[GRASS-user] Problem with v.vol.rst and 3d-interpolation
butzeck at web.de
Sat May 19 10:16:39 EDT 2007
I had a (huge) problem with the result of the v.vol.rst interpolation.
I had some contaminant samples (about 300 values for this sample area) from measuring probes, the exact coordinates and the sea level (up to 18 for each measuring probes - at different levels).
First I create a raster map with v.in.ascii from a csv-file with all values and coordinates.
# v.vol.rst input=raster_map_with_values wcolumn=contaminant dmin=0.49 segmax=700 npmin=100 elev=test_area tension=10 smooth=0.025
I used for the region top=29, bottom=0, nsres3=5 and tbres=0.5. The level from the probes is located between 1 and 28 meter about sea level. For example there could be 5 measurements at one measuring probes (e.g. 20.33, 19.83, 19.33, 15.33 about sea level) up to 18 at one measuring probes (26.11, 25.11, 24.11, 23.11, 17.11.........) Just to show that there are no fixed depths for the measurements! Maybe its important.!?
After this I used "r3.mapcalc" to eliminate all values less than null and cut the region with a 3d-buffer I created before with "r.digit" and "r.to.rast3".
At last I used "r3.out.vtk" to create a vtk-file for visualization with paraview or another 3d-program.
Now my problem!
Everything works and looks great on the first moment, but if I compare the output data with the original data from the csv-file there are big differences. It look likes v.vol.rst interpolates with the NULL-values/ No data, even if no (real) value is close to these NULL values.
For example on one measuring probes 10 probes had been taken between 13 and 23m about sea level. Above 23m and below 13m there are no data (because of no ground water above 23m and a clay bed below 12m about sea level!). At 23m there is a concentration from 0.01mg/L, at 13m 12.45mg/L (the maximum of the hole area). --------> Now I look at the vtk-file and find out, that this maximum is located at 0m (7m below the measuring point) about sea level! At 23m (the last measuring point) there is a concentration of just 5mg/L., the direction of the concentration is right but not the location.
Can anyone explain me how I can fix this problem. I don't think that it's a problem of tension and smooth parameters because I tested this since weeks.... If I used the module "v.surf.rst" you can choose an anisotropy factor ('theta' and 'scalex'), is it possible to use this also at v.vol.rst. For example that the interpolation between the vertical values have a higher weighing than the horizontal values or the other way around.
Or is this a bug? - I heard from Soeren, who designed the "r3.out.vtk"-module, that there wasn't much time to test this software?
I hope you understand my problem/questions (and my sometimes awful English;)
Regards from Potsdam.
SMS schreiben mit WEB.DE FreeMail - einfach, schnell und
kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192
More information about the grass-user