Isobel, I've used universal kriging a lot in the past, but in this case I want to use a global trend and a small search neighbourhood (since I have nearly 16000 observations), so I would be estimating the variogram from one set of residuals and kriging with another. That's why I thought of using SGS based on simple kriging with locally-varying means defined by the fitted polynomial, as it seems consistent (but there's still a problem with not accounting for variation in the polynomial on which the locally varying mean is based). However, in practice I suspect that UK would do the trick if this minor(?) inconsistency is overlooked... I could estimate the variogram of residuals from locally detrended data, but that could become complicated. I suppose IRF-k kriging is yet another option. Perhaps comparison of several approaches is the most sensible route, but I think after your advice I'll start with UK-based SGS and go from there.
Chris -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Isobel Clark Sent: 06 August 2003 21:17 To: Chris Lloyd Cc: [EMAIL PROTECTED] Subject: Re: AI-GEOSTATS: Simulation and trends Chris Could I be incredibly obvious and suggest that, if you use Universal Kriging, the trend is fitted and simulated automatically with SGS. This is one of the major advantages of SGS over approaches like Turning Bands or Monte-Carlo -- if you can krige it, you can simulate it. There is a lot of evidence in the literature, dating back to the early '80s that kriging residuals and adding back the trend gives you pretty much the same estimated surface as Universal Kriging. However, what it doesn't do is give you the right standard error since it doesn't allow for the trend fitting error. So I would hazard a guess that simulations done this way would underestimate the 'true' variability. Isobel {Clark} http://drisobelclark.ontheweb.com PS: could I take this opportunity to remind anyone interested that the IAMG 2003 is rapidly approaching. If you haven't registered yet, sort yourself out at http://www.iamg2003.com or follow the links from our page at http://ecosse.ontheweb.com/whatsnew.htm ________________________________________________________________________ Want to chat instantly with your online friends? Get the FREE Yahoo! Messenger http://uk.messenger.yahoo.com/ -- * To post a message to the list, send it to [EMAIL PROTECTED] * As a general service to the users, please remember to post a summary of any useful responses to your questions. * To unsubscribe, send an email to [EMAIL PROTECTED] with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list * Support to the list is provided at http://www.ai-geostats.org -- * To post a message to the list, send it to [EMAIL PROTECTED] * As a general service to the users, please remember to post a summary of any useful responses to your questions. * To unsubscribe, send an email to [EMAIL PROTECTED] with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list * Support to the list is provided at http://www.ai-geostats.org
