Build RELATIVE Digital Elevation Models

(currently only .BIL-Files)

For all values Z(x,y) in a DEM the following filter rule is applied:

	1. Calculating the mean Elevation Z_mean(x,y)
	   within a Radius R

	2. Z_relative(x,y) =  Z(x,y) - Z_mean(x,y)

Radius R:

The radius influences the position of  the local extreme values. Good results can be achieved by using the average autocorrelation length of an elevation grid. Normally 15m and 35m in all tested elevation models.

(-> See Tool: BIL-Analysis)

Relative DEM:

- Each point shows its relative elevation in meters above (+) / below(-) its surrounding points within 25m radius
- Upper and Lower shoreline emphazised

Relative Elevations Model

Original DEM:

based on LIDAR data and echo sounders:
- Elevation in meters above see level
- No clear lower shore line detectable

Originally measured Elevation Model


-> Visual detection of terrain features

Relative Elevation Models are exspecially usefull to detect points or edges that dominate its sourrounding. This  indecator, easy to visualize and easy to understand, can be used instead of weighted height, slope and curvature  information to detect terrain features. It works like the “unsharp mask” for picture processing.

It is useful for visual identification of features with gradual transitions into the surrounding terrain:
- Shorelines
- Flat groyne contours and heads
- Flat dams and dikes and ditches

-> Speeding up the meshing process

Relative Elevation Models are also used internally for terrain adaption during the meshing process. You can speed up the meshing process by providing a relative elevation modell, once generated bevorehand. The meshing algorithm
then can use the precalculated values instead of a on the fly calculation during every iterative meshing step.

Comments are closed.