Feature model of WIGOS metadata - wmo-im/wmdr GitHub Wiki

The feature model for WIGOS metadata (the WIGOS MetaData Representation WMDR) is a profile of ISO/OGC Observations and Measurements (O&M) and is expressed as an XML Schema Definition (XSD). O&M models an observation as a combination of various aspects of that may influence the result. The WMDR interprets the O&M model to describe either an individual observation or a set of observations. The various aspects describing an observation are combined below using the operator 'x'.

  • OM_Observation = featureOfInterest x observedProperty x parameter(s) x OM_Process
  • featureOfInterest is interpreted as domain, currently including elements {Atmosphere|Terrestrial|Earth|Ocean|Outer Space} - NB: See #277 for discussion.

Version 1.0RC9 of WMDR (currently implemented in OSCAR/Surface) describes the observedProperty as

  • observedProperty = variable x matrix x size_range

Other aspects relevant to the properties of the particle phase of aerosol can be described in OSCAR/Surface using a code list for sampling. The desire to normalize the list of entries in observedVariableAtmosphere, while remaining compatible with O&M has led to some discussion in the Working Group on Atmospheric Composition Variables WG-ACV. This has proven to be most challenging for observations relating to particles in the atmosphere. Eventually, the need to document more aspects of a measurand in a consistent manner has led to the following approach:

  • observedProperty = variable x matrix
    • matrix = the specific physical phase of interest for the observation - NB: See #278 for discussion.
    • variable = biogeochemical species or physical quantity
  • parameter(s) = geometry x particleSizeRange x sizing methodology x pressure x temperature x humidity x wavelength/frequency
    • geometry = {point|line|area|volume|vertical profile|total column} as defined in Geometry
    • particleSizeRange - see #270 for discussion.

O&M supports the description of properties as name x value pairs. These will use controlled vocabularies for the most part. In this way, the validation of WMDR metadata records can be facilitated.

  • OM_Process = sampling x processing x reporting
  • sampling = sampling_strategy (needs further discussion)
  • processing = sampling x sample_treatment x method x method_details x instrument (x operator)
    • sample_treatment = manipulation of sample prior to analysis. Most relevant to grab samples, but also applicable to others.
    • method = bio/physico/chemical principle used to make the observation
    • method_details = free text describing specifics of the method of observation
    • instrument = manufacturer x model x serial number
    • operator = an individual involved in making the observation
  • reporting = pressure x temperature x humidity x unit x ...

Comparison with other approaches

CF (Climate and Forecast)

CF uses an approach of concatenating features describing a quantity with underscores. The intention is to provide a definitive description of what the data in each variable represents, and the spatial and temporal properties of the data. The primary focus is support for the NetCDF API.

Research Data Alliance

RDA has established a Working Group on InteroperAble Descriptions of Observable Property Terminology (I-ADOPT). The approach is similar to WMDR. The exact mapping and limitations thereof need to be further explored. ...

INSPIRE

INSPIRE is a European Directive for spatial information in Europe. INSPIRE has a theme on Environmental Monitoring Facilities that is described as a UML model built on O&M.