[GRASS-dev] [GRASS GIS] #1708: r.fill.dir leaves unresolved areas

#1708: r.fill.dir leaves unresolved areas
---------------------------+------------------------------------------------
Reporter: marchesinivan | Owner: grass-dev@…
     Type: defect | Status: new
Priority: normal | Milestone: 7.0.0
Component: Raster | Version: svn-trunk
Keywords: | Platform: Linux
      Cpu: x86-64 |
---------------------------+------------------------------------------------
We have run r.fill.dir on a grid layer of 16 milions cells.
Using grass70 the procedure stops leaving a lot of unresolved areas.
Using grass64 the procedure works regularly.

--
Ticket URL: <http://trac.osgeo.org/grass/ticket/1708&gt;
GRASS GIS <http://grass.osgeo.org>

#1708: r.fill.dir leaves unresolved areas
---------------------------+------------------------------------------------
Reporter: marchesinivan | Owner: grass-dev@…
     Type: defect | Status: new
Priority: normal | Milestone: 7.0.0
Component: Raster | Version: svn-trunk
Keywords: | Platform: Linux
      Cpu: x86-64 |
---------------------------+------------------------------------------------

Comment(by marchesinivan):

Moreover, iterating the command in grass grass70 do not decrease the
unresolved areas as instead happen using grass64

--
Ticket URL: <http://trac.osgeo.org/grass/ticket/1708#comment:1&gt;
GRASS GIS <http://grass.osgeo.org>

#1708: r.fill.dir leaves unresolved areas
---------------------------+------------------------------------------------
Reporter: marchesinivan | Owner: grass-dev@…
     Type: defect | Status: new
Priority: normal | Milestone: 7.0.0
Component: Raster | Version: svn-trunk
Keywords: | Platform: Linux
      Cpu: x86-64 |
---------------------------+------------------------------------------------

Comment(by mmetz):

Replying to [comment:1 marchesinivan]:
> Moreover, iterating the command in grass grass70 do not decrease the
unresolved areas as instead happen using grass64

This seems to be different bugs in different branches related to the -f
flag, "Find unresolved areas only".

For grass64, problem areas are resolved if the flag is not given, whereas
for grass7, problem areas are resolved if the flag is given. Fixed for
trunk in r52872.

BTW, if you want to prepare a dem for hydrological analysis, r.fill.dir is
not really appropriate because it uses a method based on questionable
assumptions (sink filling). In this case maybe you could give r.hydrodem
from addons a try (impact reduction approach).

Markus M

--
Ticket URL: <http://trac.osgeo.org/grass/ticket/1708#comment:2&gt;
GRASS GIS <http://grass.osgeo.org>