[GRASS-user] unable to reproduce r.water.outlet example of GRASS book

GRASS GRASS user
1

Hello,

I'm trying to reproduce an exercise of the GRASS book, third edition, pp. 145-146, which demonstrates the usage of r.water.outlet. Using exactly the same commands as listed in the book, I get a basin which is hardly visible unless zooming very closely to it. r.report reports a size of 0.0002 ha. Is this normal ?

The relevant commands in the book are (for the NC dataset):

g.region rural_1m -p (or g.region rast=elev_lid792_1m)
r.watershed elev_lid792_1m thresh=5000 accum=accum_5K \
             drain=draindir_5K basin=basin_5K
r.water.outlet drainage=draindir_5K basin=basin_A30 \
                east=638872.6 north=220042.6

Can anybody else reproduce this ?

If yes, could changes in r.watershed have had an influence on these results ?

Moritz

2

Moritz Lennert wrote:

Hello,

I'm trying to reproduce an exercise of the GRASS book, third edition, pp.
145-146, which demonstrates the usage of r.water.outlet. Using exactly the
same commands as listed in the book, I get a basin which is hardly visible
unless zooming very closely to it. r.report reports a size of 0.0002 ha. Is
this normal ?

The relevant commands in the book are (for the NC dataset):

g.region rural_1m -p (or g.region rast=elev_lid792_1m)
r.watershed elev_lid792_1m thresh=5000 accum=accum_5K \
drain=draindir_5K basin=basin_5K
r.water.outlet drainage=draindir_5K basin=basin_A30 \
east=638872.6 north=220042.6

The version of r.watershed used for the example in the book had some
bugs, in particular a bias towards diagonal drainage direction which
is mosz notable with high-res LiDAR DEMs like elev_lid792_1m. One
consequence is that accumulation patterns shifted a bit and
east=638872.6 north=220042.6 are most probably no longer falling onto
a main channel leading to minute basins. Try plotting east=638872.6
north=220042.6 on top of accum_5K and adjusting the coords such that
they fall onto the nearest main channel.

HTH,

Markus

3

On Wed, Sep 7, 2011 at 8:22 AM, Markus Metz
<markus.metz.giswork@googlemail.com> wrote:

Moritz Lennert wrote:

Hello,

I'm trying to reproduce an exercise of the GRASS book, third edition, pp.
145-146, which demonstrates the usage of r.water.outlet. ...

The version of r.watershed used for the example in the book had some
bugs, in particular a bias towards diagonal drainage direction which
is mosz notable with high-res LiDAR DEMs like elev_lid792_1m. One
consequence is that accumulation patterns shifted a bit and
east=638872.6 north=220042.6 are most probably no longer falling onto
a main channel leading to minute basins.

...

I have added NULL support in r.water.outlet and added a working example to
the manual page [1]. Works nicely.

Markus

[1] http://trac.osgeo.org/grass/browser/grass/branches/releasebranch_6_4/raster/r.water.outlet/description.html
     http://trac.osgeo.org/grass/browser/grass/branches/releasebranch_6_4/raster/r.water.outlet/r_water_outlet.png
(in the online manual on Saturday after the automated compilation on the
GRASS server)

4

On 07/09/11 09:33, Markus Neteler wrote:

On Wed, Sep 7, 2011 at 8:22 AM, Markus Metz
<markus.metz.giswork@googlemail.com> wrote:

Moritz Lennert wrote:

Hello,

I'm trying to reproduce an exercise of the GRASS book, third edition, pp.
145-146, which demonstrates the usage of r.water.outlet. ...

The version of r.watershed used for the example in the book had some
bugs, in particular a bias towards diagonal drainage direction which
is mosz notable with high-res LiDAR DEMs like elev_lid792_1m. One
consequence is that accumulation patterns shifted a bit and
east=638872.6 north=220042.6 are most probably no longer falling onto
a main channel leading to minute basins.

...

I have added NULL support in r.water.outlet and added a working example to
the manual page [1]. Works nicely.

Yep. Thanks !

Moritz