Triply Use Case 12: Represent the coordinate resolution per axis

[wouterbeek]

Triply Use Case 12: Represent the coordinate resolution per axis

Description

Coordinates often use floating-point numbers, where the decimal numbers do not necessarily reflect the resolution at which the data was originally measured. If a user does not know the resolution of the data, they do not know which decimal positions are meaningful and which are now.

In GeoSPARQL, it is currently possible to express the precision of all dimensions with properties geo:hasSpatialResolution and geo:hasMetricSpatialResolution. However, it is not possible to express the precision per dimension.

While this is already an issue for some 2D data, this is significantly more important for 3D data, because of the following reasons:

• The Z dimension more often has a precision that is different from the X and Y dimension.
• 3D BIM data is often very detailed, e.g. a 3D model of a pipe or valve, while GEO data is often less detailed, e.g. a building circumference or street.

Actor

• Users of 3D GeoSPARQL data.

Preconditions

• 3D GeoSPARQL data that is published on the web, whose Z axis has a precision that is different from the X and Y axes.

Postconditions

• When I query 3D GeoSPARQL data that is published on the web, I can see the resolution for each dimension. Specifically, I can see that individual 3D components of a building such as windows and doors have a higher resolution than the circumference of a building or road. I can see that the precision of the Z dimension is sometimes different that the precision of the X and Y dimensions of the same object.

Steps

1. As a data publisher, I can store the resolution of coordinates in a standardized way and per dimension, by using GeoSPARQL properties.
2. As a user of that data, I can query these properties.

[situx]

Hi @wouterbeek could you elaborate on why we need new properties here and why we could not use the properties geo:hasSpatialResolution or geo:hasMetricSpatialResolution?

[wouterbeek]

@situx Thank you for your useful feedback! I have updated the description accordingly. The issue is that it is not yet possible to express the precision per dimension. This was not clear at all in my original description. Feel free to review the new version of my issue as well.

[FransKnibbe]

It could be that spatial accuracy (geo:has(Metric)SpatialAccuracy) is more in need of specialisation than spatial resolution.