aspect and drainage direction maps

[Xavier Ros]

        Many thanks to Philip Verhagen, who answered my question about aspect 
and drainage maps generation algorithms. After his help I could arrive to 
some conclusions:
        The aspect map generated by r.slope.aspect seems to be calculated 
using the Horn method (as it says the manual), but the results I got 
when I made the calculations by hand don't were exactly the same as the 
ones I got using the Grass4.1.3 version of this module. At least, if I'm aplyin
by hand the Horn method in a right way.I am following this method such as 
it is explained by Shapiro and Westerwelt in r.mapcalc algebra document. 
        I think the drainage map generated by r.watershed gives the direction 
to the lowest neighbour cell, unless there is a pit. In this case it works in 
a little different manner, so that it avoids the formation of false pits in the
drainage network. Care must be taken with the region settings to allow
it to work well.
        The drainage direction maps generation by r.direct works in
a way I don't understand. It seems clear that it doesn't avoid pits, and that 
the algorithm doesn't depend on the way the results will be expressed
(answers, agnps or grass4.0). Does anybody know anything about the
way r.direct generates its maps?


                Xavier Ros
        Unitat de Geodinamica Externa i d'Hidrogeologia
        Universitat Autonoma de Barcelona