[talk-au] Use of DEM data to generate streams

[Tom Brennan]

> Something to remember is that strahler order often used to make "bigger
> streams" is a measure of connectivity rather than importance. Like, a
> significant stream can have the same strahler order as a tiny trickle. In
> waterwaymap.org, it looks like this happens a bit (sudden jumps in stream
> width etc). Having said that I haven't looked into how the widths are
> computed.

As far as I can tell, the waterwaymap.org is based on upstream *distance* rather than something like strahler order.

> A possibly obvious contribution is that streams from a 10m DTM will be
> different from streams made with a 2m DTM. I also wonder if computing
> stream power (or some related metric) might also help? Thinking about
> rendering in terms of how much water might come here.

Some kind of rendering based on stream contribution would be better, but I think it's probably beyond what can be reasonably squeezed into OSM. The distance measure is probably not a bad proxy.

cheers
Tom
----
Canyoning? try http://ozultimate.com/canyoning
Bushwalking? try http://bushwalkingnsw.com