[STATSGRASS] RE: [GRASS-user] kriging

Alexandre Leroux alexandre.leroux at ec.gc.ca
Mon Sep 25 14:54:18 EDT 2006

Hi Maris,

Using interpolation breaklines (in the actual discussion, 'faults') does 
not provide the same results as patching two partly-masked 
interpolations. Do no forget breaklines can be non-strait complex 
curves, that's why is does not give you the same results as using masks.

Say you have:

where o = land, your breakline, and x = water. 1 and 2 are two 'x' water 
points. When masking land 'o' and interpolating, 1 and 2 will influence 
each side of the land arm (the 'o' land that goes north) while they 
shouldn't since there is land between them. I do not know the exact 
algorythm, but this can't happen when using breaklines. I hope this 
example I was clear enough to convince you the result of masks and 
breaklines is not the same... :-)



Māris Nartišs wrote:
> Hi Maciek,
> I will disagree with You - as 
> http://surfit.sourceforge.net/surfit/fault__approx_8tcl-example.html says:
> "The main idea of fault curves is that parts of resulting surface between 
> fault curve are independent of each other." 
> It's like masking areas between fault lines, interpolating them seperatly and 
> patching together. Only difference - in GRASS you must do it manualy for each 
> fault-enclosed area.
> Maris.
> On Monday 25 September 2006 15:35, Maciej Sieczka wrote:
>> Markus,
>> That's not the same. I don't you can achieve this:
>> http://surfit.sourceforge.net/surfit/fault_8tcl-example.html
>> http://surfit.sourceforge.net/surfit/fault__approx_8tcl-example.html
>> using Grass. In Grass you can't use lines as interpolation faults. The
>> method you refer to uses area masking, and it can't reproduce the
>> Surfit examples.
>> Maciek
Alexandre Leroux, M.Sc., Ing.
Environnement Canada / Environment Canada
Centre météorologique canadien / Canadian Meteorological Centre
Division de la réponse aux urgences environnementales /
Environmental Emergency Response Division
alexandre.leroux at ec.gc.ca

More information about the grass-user mailing list