[auscope-geosciml] FW: [OM2.0.swg] Statistical terms (mean, min, max) in SWE Common [SEC=UNCLASSIFIED]

Oliver.Raymond at ga.gov.au Oliver.Raymond at ga.gov.au
Mon Aug 9 08:39:30 EDT 2010


More from Alex Robin...

====================================================
True, for example in O&M we could have:

<om:Observation>
  <om:observedProperty xlink:href="http://sweet.jpl.nasa.gov/2.0/chemConcentration.owl#MassConcentration">
    <om:extension>
      <swe:statisticalOp>http://dictionary.uncertml.org/statistics/maximum</swe:statisticalOp>
    </om:extension>
  </om:observedProperty>
</om:Observation>

Provided O&M also implements the same extension placeholder as in other SWE specs.

-------------------------------------------------
Alexandre Robin
Spot Image, Web and E-Business
Tel: +33 (0)5 62 19 43 62
Fax: +33 (0)5 62 19 43 43
http://www.spotimage.com
Before printing, think about the environment



> -----Message d'origine-----
> De : Ingo Simonis [mailto:ingo.simonis at igsi.eu] Envoyé : lundi 9 août
> 2010 13:04 À : Robin, Alexandre Cc : Oliver.Raymond at ga.gov.au;
> johannes.echterhoff at igsi.eu; h.lekkerkerk at idsw.nl;
> Peter.Taylor at csiro.au; andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> dominic.lowe at stfc.ac.uk; auscope- geosciml at lists.arcs.org.au;
> om2.0.swg at lists.opengeospatial.org
> Objet : Re: [OM2.0.swg] Statistical terms (mean, min, max) in SWE
> Common [SEC=UNCLASSIFIED]
>
> I like the idea of using extensions. Even though we define it in
> SweCommon, other SWE standards could use the same approach.
>
> Cheers,
> Ingo
>
> On 09.08.2010, at 11:33, Robin, Alexandre wrote:
>
> > Hi all,
> >
> > We could investigate handling statistical properties in SWE Common
> via additional tags (as an extension).
> >
> > For example:
> >
> > <swe:field name="SiO2 concentration">  <swe:Quantity
> definition="http://sweet.jpl.nasa.gov/2.0/chemConcentration.owl#MassCo
> n
> centration">
> >    <swe:extension>
> >
> <swe:statisticalOp>http://dictionary.uncertml.org/statistics/maximum</
> s
> we:statisticalOp>
> >    </swe:extension>
> >    <swe:uom code="mg/l"/>
> >    <swe:value>55.0</swe:value>
> >  </swe:Quantity>
> > </swe:field>
> >
> > The obvious advantage of decoupling the statistical operator from
> > the
> property semantics themselves is that we don't have to define a
> separate dictionary entry for each possible combination of property
> and statistical operator.
> >
> > However the inconvenient of defining something in SWE Common is that
> other SWE standards may not benefit from it...
> >
> > Cheers,
> >
> > -------------------------------------------------
> > Alexandre Robin
> > Spot Image, Web and E-Business
> > Tel: +33 (0)5 62 19 43 62
> > Fax: +33 (0)5 62 19 43 43
> > http://www.spotimage.com
> > Before printing, think about the environment
> >
> >
> >
> >> -----Message d'origine-----
> >> De : Oliver.Raymond at ga.gov.au [mailto:Oliver.Raymond at ga.gov.au]
> >> Envoyé : lundi 9 août 2010 02:49 À : johannes.echterhoff at igsi.eu;
> >> h.lekkerkerk at idsw.nl; Peter.Taylor at csiro.au;
> >> andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> >> dominic.lowe at stfc.ac.uk Cc : om2.0.swg at lists.opengeospatial.org;
> >> Robin, Alexandre; auscope- geosciml at lists.arcs.org.au Objet :
> >> Statistical terms (mean, min, max) in SWE Common [SEC=UNCLASSIFIED]
> >>
> >> Hi all,
> >>
> >> I've been observing the recent discussion about statistical result
> >> values (eg, max, min, mean). The GeoSciML modelling team submitted
> >> a SWE change request back in January for inclusion of an explicit
> >> attribute to hold statistical descriptors for use with swe:Quantity.
> I
> >> had some discussions with Alex about options that we could use
> without
> >> changing the SWE Common model, but all involved delivering the
> >> statistical qualifier using an attribute that was not explicitly
> >> modelled to deliver statistical terms.
