[GRASS-dev] Re: [GRASS GIS] #807: r.watershed doesnt consider longer distance to diagonal neighbouring pixels

[GRASS GIS]

#807: r.watershed doesnt consider longer distance to diagonal neighbouring pixels
-----------------------+----------------------------------------------------
  Reporter:  aread     |       Owner:  grass-dev at lists.osgeo.org
      Type:  defect    |      Status:  new                      
  Priority:  critical  |   Milestone:  6.4.0                    
 Component:  Raster    |     Version:  6.4.0 RCs                
Resolution:            |    Keywords:                           
  Platform:  All       |         Cpu:  All                      
-----------------------+----------------------------------------------------
Comment (by mmetz):

 Fixed in trunk r39722 and r39724 and in devbr6 r39726 and r39727.

 I followed the description in Ehlschlaeger (1989) to the letter and the
 results seem to be correct now. I want to wait for independent
 confirmation before backporting to relbr6. Using elev_lid792_1m in the
 North Carolina dataset, the SFD results of 64 and 65 are strikingly
 different and IMHO the 64 results look wrong compared to the 65 results.
 Diagonal bias in flow direction is gone, now it's close to 50-50.

 I noticed that a correction for diagonal bias was implemented in the
 original version as in grass3.2 of 1990. Does anybody know why it was
 removed? SVN history doesn't go back far enough.

 MFD mode in grass65 and grass7 was only slightly affected, getting out of
 sinks should be improved by now.

 Markus M

-- 
Ticket URL: <https://trac.osgeo.org/grass/ticket/807#comment:1>
GRASS GIS <http://grass.osgeo.org>