Aggregate Level Terrain

Purpose

This article describes some differences between aggregate level and entity level terrains.

Rationale

In VR-Forces, the aggregate level and entity level models have different requirements for terrain.

The aggregate level model is more abstract, and typically relies on vector data.

The entity level model is more detailed, therefore terrain is represented at a higher resolution.

For example, the aggregate level models don’t care about a tree. They care that an area (defined by a vector polygon) is a forest.

Similarly, aggregate level models don’t care about a building. They care about areas that represent municipalities or industrial parks or similar things at that higher level of abstraction.

In fact, pointy rich-detailed three dimensional building models dotting some representation of a downtown urban area are disfavored by the aggregate level model, and can sometimes cause undesirable results. To the aggregate level model, those things are considered part of the surface. It’s better to simply use an elevation layer to describe the terrain surface, without all the pointy buildings. Instead, represent urban areas with a vector feature layer.

Implementation

MAK VR-Forces and its associated MAK One Data include an example aggregate level terrain called “MAK Earth Aggregate (online)”.

It consists of:

• elevation

• imagery

• a topographic map layer

• ocean vector features

• lake vector features

• river linear features

• Open Street Map road features

• Building footprints as area vector features

• A vector layer that circumscribes “Land”.

• A vector layer that defines shallow coastline.

• A vector layer that defines hilly areas.

• A vector layer that defines mountainous areas.

• A vector layer that defines alpine areas.

• A forest layer.

• A cultivated field layer.

• A layer to represent municipalities (urban areas).

• Grassland areas.

• Tundra areas.

• Swamp areas.

• Beach areas.

• Desert areas.

• Railroads.

Alternatives

“Will the aggregate level model run on any VR-Forces terrain?”

Yes, but not well.

To the extent that any of the above terrain concepts are represented in vector data, the aggregate model will use them.

The aggregate level model does not use soil type information as encoded in textures on polygons. It uses vector features to affect mobility and sensors.