> >>
> >> For example:
> >>
> >> <om:Observation>
> >>  <om:observedProperty
> >>
> xlink:href="http://sweet.jpl.nasa.gov/2.0/chemConcentration.owl#Compos
> i
> >> tion"/>
> >>    <om:result>
> >>      <swe:DataRecord>
> >>        <swe:field name="SiO2 concentration">
> >>           <swe:Quantity
> >> definition="http://dictionary.uncertml.org/statistics/mean">
> >>              <swe:uom xlink:href="#weightPercent"/>
> >>              <swe:value>53.6</swe:value>
> >>           </swe:Quantity>
> >>        </swe:field>
> >>        <swe:field name="SiO2 concentration">
> >>           <swe:Quantity
> >> definition="http://dictionary.uncertml.org/statistics/maximum">
> >>              <swe:uom xlink:href="#weightPercent"/>
> >>              <swe:value>55.0</swe:value>
> >>           </swe:Quantity>
> >>        </swe:field>
> >>     </swe:DataRecord>
> >>  </om:result>
> >> </om:Observation>
> >>
> >> or....
> >>
> >> <om:Observation>
> >>  <om:observedProperty
> >>
> xlink:href="http://sweet.jpl.nasa.gov/2.0/chemConcentration.owl#Compos
> i
> >> tion"/>
> >>    <om:result>
> >>      <swe:DataRecord>
> >>        <swe:field name="mean SiO2 concentration">
> >>           <swe:Quantity>
> >>              <swe:uom xlink:href="#weightPercent"/>
> >>              <swe:value>53.6</swe:value>
> >>           </swe:Quantity>
> >>        </swe:field>
> >>        <swe:field name="maximum SiO2 concentration">
> >>           <swe:Quantity>
> >>              <swe:uom xlink:href="#weightPercent"/>
> >>              <swe:value>55.0</swe:value>
> >>           </swe:Quantity>
> >>        </swe:field>
> >>     </swe:DataRecord>
> >>  </om:result>
> >> </om:Observation>
> >>
> >> There are other ways you could deliver this data, depending on
> whether
> >> you regard the observedProperty/result as :
> >>  a) chemical composition, (or temperature)
> >>  b) SiO2 concentration, (or sea surface temperature)
> >>  c) maximum SiO2 concentration (or maximum sea surface temperature)
> >>
> >> How would others deliver this kind of data?
> >>
> >> Regards,
> >>
> >> -------------------------------------------------------------------
> >> -
> ---
> >> -------------
> >> Ollie Raymond
> >>
> >> National Advice, Maps and Data Standards Project Geoscience
> >> Australia
> >>
> >> GeoSciML Design Group
> >> IUGS Commission for the Management and Application of Geoscience
> >> Information
> >>
> >> -------------------------------------------------------------------
> >> -
> ---
> >> -------------
> >>
> >> Address: GPO Box 378, Canberra, ACT, 2601, Australia | ABN: 80 091
> 799
> >> 039
> >> Ph: +61 2 62499575 | Fax: +61 2 62499992 | Email:
> >> Oliver.Raymond at ga.gov.au
> >> National geological maps
> >>
> http://www.ga.gov.au/minerals/research/national/nat_maps/nat_geol_maps.
> >> jsp
> >> Geoscience Australia web services
> >> http://www.ga.gov.au/resources/applications/ogc-wms.jsp
> >> -------------------------------------------------------------------
> >> -
> ---
> >> -------------
> >>
> >> --- This message was created with 100% recycled electrons ---
> >>
> >> -----Original Message-----
> >> From: om2.0.swg-
> >> bounces+oliver.raymond=ga.gov.au at lists.opengeospatial.org
> >> [mailto:om2.0.swg-
> >> bounces+oliver.raymond=ga.gov.au at lists.opengeospatial.org] On
> >> bounces+Behalf
> Of
> >> Johannes Echterhoff
> >> Sent: Friday, 6 August 2010 6:41 PM
> >> To: 'Lekkerkerk, Huibert-Jan (WD)'; Peter.Taylor at csiro.au;
> >> andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> >> dominic.lowe at stfc.ac.uk
> >> Cc: om2.0.swg at lists.opengeospatial.org;
> >> Alexandre.Robin at spotimage.fr
> >> Subject: Re: [OM2.0.swg] ***MOGELIJK SPAM*** Re: New governance for
> the
> >> (former) SWE Phenomenon?
> >>
> >> Sure, the level of detail you would like to describe the O&M
> properties
> >> is
> >> up to you.
> >>
> >> To clarify: SensorML also supports non-physical processes, i.e.
> >> pure algorithms - the ProcessModel and ProcessChain would be the
> according
> >> types
> >> to use. In that respect SensorML is not heavily focused on
> describing
> >> physical processes - they are just described in more detail as that
> is
> >> necessary to explain things like system location and relative
> component
> >> position.
> >>
> >> Coming back to the semantics of the observed property: imho this
> should
> >> be
> >> identical to what is expected from the property of the
> >> observation's feature of interest. That is exactly what the O&M
> >> model defines. If you have
> a
> >> well-defined model for that (e.g. a GML application schema) then
> >> the semantics of the observedProperty are given by that model,
> >> afaic. If you have an open-ended model (e.g. using sampling
> >> features) then it is
> up
> >> to you
> >> to define the semantics of the observedProperty ... this can but
> does
> >> not
> >> need to include indication that the result is a day-min, day-max,
> day-
> >> mean
> >> etc. computation. I leave that for the semantics gurus to debate.
