Data Maps are special masks. While Selector masks generally represent an area or selection - such as slope, angle, or curvature; Data Maps represent a specific type of data of the area - such as water flow, soil deposits, etc.
These nodes break the traditional basic data (slope, angle, curvature) + chaos (Perlin noise) method used to create pseudo-random masks from which color maps would be generated. In such situations, you need to find a good seed to make it work. With Data Maps, there is visual randomness, but it is created from systematic analysis of the terrain and follows natural principles. The resulting color maps created from such data look more believable.
Data Maps are very useful for creating color maps. For example, the water flow can be used to create a Biome texture, or used manually with a CLUTer or SatMaps node to create a texture along water flow areas.
In this example, we create a simple terrain using a Mountain node, then erode it to get some basic erosion data. On top of the eroded output, we apply a Soil Map node. The Soil Map node is fed to a SatMaps node to create a base texture for the terrain. The Flow and Wear output of the Erosion node are mixed and fed to another SatMaps node to create a second map. The two maps are mixed together to create the final texture.
// TODO: Image
See the @color-datamaps for further information.