Dear list members
I would like to share a short consideration.
Some day ago I was doing some prediction exercises by means of
ordinary kriging with linear variograms.
And doing this exercise I realized (something that probably I should
have realized many years ago!) that
things change when using linear variogram with or without a nugget effect.
Simply:
1) without nugget effect kriging weights are not influenced by the
slope of the linear variogram (is, in this
case, kriging equivalent to bilinear interpolations as one of the
list members said some time ago?)
2) if the nugget is defined, kriging weights are function of the
slope of the linear variogram.
This last point, from my perspective, should be taken into strong
considerations from those people
who suggest defaults parameter for interpolation with kriging with
linear variograms.
My opinion is that if you need quick interpolation for a first check
of data you can use a linear variogram
without nugget (or use a natural neighbor method). But once you
define a nugget effect (whose
choice should be justified) you have at least to fit the correct slope.
I would like to know your opinion about that.
Sebastiano T.
+
+ To post a message to the list, send it to ai-geostats@jrc.it
+ To unsubscribe, send email to majordomo@ jrc.it with no subject and "unsubscribe
ai-geostats" in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the
list
+ As a general service to list users, please remember to post a summary of any
useful responses to your questions.
+ Support to the forum can be found at http://www.ai-geostats.org/
