[GRASS-user] r.basin Issues Needing Resolution

[Rich Shepard]

On Sat, 7 Apr 2012, Margherita Di Leo wrote:

> r.basin produces several maps for every run, which are flow direction,
> accumulation, Horton ordered streams etc.. Prefix parameter is simply a
> string given by the user in order to distinguish all the maps produced by a
> single run of the program, i.e. every set of coords for outlet. For
> example,

madi,

   I thought this is the case, but wanted to know rather than assume.

> Threshold parameter is the same of r.watershed. 'Autothreshold ' flag'
> simply uses a threshold area of 1 km^2 and is intended as a tentative.

   Again, rather than assuming I wanted to be sure I understood.

> For what concerns r.stream.basin, it is a typo in the documentation,
> correct is r.stream.basins, thanks for pointing me out, I just updated it
> in the description file.
> r.wf.py and r.ipso.py are in the addOns repository:

   Ah, I see why they did not download for me: I added the .py extention to
the g.estension command. Now I know to leave it off.

> Please note that r.basin is intended to work with only one outlet at time.
> This means you should script it in a for cycle if you run it for several
> outlets.

   I understand this.

>  You can't give to r.basin the coordinates directly from a vector map.

   But, can I use the coordinates of a rasterized stream network?

> I have indicated to run r.stream.extract before r.basin because most of the
> times the calculated stream network doesn't match with the natural one, so
> that your coordinates should be adjusted to match the calculated stream
> network in order to get a result from r.basin. By the way, in a first run
> you can give a try with your coordinates.

   The outputs of r.stream.extract on two different project sub-watersheds
are, as far as I can see on the display, only about 5-10% of the vector
stream network, and all at the lower ends of the basins in which they
appear. I can send screen shots if you'd like to see the differences.

   After downloading and building the 'missing' three modules I'll try
v.to.rast on the natural stream network (at 1:24,000 scale, just like the
elevation data) and see what r.basin tells me.

Many thanks,

Rich