[GRASSLIST:764] Re: Working on pollution data with GRASS... interpolations

[Stuart Edwards]

I ran into some problems recently using GRASS to interpret  
groundwater pollution data -- v.surf.rst tends produce regions of  
negative pollution (overshoots) around the edge of contaminated  
zones.  This is because data sets for pollution studies are often a  
bit 'extreme'  with maybe three orders of magnitude difference of  
attribute value between some adjacent points.  My solution was two  
step -- first crank up the tension (I used values of up to 1500) to  
pull the surface as tight as possible and eliminate the negatives.   
If that didn't work, I log transformed the data to 'calm it down' for  
graphic presentation purpose (that's probably how you would contour  
it by hand anyway).  v.surf.rst does allow you to account for  
anisotropy (an elongated plume for example) with 'theta' and  
'scalex'.  But of course you have to know roughly what the anisotropy  
is first in order to use this feature.

I'd also like to know if there's a better way to deal with this problem.

Stuart
On Apr 19, 2006, at 4:57 AM, Luca Penasa wrote:

> I have a lot of points with a value of air-pollution for every  
> point (these values are lichen's biodiversity index). Now i need to  
> create a map with an interpolation procedure... iv tried using  
> v.surf.rst, but i know the model used for interpolation works fine  
> with terrain surfaces... does it work well for these pollution  
> values too? Is there another interpolation model (maybe a model  
> that works fine with pollution datas) implemented like a grass  
> module? If not, what parameters of tension and smooth have i to use  
> with v.surf.rst? Is there anybody else working with GRASS on  
> pullution datas?
>
> Thanks a lot....
>
> Luca
> --
> Email.it, the professional e-mail, gratis per te: http:// 
> www.email.it/f
> Sponsor:
> Corso multimediale sul Controllo di Gestione: impara facilmente  
> come ridurre i costi e aumentare gli utili della tua attivita'
> Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=5055&d=19-4