> >>
> >> Best,
> >> Johannes
> >>
> >>
> >>> -----Ursprüngliche Nachricht-----
> >>> Von: Lekkerkerk, Huibert-Jan (WD) [mailto:h.lekkerkerk at idsw.nl]
> >>> Gesendet: Freitag, 6. August 2010 10:20
> >>> An: johannes.echterhoff at igsi.eu; Peter.Taylor at csiro.au;
> >>> andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> dominic.lowe at stfc.ac.uk
> >>> Cc: om2.0.swg at lists.opengeospatial.org;
> Alexandre.Robin at spotimage.fr
> >>> Betreff: RE: [OM2.0.swg] ***MOGELIJK SPAM*** Re: New governance
> >>> for
> >> the
> >>> (former) SWE Phenomenon?
> >>>
> >>> Johannes,
> >>>
> >>> It probably could, the question is whether you want to go into
> >>> that
> >> much
> >>> detail for just the min/max thermometer example of Andrew. We have
> >> decided
> >>> that this would lead to an extensive amount of overhead in the
> >> datastructure
> >>> that does not solve any practical problem that isn't solved by the
> >> 'simple'
> >>> solution. The other thing is that the focus of SensorML is so much
> >> geared
> >>> towards 'field' observations where we deal for a large part with
> >> 'derived'
> >>> observations or even conclusions such as indicators for e.g. the
> >> state of
> >> the
> >>> waters that SensorML does not really support.
> >>>
> >>> Met vriendelijke groeten,
> >>>
> >>> Huibert-Jan Lekkerkerk
> >>> Sr. Projectleider standaarden IDsW
> >>> ____________________________________
> >>> InformatieDesk standaarden Water
> >>> Postbus 17, 8200 AA Lelystad
> >>> Zuiderwagenplein 2, 8224 AD Lelystad T + 31 (0) 320 298 595 M + 31
> >>> (0) 613 71 8239 F + 31 (0) 320 298366 E h.lekkerkerk at idsw.nl I
> >>> www.idsw.nl ____________________________________
> >>>
> >>> IDsW is een samenwerkingsverband van vijf waterbeherende overheden
> >> (Unie
> >> van
> >>> Waterschappen, Rijkswaterstaat, InterProvinciaal Overleg,
> Planbureau
> >> voor
> >> de
> >>> Leefomgeving en LNV). IDsW beheert en ontwikkelt
> >> informatiestandaarden
> >> voor
> >>> het Nederlandse Waterbeheer. Op deze manier zet IDsW zich in voor
> de
> >>> stroomlijning van de informatievoorziening van de sector water.
> >>>
> >>>
> >>> -----Oorspronkelijk bericht-----
> >>> Van: Johannes Echterhoff [mailto:johannes.echterhoff at igsi.eu]
> >>> Verzonden: vrijdag 6 augustus 2010 10:16
> >>> Aan: Lekkerkerk, Huibert-Jan (WD); Peter.Taylor at csiro.au;
> >>> andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> dominic.lowe at stfc.ac.uk
> >>> CC: om2.0.swg at lists.opengeospatial.org;
> Alexandre.Robin at spotimage.fr
> >>> Onderwerp: AW: [OM2.0.swg] ***MOGELIJK SPAM*** Re: New governance
> for
> >> the
> >>> (former) SWE Phenomenon?
> >>>
> >>> "a single procedure can lead to more results that may differ"
> >>>
> >>> Such a procedure can be described with a SensorML System or
> >> ProcessChain,
> >> can
> >>> it not?
> >>>
> >>> Regards,
> >>> Johannes
> >>>
> >>>
> >>>
> >>>> -----Ursprüngliche Nachricht-----
> >>>> Von:
> >>> om2.0.swg-
> >> bounces+johannes.echterhoff=igsi.eu at lists.opengeospatial.org
> >>>> [mailto:om2.0.swg-
> >>>> bounces+johannes.echterhoff=igsi.eu at lists.opengeospatial.org] Im
> >>>> bounces+Auftrag
> >>> von
> >>>> Lekkerkerk, Huibert-Jan (WD)
> >>>> Gesendet: Freitag, 6. August 2010 09:33
> >>>> An: Peter.Taylor at csiro.au; andrew.woolf at stfc.ac.uk;
> >>>> Simon.Cox at csiro.au; dominic.lowe at stfc.ac.uk
> >>>> Cc: om2.0.swg at lists.opengeospatial.org;
> >> Alexandre.Robin at spotimage.fr
> >>>> Betreff: Re: [OM2.0.swg] ***MOGELIJK SPAM*** Re: New governance
> for
> >>>> the
> >>>> (former) SWE Phenomenon?
> >>>>
> >>>> All,
> >>>>
> >>>> Just adding my two pence worth of thoughts. I'm still recent to
> >> these
> >>>> discussions, so maybe some of my remarks have already been
> >> discussed,
> >>>> but anyway.
