Step 2: Temperature Parameter

Describing what our data values represent is fairly simple in CoverageJSON and is done with a Parameter:

{
  "type" : "Parameter",
  "unit" : {
    "symbol" : "°C"
  },
  "observedProperty" : {
    "label" : {
      "en": "Air temperature",
      "de": "Lufttemperatur"
    }
  }
}

There are two pieces of information in here:

• unit.symbol is an informal textual notation of the units of measurement, here degrees Celsius.
• observedProperty describes the abstract concept that is measured or modelled, independent of what units are used.

If data values have no units, then unit would be omitted. observedProperty.label however is always required and is a mapping from language tags to strings.

While the above is enough for humans to get a basic understanding, CoverageJSON allows us to describe a parameter in more detail, partially to make it easier for machines to identify related datasets and also to compare them with each other:

{
  "type" : "Parameter",
  "description": {
    "en": "Air temperature measured in degrees Celsius."
  },
  "unit" : {
    "label": {
      "en": "Degree Celsius"
    },
    "symbol": {
      "value": "Cel",
      "type": "http://www.opengis.net/def/uom/UCUM/"
    }
  },
  "observedProperty" : {
    "id": "http://vocab.nerc.ac.uk/standard_name/air_temperature/",
    "label" : {
      "en": "Air temperature",
      "de": "Lufttemperatur"
    },
    "description": {
      "en": "Air temperature is the bulk temperature of the air, not the surface (skin) temperature."
    }
  }
}

There are two additions in there which are specifically meant for machines:

• Units are now expressed using a standardised syntax scheme, in this case the Unified Code for Units of Measure (UCUM). The scheme is referenced by its global identifier http://www.opengis.net/def/uom/UCUM/. Machines are now able to understand the units and do transformations on them.
• The observed property now references a global identifier that uniquely identifies the concept and lets machines match it up with other datasets.

It is important to note that the observedProperty is typically reused across many datasets, while the rest of the parameter is often customized for a given dataset. This is why a parameter itself often does not have a global identifier.

Next stop, data values!