I have the boundaries and centerline of a river.

The Thalweg line defines the deepest elevation of a watercourse (see Image taken from Google). I do not have accurate enough bathymetry or Digital Elevation Models to determine the proper deepest elevation of a river channel.

I would like to know if it is possible determine the Thalweg line of the river without the bathmetry?

enter image description here

It looks like the Thalweg line follows a predictable pattern where it hugs the outside curve of a river when it meanders and is in the center of the river with it is straight.

I am using ArcGIS 10.1 with all licenses available, and I am familiar with python.

  • 4
    The thalweg is going to be closer to the outside of the bend in a meander, but its exact location can vary. The straight line segments are the bigger issue, where the thalweg can meander considerably (which ultimately creates new meanders in the river). Cross-section stations would help considerably if you can get those, otherwise it is not really a solvable problem with just centerline and boundary. Feb 2 '16 at 18:56
  • I am unable to get any cross-section stations. I am more concerned about determining the outside curve of a river using the Thalweg (Highest chance of erosion) than where it falls on a straight section.
    – JC11
    Feb 2 '16 at 19:18
  • 1
    @JC11 if position of outside curve is main objective you might want to calculate angles at each vertex
    – FelixIP
    Feb 2 '16 at 19:22

You don't really give a scale at which you are working at? The entire Amazon or a section of a river as suggested by the image above? Without Bathemetry data or cross sections I would imagine computing a Thalweg would be very difficult, if not impossible (personally never done it). Also what about the geology, bedrock outcrops would constrain the rivers planform.

Just looking at the image I started thinking about radius of curvature as a possible way of measuring erosion rates? This article on Grays River talks about radius of curvature. A quick google search will throw up other articles/journal papers.

If the extent of your data is really what is in your image then you can simply draw the circles by hand?

Just an idea?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.