[GRASS-user] Re: understanding r.watershed

[Markus Metz]

Georg Kaspar wrote:
> ok, so I only need to specify depressions that do not loose water!? 
> Artificial depressions like postholes or pits for example!? 
Rather something with a few hectares in size (at least), e.g. water 
reservoirs where water is pumped out, no overground outflow. Postholes 
and small pits are not able to distort r.watershed. The best example for 
a real depression is the Dead Sea, a big lake with no outflow.
> What would be 
> the difference between purging the DEM from those depressions and 
> providing them as binary input for r.watershed?
>   
Generally there is no need to purge the input DEM from depressions, why 
do you insist on doing so?
The difference between purging a DEM and providing  an option 
depression=<map with real depressions> has been explained before in this 
thread [1], [2].

[1] http://lists.osgeo.org/pipermail/grass-user/2009-January/048604.html
[2] http://lists.osgeo.org/pipermail/grass-user/2009-January/048605.html