[STATSGRASS] Another problem with kriging (gstat)

[Jarek Jasiewicz]

Edzer J. Pebesma napisał(a):
> Jarek Jasiewicz wrote:
>
>> After kriging (in fact, with strongly chaotic data) I recived 
>> atrefacts (noisy elipses with values far from data range) in some 
>> places.
>>
>> For now I can't place image on www but if somebody is intersting, I 
>> can send image with addational patameters about kriging made
>
> This may be caused by several factors; without more knowledge about 
> your analysis we can't say anything. One possible reason is 
> observations that are very close in space (in combination with no 
> nugget effect).
>
> Best regards,
> -- 
> Edzer
no, nugget is very LARGE (data are chaotic, almost random), in my 
opinion data are not to do kriging, the aim is slighty different, I only 
interested in reasons of such element like in attachment
-------------- next part --------------
A non-text attachment was scrubbed...
Name: krig.jpg
Type: image/jpeg
Size: 53382 bytes
Desc: not available
Url : http://lists.osgeo.org/pipermail/grass-stats/attachments/20060904/039230ba/krig.jpg