Skip to content

KnickPoints

Summary

Knickpoints detection based on Relative Slope Extension Index (RSE)

Tags:
longitudinal profile, elevation, slope, anomaly detection

Usage

Knickpoints correspond to anomalies in the elevation profile, and are expected to exhibit a sudden change in slope.

The elevation profile of a stream is normally such that the slope is inversely proportional to the distance to the source, which can be written :

$$ H = k . \ln L $$

where \(H\) is the height relative to the source, \(L\) the distance to the source, and \(k\) is a constant.

RSE is the ratio between the gradient index \(G\) and \(k\). \(G\) is computed for every stream stretch of length \(\Delta L\) as :

$$ G = \frac{\Delta H}{\Delta L} L $$

The higher the value of RSE at some point in the elevation profile, the more likely this point is a knickpoint.

Parameters

Parameter Description Type
Stream Network Aggregated by Hack Order with Z Coordinate
INPUT 		Stream network aggregated by Hack order (length-wise)
with elevation as z-coordinate,
oriented from upstream to downstream.
 LineStringZ(M)
No Data Value for Z
NODATA 		No-data value for z Number
Contour Interval
MIN_DZ 		Minimum z difference between two knickpoints
Default value: 5.0 		Number
Minimum Knickpoints RSE Value
MIN_RSE 		Output knickpoints having a RSE greater or equal than this value.
Default value: 2.0 		Number
Minimum RSE Total Value
MIN_RSE_TOTAL 		Process only tributaries having a global RSE
greater or equal than this threshold.
Default value: 1.0 		Number
Knickpoints
OUTPUT 		Detected knickpoints
with fields:
  • existing fields in INPUT
  • L: upstream channel length (ie. stream distance to source)
  • H: height relative to source = source z - z at knickpoint
  • DL: stretch length
  • DH: stretch height = z at start - z at end
  • HGI: Hack gradient index
  • RSE: Relative Slope Extension index
  • RSET: RSE total = profile height / ln(stream length)
 PointZ(M)

Example

Lorem ipsum …

See Also

References