I am trying to create a DEM from a bathymetric contour lines. Contours are in vector format and the elevation value is negative. To create the surface I have converted the contours in raster (v.to rast) and then I have used r.surf.contour, but It doesn't work. I think that it would be because the negative values of the created raster. Could someone help me?
Add a comment
|
2 Answers
What are your region settings? I tried this set of commands, and succeeded to create a DEM with negative values, with no problems:
# Set region to low resoution
GRASS 7.0.0 (ITM):~ > g.region -p res=5
# The contour vector
GRASS 7.0.0 (ITM):~ > v.db.select test_ctours
cat|elev
1|-100
2|-80
3|-60
4|-40
5|-20
# Create rasterized contours
GRASS 7.0.0 (ITM):~ > v.to.rast test_ctours type=line use=attr attribute=elev out=test_ctours --o
# Create DEM
GRASS 7.0.0 (ITM):~ > r.surf.contour test_ctours out=test_dem
GRASS 7.0.0 (ITM):~ > r.info test_dem
+----------------------------------------------------------------------------+
| Map: test_dem Date: Fri May 8 13:02:06 2015 |
| Mapset: Arava Login of Creator: micha |
| Location: ITM |
| DataBase: /home/micha/GIS/grass |
| Title: ( test_dem ) |
| Timestamp: none |
|----------------------------------------------------------------------------|
| |
| Type of Map: raster Number of Categories: 0 |
| Data Type: DCELL |
| Rows: 200 |
| Columns: 300 |
| Total Cells: 60000 |
| Projection: Transverse Mercator |
| N: 500000 S: 499000 Res: 5 |
| E: 211500 W: 210000 Res: 5 |
| Range of data: min = -100 max = -20 |
| |
| Data Description: |
| generated by r.surf.contour |
| |
| Comments: |
| r.surf.contour input="test_ctours" output="test_dem" |
| |
+----------------------------------------------------------------------------+
The point is that r.surf.contour is very slow when there are large gaps between the contour lines. So, either very dense contours, or a rougher resolution setting might help.
-
In addition, see also grasswiki.osgeo.org/wiki/Contour_lines_to_DEM– markusNCommented May 15, 2015 at 14:00
This are the settings of my vector layer:
g.region -p vector=CT_ISOBATAS_5m_ETRS89@froga
projection: 1 (UTM)
zone: 30
datum: etrs89
ellipsoid: grs80
north: 4818925.16988824
south: 4791485.92527958
west: 486512.5000001
east: 601274.08203047
nsres: 1.00011826
ewres: 1.00010965
rows: 27436
cols: 114749
cells: 3148253564
After executing the v.to.rast command, I get a raster, and it seems that is correct (pixels have the value of the elevation)
r.info isobatas5rast
+----------------------------------------------------------------------------+
| Map: isobatas5rast Date: Fri May 8 14:45:44 2015 |
| Mapset: froga Login of Creator: gepalgoi |
| Location: froga1 |
| DataBase: /gscratch/gepalgoi |
| Title: Categories ( isobatas5rast ) |
| Timestamp: none |
|----------------------------------------------------------------------------|
| |
| Type of Map: raster Number of Categories: 0 |
| Data Type: DCELL |
| Rows: 27436 |
| Columns: 114749 |
| Total Cells: 3148253564 |
| Projection: UTM (zone 30) |
| N: 4818925.16988824 S: 4791485.92527958 Res: 1.00011826 |
| E: 601274.08203047 W: 486512.5000001 Res: 1.00010965 |
| Range of data: min = -115 max = -5 |
| |
| Data Source: |
| Vector Map: CT_ISOBATAS_5m_ETRS89@froga |
| Original scale from vector map: 1:1 |
| |
| Data Description: |
| generated by v.to.rast |
| |
| Comments: |
| v.to.rast input="CT_ISOBATAS_5m_ETRS89@froga" layer="1" type="line" \ |
| output="isobatas5rast" use="attr" attribute_column="CONTOUR" value=1\ |
| memory=300 |
| |
+----------------------------------------------------------------------------+
Then, I interpolate the surface, but the resulting raster is exactly the same as the previos one. There is no an interpolate surface (although is hasn't give any error message):
r.surf.contour --overwrite input=isobatas5rast@froga output=mdt_isobata5
-
Note the non-square pixels - better to use the align flag -a of g.region with res=XX to set it properly (you need to do that at the very beginning of your workflow to not propagate odd resolution throughout your work).– markusNCommented May 15, 2015 at 13:59
-
I have used the flag -a as you said but it doesn't work. The resulting raster is again exactly the same as the one I have rasterised.– IrantzuCommented May 15, 2015 at 17:58
-
g.region -p -a raster=isobatas5rast@froga nsres=1 ewres=1 v.to.rast --overwrite input=CT_ISOBATAS_5m_ETRS89@froga type=line output=isobatas5rast use=attr attribute_column=CONTOUR– IrantzuCommented May 15, 2015 at 17:59
-
You region settings are giving you rasters of over 3 billion cells. Are you sure you need such high resolution? If I'm not mistaken, each raster layer will take 25GB of disk space.– MichaCommented May 15, 2015 at 18:50
-
Yes, I need a DEM of very high resolution of a wide area. I am using a computer with large capacities for it.– IrantzuCommented May 16, 2015 at 13:09