According to the ISO standard 19109:2015, Geographic (geospatial) data is defined as data with implicit or explicit reference to a location relative to the Earth. In this context, we can think of the “data” as the “what” and the reference to a location relative to the Earth as the where. The when is also generally considered to be part of the data.
Implicit Reference to a Location Relative to the Earth.
While this formulation seems strange, it is rather simple. Implicit reference means that the data refers to a location on Earth using some name or code that refers to a location that is explicitly defined somewhere else. For instance, on the website of the Danish Statistical Office, we can find data describing the number of Contacts with a doctor covered by public health insurance by sex, region and time. In this data set, the location relative to the Earth is given implicitly by the name of the municipality while the data is the columns Male, female and Year.
| Minicipality | Male | female | Year |
|---|---|---|---|
| Compehagen | 2236743 | 3693462 | 2023 |
| Frederiksberg | 393603 | 681149 | 2023 |
| … |
For implicit referencing to work, there must be some standards for how different typical geometries, such as countries, municipalities, etc., are referenced. This is done through national and international spatial coding standards.
Explicit Reference to a Location Relative to the Earth
The following text is a geojson specification of the location of my office at the campus.
{ “type”: “Feature”, “properties”: { “What”: “Office” }, “geometry”: { “type”: “Point”, “coordinates”: [ 12.1413, 55.652] } }`
In this geojson text, the location relative to the Earth is given explicitly by the coordinates (Longitude and Latitude ). To ensure a correct reference relative to the Earth the coordinates alone, we also need to specify the coordinate system used for the coordinates. In the context of geospatial data, such a coordinate system is named a Coordinate Reference System (CRS).
In the above example, the coordinates represent a point in 2D, but more complex geometry types (lines and polygons) and higher dimensions can be used. While terms like Point, Line(Sting) and [Polygon] seem well-defined, you quickly run into more exotic terms such as “MultiCurve” or “MultiPoint”. In order to ensure the Interoperability and Reusability of data (the I and R of Open Geospatial Consortium (OGC)‘s FAIR principles) While Geospatial data typically consist of 0, 1 and 2-dimensional objects (Points, lines and polygons), 3D objects are beginning to appear, especially in geospatial data for urban planning (for more, see the note on CityGML) it is, however,, important to be clear that there is a distinction between the dimensions of the space and the dimension of the objects. The official Danish infrastructure data (GeoDanmark)is 0, 1 and 2D objects but represented in a 3D space, i.e. using XYZ coordinates. For more information on this topic, see Dimensions in Geospatial Data.
When it comes to working with more than 2D geometries, there are especially three use cases to take note of, namely:
A emerging area is the incorporation of real-time sensor dat in a geospatial context combining real-time data and 3D geospatial data comes close to the concept of a Digital Twin