Fundamentals of 3D data
Three-dimensional GIS data incorporates and extra dimension—a z-value—into its definition (x,y,z). Z-values have units of measurement and allow the storage and display of more information than traditional 2D GIS data (x,y). Even though z-values are most often real-world elevation values—such as the height above sealevel or geological depth—there is no rule that enforces this methodology. Z-values can be used to represent many things, such as chemical concentrations, the suitability of a location, or even purely representative values for hierarchies.
There are two basic types of 3D GIS data: feature data and surface data.
3D feature data
Feature data represents discrete objects, and the 3D information for each object is stored in the feature's geometry.
Three-dimensional feature data can support potentially many different z-values for each x,y location. For example, a vertical line has an upper vertex and a lower vertex, each with the same 2D coordinate, but each having different z-values. Another example of 3D feature data would be a 3D multipatch building, whose roof, interior floors, and foundation would all contain different z-values for the same 2D coordinate. Other 3D feature data, such as an aircraft's 3D position or a walking trail up a mountain, would only have a single z-value for each x,y location.
Surface data
Surface data represents height values over an area, and the 3D information for each location within that area can be either stored as cell values or deduced from a triangulated network of 3D faces.
Surface data is sometimes referred to as 2.5D data because it supports only a single z-value for each x,y location. For example, the height above sealevel for the surface of the earth will only return a single value.
When to model GIS data in 3D
Since 3D GIS data can be more difficult to create and maintain than 2D data, modeling your data in three dimensions should only be done when the extra effort will add value to your work. While some GIS features, such as aircraft locations or underground wells, naturally lend themselves to being modeled in 3D, other data can be just as effective in 2D as in 3D. For example, having a road network modeled in 3D might seem useful for investigating gradients, but the additional effort to maintain z-values might outweigh the benefits.
These are some important considerations when deciding to model your data in 3D:
- GIS data does not have to be modeled in 3D to be displayed inside a 3D view.
- Height values from a surface can easily be added to 2D objects, when you need them, through the use of geoprocessing tools.
- If the source of your z-values is a surface, consider how often that underlying surface changes. The more it changes, the less useful it will be to store z-values for features generated against it.
If you decide to model some or all of your data in three dimensions, the most important decision will be the units of the z-values. A solid understanding of what your z-values represent will be critical when you start editing and maintaining them. A general rule to follow whenever possible is that the z-units should match your x,y units. For example, if your data is in a (meter-based) UTM zone, you should also model your z-values as meters. This will help you interact with the data in an intuitive way, such as when you measure 3D distances or move objects in x, y, and z.