[GRASS-user] r.los

[Michael Barton]

Richard,

So this gives you yet a 3rd method. While interpolating a smooth DEM from
contour lines is a difficult problem (similar or worse problems reported
with ArcGIS), all 3 of these methods will do a much better job than using
v.extrude. 

You can also use a 2 step process. Use any of these 3 methods and take a
look at the results. To improve them you can use a neighborhood smoothing
routine, as Maciek mentions below. Or you can generate a set of random
points across your somewhat 'steppy' surface and do a new interpolation on
the basis of the random points to smooth things out. There are even more
routines to interpolate a surface from points.

The bottom line is that there are a variety of ways to solve this problem in
GRASS, giving you a lot of options to achieve the results you need.

Michael


On 7/31/07 4:14 PM, "Maciej Sieczka" <tutey at o2.pl> wrote:

> Michael Barton wrote:
>> Thanks for the clarification. This is good to know. Any other options
>> besides r.surf.contour?
> 
> I prefer r.surf.nnbathy, as I find natural neighbor or plain
> triangulation most robust for interpolating contour lines. However,
> these methods have their own problems - the output surface is not
> smooth (only partly - in case on nat. neigh.), kidney shape contours
> might be interpolated as plateus. But at least the result is
> predictable. If I want it smoother, I filter the DEM afterwards.
> 
> But that's just me.
> 
> Maciek

__________________________________________
Michael Barton, Professor of Anthropology
Director of Graduate Studies
School of Human Evolution & Social Change
Center for Social Dynamics and Complexity
Arizona State University

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton