Property Reference

GeoCroissant introduces a set of specialised metadata properties that extend the core Croissant vocabulary with geospatial concepts. This page provides a full technical reference for each property, including its domain, range, cardinality, and usage examples.

Namespaces

The GeoCroissant vocabulary lives under its own namespace:

http://mlcommons.org/croissant/geo/

This is typically abbreviated using the prefix geocr. The table below lists all namespaces used in GeoCroissant metadata.

Prefix IRI Description
sc http://schema.org/ The schema.org namespace
cr http://mlcommons.org/croissant/ MLCommons Croissant base namespace
geocr http://mlcommons.org/croissant/geo/ GeoCroissant extension namespace
dct http://purl.org/dc/terms/ Dublin Core Terms

The specification itself is versioned via its URI: http://mlcommons.org/croissant/geo/1.0. Note that the vocabulary namespace is not versioned — property IRIs remain stable across versions of the specification.


Spatial Metadata

geocr:coordinateReferenceSystem

The coordinate reference system (CRS) or projection used to interpret spatial data. This is essential for any dataset whose records need to be correctly aligned and compared across tools and platforms.

Type sc:Text
Domain sc:Dataset
Cardinality ONE
{
  "@type": "Dataset",
  "geocr:coordinateReferenceSystem": "EPSG:4326"
}

geocr:spatialResolution

The nominal ground sampling distance — how much real-world area one pixel covers. Can be set at dataset level for uniform imagery, or at record level when resolution varies across records (e.g., Harmonized Landsat/Sentinel-2).

Type sc:Text or sc:QuantitativeValue
Domain sc:Dataset, cr:RecordSet
Cardinality ONE

Dataset level:

{
  "geocr:spatialResolution": { "@type": "QuantitativeValue", "value": 30, "unitText": "m" }
}

Record level (when resolution varies):

{
  "recordSet": [{
    "@type": "cr:RecordSet",
    "field": [{ "@type": "cr:Field", "name": "spatialResolution", "dataType": "sc:Text" }],
    "data": [
      { "spatialResolution": "10m" },
      { "spatialResolution": "30m" }
    ]
  }]
}

geocr:spatialIndex

Precomputed spatial index tokens (e.g., H3 cells, DGGS identifiers, geohashes) that support fast coarse spatial filtering before running exact geometry operations. Particularly useful for large catalogs.

Type sc:Text
Domain sc:Dataset, cr:RecordSet
Cardinality MANY
{
  "geocr:spatialIndex": ["h3:8a2a1072b59fff", "h3:8a2a1072b5bfff"]
}

Spectral & Band Metadata

geocr:bandConfiguration

Describes how bands are organised within a raster dataset — how many there are, in what order, and what they represent. Can be set at dataset or field level.

Type geocr:BandConfiguration
Domain sc:Dataset, cr:Field
Cardinality ONE
{
  "geocr:bandConfiguration": {
    "@type": "geocr:BandConfiguration",
    "geocr:totalBands": 6,
    "geocr:bandNameList": ["Blue", "Green", "Red", "NIR", "SW1", "SW2"]
  }
}

geocr:totalBands

The total number of bands in the raster asset.

Type sc:Integer
Domain geocr:BandConfiguration
Cardinality ONE
"geocr:totalBands": 6

geocr:bandNameList

An ordered list of band names corresponding to the raster band stack.

Type sc:Text
Domain geocr:BandConfiguration
Cardinality MANY
"geocr:bandNameList": ["Blue", "Green", "Red", "NIR", "SW1", "SW2"]

geocr:spectralBandMetadata

Per-band spectral descriptors — center wavelength, bandwidth, and similar physical quantities. Use this when downstream systems or models need band physics beyond simple names.

Type geocr:SpectralBand
Domain sc:Dataset, cr:Field
Cardinality MANY
{
  "geocr:spectralBandMetadata": [
    {
      "@type": "geocr:SpectralBand",
      "name": "Blue",
      "geocr:centerWavelength": { "@type": "QuantitativeValue", "value": 490, "unitText": "nm" },
      "geocr:bandwidth": { "@type": "QuantitativeValue", "value": 98, "unitText": "nm" }
    },
    {
      "@type": "geocr:SpectralBand",
      "name": "NIR",
      "geocr:centerWavelength": { "@type": "QuantitativeValue", "value": 865, "unitText": "nm" }
    }
  ]
}

geocr:centerWavelength

The center wavelength for a spectral band.

Type sc:QuantitativeValue
Domain geocr:SpectralBand
Cardinality ONE
"geocr:centerWavelength": { "@type": "QuantitativeValue", "value": 490, "unitText": "nm" }

geocr:bandwidth

The spectral bandwidth (full-width at half-maximum) for a band entry.