> >>>>
> >>>> In the implementation of O&M in the Netherlands we have run into
> >>>> similar issues. We have solved this (maybe partially) by adding a
> >>>> 'valueDeterminationMethod' attribute in the proces description
> >> where
> >>>> the statistical parameters are included in a codelist that is
> >>>> references. That could / would solve the min / max / mean issues.
> >> Two
> >>>> values out of a
> >>> single
> >>>> instrument are recognized in our extension as a 'value' series
> with
> >>> different
> >>>> processes attached (I do know that our definition of process is
> not
> >>> identical
> >>>> to the latest of O&M; ours is linked directly to the result and
> >>>> is
> >> the
> >>>> process that described the result (and not the observation!).
> >>>>
> >>>> We also recognize in this specific process that there are other,
> >>>> important aspects of result determination such as interpretation
> >> conditions
> >>> (e.g.
> >>>> observation is concentration of PO4 with the result expressed in
> >> units
> >>>> of ug/l as P; this is quite common in nutrient determination).
> >>>>
> >>>> I think the time problem is already solved in O&M (monthly etc)
> >>>> although
> >>> it
> >>>> does need an exact reference to the time period used (which is
> >>>> not
> >> a
> >>>> bad idea; if it is given as a single time for e.g. the middle /
> >> start
> >>>> / end of
> >>> a
> >>>> period in the observation then I do believe this is part of the
> >> procedure!
> >>>> But putting the statistics in the procedure is not correct as a
> >> single
> >>>> procedure can lead to more results that may differ.
> >>>>
> >>>> The use of a central dictionary for this is always better than
> >>>> localised versions.
> >>>>
> >>>> Kind regards,
> >>>>
> >>>> HJ
> >>>>
> >>>> Huibert-Jan Lekkerkerk
> >>>> Sr. Projectmanager standardisation IDsW
> >>>> ____________________________________
> >>>> InformationDesk standards Water
> >>>> Postbus 17, 8200 AA Lelystad
> >>>> Zuiderwagenplein 2, 8224 AD Lelystad T + 31 (0) 320 298 595 M +
> >>>> 31 (0) 613 71 8239 F + 31 (0) 320 298366 E h.lekkerkerk at idsw.nl I
> >>>> www.idsw.nl ____________________________________
> >>>>
> >>>> IDsW is a partnership between five water managing authorities in
> >> the
> >>>> Netherlands (Union of Waterboards, Ministry of Transportation and
> >>>> Watermanagement, InterProvincial Board, Environmental Planning
> >> Agency
> >>>> and
> >>> the
> >>>> Ministry of Agriculture and Food). IDsW manages and develops
> >>>> information standards for water management in the Netherlands.
> This
> >>>> way IDsW helps developing a streamlined informationsystem for the
> >> water
> >>> sector.
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> -----Oorspronkelijk bericht-----
> >>>> Van: om2.0.swg-
> >> bounces+h.lekkerkerk=idsw.nl at lists.opengeospatial.org
> >>>> [mailto:om2.0.swg-
> >> bounces+h.lekkerkerk=idsw.nl at lists.opengeospatial.or
> >>>> g]
> >>>> Namens Peter.Taylor at csiro.au
> >>>> Verzonden: donderdag 5 augustus 2010 13:20
> >>>> Aan: andrew.woolf at stfc.ac.uk; Simon.Cox at csiro.au;
> >>>> dominic.lowe at stfc.ac.uk
> >>>> CC: om2.0.swg at lists.opengeospatial.org;
> >> Alexandre.Robin at spotimage.fr
> >>>> Onderwerp: ***MOGELIJK SPAM*** Re: [OM2.0.swg] New governance for
> >> the
> >>>> (former) SWE Phenomenon?
> >>>>
> >>>> Yes we have had similar discussion with WaterML2: balancing
> between
> >>>> what
> >>> is
> >>>> 'right' (conceptually) and what people expect (based on how they
> >>>> normally structure and interpret data). We've described the
> >>>> nature
> >> of
> >>>> such results (time series, average/min/max over an interval and
> >> which
> >>>> direction the interval holds) with the result; mainly because it
> is
> >> so
> >>>> important when interpreting the data and people don't normally
> look
> >> in
> >>>> the process description for it. A lot of existing standards for
> >>>> hydrological data
> >>> don't
> >>>> even maintain references to the generating process (not
> >>>> suggesting
> >> we
> >>>> shouldn't though, I'm all for it).
> >>>>
> >>>> Often a system will maintain a tree of the dependent results: raw
> >> as
> >>>> measured, av/min/max hourly, daily, monthly etc, i.e. a full
> >>>> suite
> >> of
> >>>> pre- calculated summary statistics. In waterml2 we are coming up
> >>>> against how
> >>> best
> >>>> to represent these in O&M, so a consistent approach would be
> >>>> worth developing.
> >>>>
> >>>> Pete.
