[GRASS-user] r.surf.contour (again)

GRASS GRASS user
1

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I'd like to create a DEM from contours using r.surf.contours. Maybe I
am trying something to ambitious here because the region is quite
large and the relief is rather complex with lots of human made
modifications to the terrain. I used standard contours and contours
describing terrain features in combination with elevation points.

The region is:

nsres: 1
ewres: 1
rows: 12000
cols: 16000
cells: 192000000

and the calculation took amazing five days (or was it six days?).
Anyway, the result is unfortunately not what I would like to have
gotten. There links to three images with accompanying contour lines
[1][2][3] for some of the problems I encounter with the result, which
is btw. a hillshade computed from the DEM. To me it seems that the
contour lines are sometimes to dense and sometimes to far spaced for
the algorithm. And there are those straight line artefacts you can
find all over the place and some patterns that vaguely look like
quadtree pattern. And then there are those contour like black lines,
especially in [1] and [3] which I really can't explain.

Well I am willing to give it a second try and now I wonder what I can
do better. Increase resolution? Leave out some contour lines?

So any help is greatly appreciated :slight_smile:

Frank

PS
Grass is version 6.4.3 on Ubuntu 12.04

[1] http://www.geometrico.lu/devel/grass/hillshade1.jpeg
[2] http://www.geometrico.lu/devel/grass/hillshade2.jpeg
[3] http://www.geometrico.lu/devel/grass/hillshade3.jpeg

- --
Frank BRONIEWSKI

METRICO s.à r.l.
géomètres
technologies d'information géographique
rue des Romains 36
L-5433 NIEDERDONVEN

tél.: +352 26 74 94 - 28
fax.: +352 26 74 94 99
http://www.metrico.lu
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJTYLuXAAoJEHopqW0d1cQVPU8IAKDS5E2lC8txxVrMOSkzEBAA
cKVTTOghKJt2K63hVO7TGrAwaK2LFVAKsdeb+wLMS/TisnP5cFlCrN8lZNzGNxg8
v1+o4qZBgUHZP6RmIHsAQzOgTlNHqjlK8vxvWWdtTOaf4GESQlBbUehzF2nW6eRg
cIbPp1H3Tbel2wSnEM9nuc/Kqy9/wgGwrf9G0N2C7XeM2CnfMFRyj45e4BILjQMb
tPpyLfnhdjOk3Eg+3PiC5CtZDpn6AFBxFqq6F6yDUxm5MrUJe6uvKqADTd8dReDs
Nwj2xOwLRGQdIXayeHHnam2HYtzjXQO+UEhu+v+SRqWJEb6lSR1u8OvP3uG0PbE=
=RAFt
-----END PGP SIGNATURE-----

2

Hi.

Well I am willing to give it a second try and now I wonder what I can
do better. Increase resolution? Leave out some contour lines?

To speed up the process to see if the results are desirable, you could
increase the resolution and try the add-on r.surf.nnbathy. Then decrease
resolution if necessary and applicable.
Mark

3

Am 2014-05-01 15:27, schrieb Mark Seibel:

Hi.

Well I am willing to give it a second try and now I wonder what I can
do better. Increase resolution? Leave out some contour lines?

To speed up the process to see if the results are desirable, you could
increase the resolution and try the add-on r.surf.nnbathy. Then decrease
resolution if necessary and applicable.
Mark

Thanks for the tip, I will try the add-on. Any idea what is causing the
distortion in the maps? It would interest me if it is a resolution issue
or the data itself ....

Frank

--
Frank BRONIEWSKI

METRICO s.à r.l.
géomètres
technologies d'information géographique
rue des Romains 36
L-5433 NIEDERDONVEN

tél.: +352 26 74 94 - 28
fax.: +352 26 74 94 99
http://www.metrico.lu

4

Thanks for the tip, I will try the add-on. Any idea what is causing the
distortion in the maps? It would interest me if it is a resolution issue
or the data itself …

It’s a function of the contour lines and interpolator. Contour lines are poor input for building DEMs because they have high density of input points along the line with no data between them. When the distance is too large some interpolator create sharp angle artifacts in the process. Other interpolators smooth the transition areas better.

Mark