Type sc:QuantitativeValue
Domain geocr:SpectralBand
Cardinality ZERO or ONE
"geocr:bandwidth": { "@type": "QuantitativeValue", "value": 98, "unitText": "nm" }

Temporal Metadata

geocr:temporalResolution

The temporal cadence of the collection — how frequently observations are taken. Can be set at dataset or record level when cadence varies.

Type sc:Text or sc:QuantitativeValue
Domain sc:Dataset, cr:RecordSet
Cardinality ZERO or ONE
"geocr:temporalResolution": { "@type": "QuantitativeValue", "value": 1, "unitText": "month" }

geocr:timeSeriesIndex

Identifies the field within a RecordSet that provides the temporal ordering key — typically a timestamp. This allows downstream systems to sort, slice, and batch time-series records reliably.

Type sc:Text
Domain cr:RecordSet
Cardinality ZERO or ONE
{
  "@type": "cr:RecordSet",
  "geocr:timeSeriesIndex": { "@id": "timeseries_recordset/timestamp" },
  "field": [
    { "@type": "cr:Field", "@id": "timeseries_recordset/timestamp", "name": "timestamp", "dataType": "sc:DateTime" }
  ]
}

Responsible GeoAI

geocr:spatialBias

A plain-language description of any known geographic representativeness limitations that could affect model generalisation — for example, if coverage is concentrated in one region or continent.

Type sc:Text
Domain sc:Dataset
Cardinality ZERO or ONE
"geocr:spatialBias": "Coverage is concentrated over the contiguous United States; high-latitude regions and other global biomes are under-represented."

geocr:samplingStrategy

Describes how training samples were constructed — including windowing strategies, filtering criteria, and any design choices that influence what the model learns.

Type sc:Text
Domain sc:Dataset
Cardinality ZERO or ONE
"geocr:samplingStrategy": "512×512 chips centred on burn polygons; cloudy scenes removed."

Programmatic Access

geocr:recordEndpoint

A URL pointing to a service where the dataset’s metadata records can be queried and retrieved programmatically. This is useful when GeoCroissant metadata is served dynamically from a catalog or registry (e.g., an OGC API – Records endpoint).

Clients can use this endpoint to list records, page through results, and apply spatial or temporal filters — for example by appending ?bbox=-125,24,-66,49 or ?datetime=2018-01-01/2021-12-31.

Type sc:URL
Domain sc:Dataset
Cardinality ZERO or ONE
"geocr:recordEndpoint": "https://example.org/ogcapi/records"

Space Weather Metadata

These properties support heliophysics and space weather datasets, where observations are described using wavelength channels and instrument characterisation rather than geographic coordinates.

geocr:multiWavelengthConfiguration

A composite wrapper describing the wavelength channels used in a space weather or heliophysics dataset, designed to simplify authoring and support mapping to SPASE constructs.

Type geocr:MultiWavelengthConfiguration
Domain sc:Dataset
Cardinality ZERO or ONE
{
  "geocr:multiWavelengthConfiguration": {
    "@type": "geocr:MultiWavelengthConfiguration",
    "geocr:channelList": ["171Å", "193Å"]
  }
}

geocr:solarInstrumentCharacteristics

A composite wrapper for solar instrument characterisation — observatory and instrument identifiers — to support mapping to SPASE.

Type geocr:SolarInstrumentCharacteristics
Domain sc:Dataset
Cardinality ZERO or ONE
{
  "geocr:solarInstrumentCharacteristics": {
    "@type": "geocr:SolarInstrumentCharacteristics",
    "geocr:observatory": "SDO",
    "geocr:instrument": "AIA"
  }
}

geocr:observatory

Name or identifier of the observatory or platform.

"geocr:observatory": "SDO"

geocr:instrument

Name or identifier of the instrument producing the measurement.

"geocr:instrument": "AIA"

geocr:channelList

The wavelength channels used in a multi-wavelength dataset.

"geocr:channelList": ["171Å", "193Å"]

Adding Custom Properties

If a metadata attribute is already defined in an established vocabulary (GeoSPARQL, SPASE, Dublin Core), reuse that term directly by declaring the appropriate prefix in @context.

If no suitable external term exists, express the attribute using sc:additionalProperty with sc:PropertyValue:

{
  "additionalProperty": [
    {
      "@type": "PropertyValue",
      "name": "cloudCoverMedian",
      "value": { "@type": "QuantitativeValue", "value": 12.3, "unitText": "%" }
    },
    {
      "@type": "PropertyValue",
      "name": "tilingScheme",
      "value": "MGRS"
    }
  ]
}