> >>>> ________________________________________
> >>>> From: andrew.woolf at stfc.ac.uk [andrew.woolf at stfc.ac.uk]
> >>>> Sent: Thursday, August 05, 2010 7:02 PM
> >>>> To: Cox, Simon (CESRE, Kensington); dominic.lowe at stfc.ac.uk;
> >> Taylor,
> >>>> Peter (ICT Centre, Hobart)
> >>>> Cc: om2.0.swg at lists.opengeospatial.org;
> >> Alexandre.Robin at spotimage.fr
> >>>> Subject: RE: [OM2.0.swg] New governance for the (former) SWE
> >> Phenomenon?
> >>>>
> >>>> Simon,
> >>>>
> >>>> I agree with this - to a point (especially for the SST case).
> >>>>
> >>>> However, we have common enough cases where it would be forcing
> >> matters
> >>>> for the sake of the model to insist on such decomposition. For
> >>>> instance, an instrument like a maximum-minimum thermometer really
> >>>> measures maximum
> >>>> (minimum) temperature over some time interval, it doesn't measure
> >>> temperature
> >>>> and calculate maximum in a separate (sub)process. Likewise we
> >>>> have instruments that directly measure a bulk property (usually
> >>>> an
> >> average)
> >>> over
> >>>> some region of space. To require the statistical summarisation
> mode
> >> to
> >>>> be modelled as part of the procedure feels somewhat pedantic,
> >>>> certainly unnatural, and possibly misleading (in the context of
> >> 'procedure'
> >>>> practitioners would expect details of the measurement
> >> instrumentation,
> >>>> for which it would not be normal practice in such cases to regard
> >>>> 'statistical summarisation' characteristics as
> distinct/independent
> >>> parameters).
> >>>>
> >>>> As well, it is conventional practice to consider statistically
> >>> representative
> >>>> phenomena independent from the procedure (i.e. to talk about the
> >>>> January minimum temperatures over a decade without reference to
> the
> >>> procedure).
> >>>>
> >>>> Some years ago we started (but never completed) work on applying
> >> SWE
> >>>> Phenomenon to model such 'statistically representative' phenomena
> >> (see
> >>>> http://portal.opengeospatial.org/files/?artifact_id=25716 and
> >> attached
> >>> doc).
> >>>> I still think it would be useful to recognise this as a broadly
> >> useful
> >>>> specific class of constrained phenomena (and I like the phrase
> >>>> 'statistically-representative phenomena') and develop a model for
> >> it;
> >>>> and
> >>> I
> >>>> think this is the right place.
> >>>>
> >>>> Andrew
> >>>>
> >>>> -----Original Message-----
> >>>> From:
> >>>> om2.0.swg-bounces+andrew.woolf=stfc.ac.uk at lists.opengeospatial.or
> >>>> g
> >>>>
> >>>
> >> [mailto:om2.0.swg-
> >> bounces+andrew.woolf=stfc.ac.uk at lists.opengeospatial.org]
> >>>> On Behalf Of Simon.Cox at csiro.au
> >>>> Sent: 04 August 2010 15:36
> >>>> To: Lowe, Dominic (STFC,RAL,SSTD); Peter.Taylor at csiro.au
> >>>> Cc: om2.0.swg at lists.opengeospatial.org;
> >> Alexandre.Robin at spotimage.fr
> >>>> Subject: Re: [OM2.0.swg] New governance for the (former) SWE
> >> Phenomenon?
> >>>>
> >>>> You have to be a bit careful with these constrained phenomena.
> >>>> Using O&M terminology, in most cases they actually mix the
> >>> observedProperty
> >>>> with either the procedure (e.g. monthly-mean temperature) or with
> >> the
> >>> feature
> >>>> of interest (sea surface temperature).
> >>>> 'Temperature' is the true semantics.
> >>>> Monthly mean is procedure/protocol/sampling-regime.
> >>>> Sea surface is foi.
> >>>>
> >>>> So we should ask the question as to whether we should be burying
> >>>> procedure/foi in the phenomenon, or rather encouraging people to
> >> split
> >>> these
> >>>> factors out correctly.
> >>>>
> >>>> ---
> >>>> simon.cox at csiro.au
> >>>> CSIRO Earth Science and Resource Engineering Australian Resources
> >>>> Research Centre PO Box 1130, Bentley WA 6102 Australia
> >>>> T: +61 8 6436 8639
> >>>> F: +61 8 6436 8555
> >>>> ________________________________________
> >>>> From: Dominic Lowe [dominic.lowe at stfc.ac.uk]
> >>>> Sent: Wednesday, 4 August 2010 9:31 PM
> >>>> To: Taylor, Peter (ICT Centre, Hobart)
> >>>> Cc: ingo.simonis at igsi.eu; om2.0.swg at lists.opengeospatial.org;
> >>>> Alexandre.Robin at spotimage.fr; Cox, Simon (CESRE, Kensington)
> >>>> Subject: Re: [OM2.0.swg] New governance for the (former) SWE
> >> Phenomenon?
