Minutes for the MetOcean DWG Teleconference on 16th January 2012, 16:00 - 17.04 UTC

Attendees:

Matthias Müller (University of Dresden)

Steve Foreman (WMO)

Stephan Siemen (ECMWF)

Ilkka Rinne (Spatineo Inc. on behalf of the Finnish Meteorological Institute)

Frederic Guillaud (Météo-France)

Marie-Françoise Voidrot (Météo-France)

Minutes taken down by Ilkka Rinne

During this teleconference the proposals for standardizing the use of certain elements of WMS 1.3 capabilities document within the scope of the INSPIRE Data Specification for the themes "Atmospheric Conditions" (AC) and "Meteorological Geographical Features" (MF) were discussed. A member if the INSPIRE AC-MF thematic working group Ilkka Rinne had previously sent an example WMS 1.3 capabilities document with the proposed elements to the OGC MetOcean DWG mailing list to get community feedback on these proposals before writing them down in final version (3.0) of the INSPIRE Data Specification, due to be ready on 20th April 2012.

Topics in the order of discussion:

The sampling dimension for the analysis time

The proposal of fixing WMS sampling dimension named "DIM_ANALYSIS_TIME" as the indicator of the results of a certain numerical model run was discussed. It was agreed that the sampling dimension name should should no the the "DIM_" prefix. The "DIM_" prefix should only be used in the WMS GetMap request for specifying the value of that dimension.

Mapping between the WMS layers and the property names (like air temperature) they visualize

A combination of WMS Layer elements "Identifier" and "AuthorityURL" were used in the Ilkka's capabilities document example for identifying the underlying visualized properties of a WMS layer. The "authority" attribute of Identifier element would point to a namespace of a certain standardized parameter vocabulary, like WMO GRIB codes or CF convention standard names. In the "AuthorityURL" element these namespaces would be bound to online resources for the vocabularies.

According to the Rinne's suggestion the values of the WMS Layer "Name" elements would not have an agreed or standardized meaning in the context of the INSPIRE AC-MF datasets.

Matthias Müller suggested that the "Keyword" element should be used to define this "class" type definitions instead of the "Identifier" more appropriate for an "instance" type identity management. Ilkka argued that it would probably not as be easy to point into an online resource for the vocabulary using Keyword elements than using the Identifier element combined with the AuthorityURL element.

It was agreed that the Keyword element would more suitable for this kind of definitions that the Identifier, even if it would be technically more difficult to discover the online resource defining those vocabularies (the need of an external registry would be needed).

Marie-Françoise Voidrot brought up that it was discussed in the previous OGC MetOcean telcos that the WMS layer names could be used for encoding the time and aggregation semantics of climatological data. This would mean that the internal structure of the Layer "Name" element would be used to convey the information of the used climatological time-related attributes for the dataset visualized in that layer. Rinne suggested that it would be better if there would be no internal data structure in the Name element, but a "Keyword" element with a specific "vocabulary" attribute could be used for this information. Frederic Guillaud agreed that this would seem to be a viable option.

Standardizing the use of MetadataURL, DataURL and FeatureListURL element within the AC-MF INSPIRE context

Ilkka Rinne presented examples of suggested use the WMS Layer elements MetadataURL, DataURL and FeatureListURL in the INSPIRE AC-MF context:

• MetadataURL: link a metadata record describing the underlying dataset of the layer, either using the INSPIRE Discovery Service or an online stand-alone metadata record in a standard format
• FeatureListURL: link the the INSPIRE Download Service Get Spatial Object operation retuning the OM_Observation instance the result (coverage) of which this layer is visualizing
• DataURL: direct link to downloading the result coverage as data that this layer is visualizing

Matthias Müller commented that the presented solution seem to very handy and neat for certain "smart" clients, but requiring or even recommending a particular way of using these element should be out of scope for an INSPIRE Data Specification, because it would make things more complicated for the data providers. Marie-Françoise Voidrot also thought that specifying the use of these elements is probably too detailed for the INSPIRE Data Specification.

Rinne promised to take these comments back to the INSPIRE AC-MF thematic working group discussion.

-- IlkkaRinne - 17 Jan 2012