to be "up" or "down" in standard names. As you probably
know, there are already standard names containing away_from to indicate their
sign convention, just as you suggest.

I agree that there is sometimes a need to know how to relate quantities with
opposite sign conventions in their standard names. The standard names are
mostly systematically constructed, but for use by humans; they aren't designed
to be parsed by machines. If there is a need for some means to deal with this,
I would favour recording the sign convention as a machine-readable extra piece
of information in the standard name table. If we put it in the table, it must
be consistent with the standard name; you'd just have to look it up, instead of
trying to extract it from the standard name.

Best wishes

Jonathan

----- Forwarded message from "Kehoe, Kenneth E." <kkehoe at ou.edu<mailto:kkehoe at ou.edu>> -----

Date: Wed, 27 Sep 2017 22:49:20 +0000
From: "Kehoe, Kenneth E." <kkehoe at ou.edu<mailto:kkehoe at ou.edu>>
To: "CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>" <CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>>
Subject: [CF-metadata] positive attribute expansion of use and reserved
values

CF-metadata,

I would like to propose an expansion of the use and reserved values for the “positive” attribute (section 4.3), specifically to include values in addition to the two reserved values of “up” and “down” to include “towards” and “away”.

Most variables define direction in the standard_name, but with instrumentation a standard name is often not available or the correct definition of the name exists with the wrong positive direction. Also, needing to understand or extract direction from the standard_name can be difficult for a simple tool not wanting to review the standard_name definition just to see if a transformation is needed. Expanding the use of the “positive” attribute would reduce the number of standard names by not needing to include positive direction in the definition. This would also follow the recommendation of being consistent with the definition between the standard_name and positive attribute. The “positive” attribute is currently reserved for use with coordinate dimensions only, but the same logic can be used with data variables to indicate direction. For example rate of speed for vertical velocities could be described by indicating positive = “up” when vertical velocities are positive when moving away from the surface.

Instruments are also often not installed perpendicular to the surface, and the coordinate system is better described as towards or away from the instrument. Specifically for radial instruments. Errors in misunderstanding direction with radial velocities or accelerations are comment when not specifically defined. There is no vendor standard.

I’ll leave my suggestion at towards and away, but this could also be expanded to include cardinal direction for East-West/North-South directions.

Thanks,

Ken




Kenneth E. Kehoe
 Research Associate - University of Oklahoma
 Cooperative Institute for Mesoscale Meteorological Studies
 ARM Climate Research Facility - Data Quality Office
 e-mail: kkehoe at ou.edu<mailto:kkehoe at ou.edu><mailto:kkehoe at ou.edu> | Office: 303-497-4754 | Cell: 405-826-0299

_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

----- End forwarded message -----
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

--
CICS-NC <http://www.cicsnc.org/> Visit us on
Facebook <http://www.facebook.com/cicsnc> *Jim Biard*
*Research Scholar*
Cooperative Institute for Climate and Satellites NC <http://cicsnc.org/>
North Carolina State University <http://ncsu.edu/>
NOAA National Centers for Environmental Information <http://ncdc.noaa.gov/>
/formerly NOAA’s National Climatic Data Center/
151 Patton Ave, Asheville, NC 28801
e: jbiard at cicsnc.org<mailto:jbiard at cicsnc.org> <mailto:jbiard at cicsnc.org>
o: +1 828 271 4900

/Connect with us on Facebook for climate
<https://www.facebook.com/NOAANCEIclimate> and ocean and geophysics
<https://www.facebook.com/NOAANCEIoceangeo> information, and follow
us on Twitter at @NOAANCEIclimate
<https://twitter.com/NOAANCEIclimate> and @NOAANCEIocngeo
<https://twitter.com/NOAANCEIocngeo>. /



_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata


----- End forwarded message -----
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu<mailto:CF-metadata at cgd.ucar.edu>
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata



Kenneth E. Kehoe
  Research Associate - University of Oklahoma
  Cooperative Institute for Mesoscale Meteorological Studies
  ARM Climate Research Facility - Data Quality Office
  e-mail: kkehoe at ou.edu<mailto:kkehoe at ou.edu> | Office: 303-497-4754 | Cell: 405-826-0299

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20170928/7456ac3f/attachment-0001.html>


More information about the CF-metadata mailing list