> >>>>
> >>>> Hi Pete,
> >>>>
> >>>> Great - it would be good to have you involved.
> >>>>
> >>>> In the first instance I'm mainly interested in keeping the
> >>>> current model alive :)
> >>>>
> >>>> Beyond that, one development that's of particular interest to me
> >> would
> >>>> be
> >>> the
> >>>> idea that we don't just want to say 'temperature' but might want
> to
> >>>> say
> >>> the
> >>>> phenomenon is 'monthly mean temperature', or 'maximum
> temperature'.
> >> So
> >>>> you have the concept of the base phenomenon (which may be
> described
> >>> externally)
> >>>> and then you have some sort of function. In climate data we use a
> >> lot
> >>>> of combined averaging mechanisms that I'm not sure can be
> expressed
> >> in
> >>>> the current model.
> >>>>
> >>>> Also as Alex suggested, alternate encodings may be useful
> >> (especially
> >>>> if gml:Dictionary becomes deprecated in future versions).
> >>>>
> >>>> Cheers
> >>>> Dom
> >>>>
> >>>> On 04/08/10 11:52, Peter.Taylor at csiro.au wrote:
> >>>>> I'm interested in this as well, I'd be happy to provide some
> >> input
> >>>>> if I
> >>>> can. Dom: do you have an idea of what the scope or features you
> are
> >>>> interested in would be? I seem to recall the current model
> supports
> >>>> constraints and compounds, with a reference id to an external
> >> definition.
> >>> I
> >>>> understood Alex's reasoning behind removing it (in light of using
> >>>> plain references to semantic resources), but it is handy
> >>>> sometimes
> >> to
> >>>> have extra metadata that is closer to the reference; resolution
> >>>> of
> >> the
> >>>> semantic reference is often not supported or available.
> >>>>>
> >>>>> Cheers,
> >>>>> Pete.
> >>>>> ________________________________________
> >>>>> From:
> >>>>> om2.0.swg-bounces+peter.taylor=csiro.au at lists.opengeospatial.org
> >>>>> [om2.0.swg-
> >> bounces+peter.taylor=csiro.au at lists.opengeospatial.org]
> >>>>> On Behalf Of Ingo Simonis [ingo.simonis at igsi.eu]
> >>>>> Sent: Wednesday, August 04, 2010 7:35 PM
> >>>>> To: Dominic Lowe
> >>>>> Cc: om2.0.swg at lists.opengeospatial.org; Robin,  Alexandre; Cox,
> >>>>> Simon (CESRE, Kensington)
> >>>>> Subject: Re: [OM2.0.swg] New governance for the (former) SWE
> >> Phenomenon?
> >>>>>
> >>>>> They main reason I want to avoid having another SWE SWG is that
> >> we
> >>> already
> >>>> have major difficulties to find time slots at the TC meetings...
> As
> >>>> the semantics are a SWE-wide issue, it would fit into SweCommon,
> >>>> I
> >> would
> >>> say.
> >>>>>
> >>>>> Best regards,
> >>>>> Ingo
> >>>>>
> >>>>>
> >>>>> On 04.08.2010, at 11:19, Dominic Lowe wrote:
> >>>>>
> >>>>>> Alex, Simon, Ingo, Dan,
> >>>>>>
> >>>>>> Many thanks for your input on this.
> >>>>>>
> >>>>>> Logically I also like the idea of a SWE Semantics SWG, but am
> >> also
> >>>>>> slightly wary of the overhead involved of a new SWG (both at
> >>>>>> startup and ongoing).
> >>>>>>
> >>>>>> As Alex suggests the more pragmatic solution may be to extend
> >> the
> >>>>>> SWG Common charter and start the work sooner. We could always
> >> set
> >>>>>> up SWE Semantics at a later date if needs be. (On the other
> >>>>>> hand
> >> if
> >>>>>> we know we are going to need SWE Semantics SWG at some point
> >> then
> >>>>>> there is a case for doing it from the offset...)
> >>>>>>
> >>>>>> Basically I would be happy with either solution - so unless
> >> there
> >>>>>> are any other suitable alternatives I guess it is up to the SWE
> >>>>>> Common SWG at this point?
> >>>>>>
> >>>>>> I think we agree that the Phenomenon model should have it's own
> >>>>>> namespace regardless of which SWG develops it.
> >>>>>>
> >>>>>>> In the mean time, why can't you use version 1.0 of the
> >> phenomenon
> >>> model?
> >>>>>>>
> >>>>>> Yes thinking about it this does seem sensible - instead of
> >>>>>> reproducing it elsewhere :)
> >>>>>>
> >>>>>> Regards
> >>>>>>
> >>>>>> Dominic
> >>>>>>
> >>>>>>
> >>>>>> On 04/08/10 08:55, Robin, Alexandre wrote:
> >>>>>>> Hi Dominic,
> >>>>>>>
> >>>>>>> I agree that a SWE semantics SWG would be ideal because that
> >> would
> >>> give
> >>>> us a chance to seriously study the interactions between the SWE
> >>>> phenomenon model and other ontology languages such as RDF and OWL.
