#807: r.watershed doesnt consider longer distance to diagonal neighbouring pixels ---------------------+------------------------------------------------------ Reporter: aread | Owner: [email protected] Type: defect | Status: new Priority: major | Milestone: 6.4.0 Component: Raster | Version: 6.4.0 RCs Resolution: | Keywords: r.watershed Platform: All | Cpu: All ---------------------+------------------------------------------------------ Comment (by mmetz):
Replying to [comment:5 hamish]: > Recommendation: wait to backport this to the 6.4.x line until after 6.4.0 has shipped, > and have it go out in the 6.4.1 release instead. otherwise we only have time for > cursory checks. Waiting for 6.4.1 buys us some time. Hmm, about time, nobody checked in the last 17 years or so? grass40 was released in 1992 I learned, and there the correction for diagonal flow bias was already disabled. It was probably not related to lat/lon projections because 1) I'm not sure if grass supported lat/lon back then, 2) there is a lot of slope calculation for the RUSLE factors that doesn't care about lat/lon locations and is still active. Making r.watershed lat/lon aware is (another) major rewrite... It was more probably related to a severe speed penalty imposed by the original implementation, avoided in the new implementation (which is also more in line with A* Search logic). It is really a pity that the commit comments from grass32 to grass40 are no longer available. I also did not find anything in the mailing list archives, although r.watershed was apparently regularly used. Hydrologists at CSIRO in Australia are currently looking at r.watershed in depth and tear the algorithm apart, looking at all sorts of special cases. So far the corrected version produces pretty much optimal results. > We could add a BUGS section to the 6.4.0 help page if needed. BTW, there is another bug in SECTION 4: Watershed determination, I'm busy fixing it. Markus M -- Ticket URL: <https://trac.osgeo.org/grass/ticket/807#comment:6> GRASS GIS <http://grass.osgeo.org>
_______________________________________________ grass-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-dev
