0

I'm in the process of tracing various river systems in Great Britain, and I've arrived at a problem. At the town of Earith, the Old and New Bedford Rivers diverge from the main course of the River Great Ouse (5th order), and I have no idea how to continue my stream order system from here.

Do all three distributaries retain fifth order? If so, what happens when tributaries join and how do I avoid artificial addition of stream orders downstream? How do I justify keeping the original Great Ouse course 5th order when it is clearly a smaller channel at that point (owing to drainage down the other channels)?

I should clarify I am using the Strahler number system.

Illustration

Shaun
  • 44,997
Will
  • 1
  • It would seem your rivers form a tree graph, which establishes a partial order. (https://en.wikipedia.org/wiki/Tree_(graph_theory) and https://en.wikipedia.org/wiki/Partially_ordered_set). – David G. Stork Nov 09 '20 at 21:48
  • You don't say which stream order you are using. This https://en.wikipedia.org/wiki/Stream_order might help you find the name of the order you are using. (You may also find the answer there.) – Eric Towers Nov 09 '20 at 21:48
  • @DavidG.Stork I'm not familiar with partially ordered sets, but I can see what its getting at. However, (referring to the image at the top of the Wikipedia article on partially ordered sets) I don't know what my {x, y, z} is because the confluence of the three distributaries carries more water than it did before (due to tributaries). – Will Nov 09 '20 at 21:57
  • The $x,y,z$ coordinates are irrelevant to the tree graph. All that matters is the flow of water... from which stream into which stream. Think about this: A geneological (family) tree, from grandparents, to parents (and uncles) to children. This forms a partial order in that a father is "before" (above) children, but it makes no sense to say which of distance cousins is "above" another. – David G. Stork Nov 09 '20 at 22:01
  • @DavidG.Stork I don't see how that's relevant to the question. I can't just neglect a watercourse because it's a distributary. – Will Nov 10 '20 at 20:56
  • @Will: No one is asking you to "neglect a watercourse." Exactly the opposite. Include all watercourses, but IGNORE THE SPATIAL LOCATIONS OF THE INTERSECTIONS. This is a problem in graph theory (or topology) where you could distort the landscape however you like (without tearing or "sewing"), and the watercourse relationships remain unchanged. – David G. Stork Nov 10 '20 at 21:01
  • You mean to treat the separate channels as one in terms of stream order? That's what I started to think about. I believe I may have solved the issue by simply subtracting the stream order by one at the point of divergence. That should avoid artificial additions when they recombine (and indeed seems to). Thanks for your help. – Will Nov 12 '20 at 10:33

0 Answers0