> >> For
> >>>> instance it is worth seriously investigating how the conceptual
> >> model
> >>>> could be
> >>> implemented
> >>>> in OWL rather than plain XML, etc...
> >>>>>>>
> >>>>>>> Since work on the SWE Common Data Model and SWE Service Common
> >> is
> >>> coming
> >>>> to an end in the SWE Common SWG, there is also a possibility that
> >>>> version
> >>> 2.0
> >>>> of the phenomenon model can be developed there, provided we
> >>>> update
> >> the
> >>>> charter. I will probably not be available to chair the group
> >> anymore
> >>>> but
> >>> I'm
> >>>> sure we can find somebody else for that. I'm proposing this
> because
> >> it
> >>>> can
> >>> be
> >>>> a quicker thing than creating a whole new SWG and work can start
> >>> immediately.
> >>>>>>>
> >>>>>>> In the mean time, why can't you use version 1.0 of the
> >> phenomenon
> >>> model?
> >>>>>>>
> >>>>>>> Cheers,
> >>>>>>>
> >>>>>>> -------------------------------------------------
> >>>>>>> Alexandre Robin
> >>>>>>> Spot Image, Web and E-Business
> >>>>>>> Tel: +33 (0)5 62 19 43 62
> >>>>>>> Fax: +33 (0)5 62 19 43 43
> >>>>>>> http://www.spotimage.com
> >>>>>>> Before printing, think about the environment
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>> -----Message d'origine-----
> >>>>>>>> De : om2.0.swg-
> >>>>>>>> bounces+alexandre.robin=spotimage.fr at lists.opengeospatial.org
> >>>>>>>> [mailto:om2.0.swg-
> >>>>>>>> bounces+alexandre.robin=spotimage.fr at lists.opengeospatial.org
> >>>>>>>> bounces+]
> >> De
> >>>>>>>> bounces+la
> >>>>>>>> part de Simon.Cox at csiro.au
> >>>>>>>> Envoyé : lundi 2 août 2010 20:28 À : dominic.lowe at stfc.ac.uk;
> >> om2.0.swg at lists.opengeospatial.org
> >>>>>>>> Objet : Re: [OM2.0.swg] New governance for the (former) SWE
> >>> Phenomenon?
> >>>>>>>>
> >>>>>>>> Dom -
> >>>>>>>>
> >>>>>>>> I understand your reasoning: the O&M activity is a plausible
> >>> umbrella.
> >>>>>>>> As you are aware, it originated in O&M pre v1.0 and was only
> >>>>>>>> moved into SWE Common relatively late in the piece, when it
> >>>>>>>> appeared that all the 'common' elements should be there.
> >>>>>>>>
> >>>>>>>> Now SWE Common has been re-cast as a data-typing exercise the
> >>>>>>>> 'phenomenon' model deals with some semantics which appears to
> >> be
> >>>>>>>> beyond the scope of SWE Common.
> >>>>>>>>
> >>>>>>>> But it needs its own namespace, as it is equally important
> >>>>>>>> for SensorML.
> >>>>>>>> To progress this, I'm not sure that the O&M SWG is formally
> >> the
> >>>>>>>> right place as the charter would need a bit of 'interpreting'
> >> to
> >>>>>>>> take this on.
> >>>>>>>> Maybe a 'SWE Semantics' SWG?
> >>>>>>>>
> >>>>>>>> Simon
> >>>>>>>>
> >>>>>>>> -----Original Message-----
> >>>>>>>> From:
> >>>>>>>> om2.0.swg-bounces+simon.cox=csiro.au at lists.opengeospatial.org
> >>>>>>>> [mailto:om2.0.swg-
> >> bounces+simon.cox=csiro.au at lists.opengeospatial
> >>>>>>>> .o
> >>>>>>>> rg]
> >>>>>>>> On Behalf Of Dominic Lowe
> >>>>>>>> Sent: 29 July 2010 11:33
> >>>>>>>> To: om2.0.swg at lists.opengeospatial.org
> >>>>>>>> Subject: [OM2.0.swg] New governance for the (former) SWE
> >> Phenomenon?
> >>>>>>>>
> >>>>>>>> All,
> >>>>>>>>
> >>>>>>>> I would like to propose that a new standard or document is
> >>>>>>>> created that provides a formal OGC definition of the model
> >>>>>>>> and schemas formerly known as the SWE Phenomenon model.
> >>>>>>>>
> >>>>>>>> As you know, this was removed from SWE recently as (according
> >> to
> >>>>>>>> my
> >>>>>>>> understanding) it was felt SWE was not the right place to
> >> govern
> >>>>>>>> this very general model. However the need for such a model
> >>>>>>>> remains; we have been using the SWE Phenomenon for a while
> >>>>>>>> now
> >> in
> >>>>>>>> the context of CSML (Climate Science Modelling Language).
