Dear list, 

I received a few more replies from Jon Skoien, Gerald van den Boogaart,
Denis Marcotte and Yücel Tandoğdu,  to my question on the correlation
between kriging residuals and my input data. So many thanks to all for
the replies that all go in the same direction (please send your replies
to the list, not to me as the mail archive is supposed to contain all
our discussions!). 

Still, part of the question remained unanswered. We all agree that in
the case of a pure nugget effect, the predictions would be equal to the
mean of the observations leading so to as many overestimations as
underestimations (hence the correlation between residuals and
observations = 1 or -1 depending on the definition of the residuals).
The real issue was however the fact that the investigated dataset was
not considered to present a pure nugget efffect!

The problem was probably better summarized by Jon: "If there is a
spatial correlation between your observations, the correlation between
observations and residuals will generally decrease with increasing
spatial correlation."  In addition to my problem of considering whether
there was or wasn't a spatial correlation in the data, the interesting
part of the problem is thus the "will generally decrease". 

What are thus the requirements for not observing this decrease and how
(linearly?) are both correlations (spatial correlation of observed
values versus the correlation between observations and residuals)
linked? 

Many thanks again for your replies and sorry for my late feedback (I'm
currently moving office)

Best regards,

Gregoire

__________________________________________
Gregoire Dubois (Ph.D.)

European Commission (EC)
Joint Research Centre (DG JRC)
WWW: http://rem.jrc.cec.eu.int/

"The views expressed are purely those of the writer and may not in any
circumstances be regarded as stating an official position of the
European Commission."




+
+ To post a message to the list, send it to [email protected]
+ 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/

Reply via email to