> >> There
> >>>>>>>> has also recently been a need to use it in the ESA HMA Follow
> >> On
> >>>>>>>> project - and as a result of it's current non-existence we
> >>>>>>>> are now having to replicate the old SWE model in the HMA
> >> namespace.
> >>>>>>>>
> >>>>>>>> Clearly it is a useful and practical model and it would be
> >> good
> >>>>>>>> to have an agreed definition of phenomena in OGC that could
> >>>>>>>> be used in a variety of contexts.
> >>>>>>>>
> >>>>>>>> I would be happy to take on the responsibility for taking
> >>>>>>>> this forward whether it be as a best practice paper, or a
> >> standalone
> >>>>>>>> OGC standard - starting with the old SWE model as a baseline.
> >>>>>>>> However it needs to have a 'home' within OGC. I have
> >>>>>>>> discussed this with Johannes and we both agree that the O&M
> >>>>>>>> SWG may be a logical place for the governance of this model -
> >>>>>>>> it is very generic and is an important part of the
> >>>>>>>> observations
> >> framework.
> >>>>>>>> However I don't imagine that it would necessarily be under
> >>>>>>>> the O&M namespace at the implementation level (although that
> >>>>>>>> is clearly a discussion to have).
> >>>>>>>>
> >>>>>>>> So this email is to canvass initial support for (or against)
> >> the
> >>>>>>>> idea within the O&M SWG. What are your thoughts..?
> >>>>>>>>
> >>>>>>>> Regards,
> >>>>>>>>
> >>>>>>>> Dominic
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> _______________________________________________
> >>>>>>>> OM2.0.swg mailing list
> >>>>>>>> OM2.0.swg at lists.opengeospatial.org
> >>>>>>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>>>>>>
> >>>>>>>> _______________________________________________
> >>>>>>>> OM2.0.swg mailing list
> >>>>>>>> OM2.0.swg at lists.opengeospatial.org
> >>>>>>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>> _______________________________________________
> >>>>>> OM2.0.swg mailing list
> >>>>>> OM2.0.swg at lists.opengeospatial.org
> >>>>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> OM2.0.swg mailing list
> >>>>> OM2.0.swg at lists.opengeospatial.org
> >>>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>> _______________________________________________
> >>>> OM2.0.swg mailing list
> >>>> OM2.0.swg at lists.opengeospatial.org
> >>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>>
> >>>> --
> >>>> Scanned by iCritical.
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> OM2.0.swg mailing list
> >>>> OM2.0.swg at lists.opengeospatial.org
> >>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>>
> >>>>
> >>>> Disclaimer
> >>>>
> >>
> **********************************************************************
> >>>> ** Aan dit bericht kunnen geen rechten worden ontleend. Dit
> bericht
> >> is
> >>>> uitsluitend bestemd voor de geadresseerde. Als u dit bericht per
> >> abuis
> >>>> hebt ontvangen, wordt u verzocht het te vernietigen en de
> >>>> afzender
> >> te
> >>>> informeren. Wij adviseren u om bij twijfel over de juistheid of
> >>>> de volledigheid van de mail contact met afzender op te nemen.
> >>>>
> >>>> This message shall not constitute any rights or obligations.
> >>>> This message is intended solely for the addressee.
> >>>> If you have received this message in error, please delete it and
> >>>> notify the sender immediately. When in doubt whether this message
> >> is
> >>>> correct or complete, please contact the sender.
> >>>>
> >>
> **********************************************************************
> >>>> **
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> OM2.0.swg mailing list
> >>>> OM2.0.swg at lists.opengeospatial.org
> >>>> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >>>
> >>>
> >>>
> >>> Disclaimer
> >>>
> >>
> **********************************************************************
> *
> >> *
> >>> Aan dit bericht kunnen geen rechten worden ontleend. Dit bericht
> >>> is uitsluitend bestemd voor de geadresseerde. Als u dit bericht
> >>> per
> >> abuis
> >>> hebt ontvangen, wordt u verzocht het te vernietigen en de afzender
> te
> >>> informeren. Wij adviseren u om bij twijfel over de juistheid of de
> >>> volledigheid van de mail contact met afzender op te nemen.
> >>>
> >>> This message shall not constitute any rights or obligations.
> >>> This message is intended solely for the addressee.
> >>> If you have received this message in error, please delete it and
> >>> notify the sender immediately. When in doubt whether this message
> >>> is correct or complete, please contact the sender.
> >>>
> >>
> **********************************************************************
> *
> >> *
> >>
> >>
> >> _______________________________________________
> >> OM2.0.swg mailing list
> >> OM2.0.swg at lists.opengeospatial.org
> >> https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
> >
> >
> >
> > _______________________________________________
> > OM2.0.swg mailing list
> > OM2.0.swg at lists.opengeospatial.org
> > https://lists.opengeospatial.org/mailman/listinfo/om2.0.swg
>





More information about the GeoSciML mailing list