AI-GEOSTATS: non-normal data distribution
- Dear Members,
I have been reading the archives on dealing with non-normal data
distributions in kriging with great interest, yet a complete understanding
of the normal-score transform (or rather the back-transform) eludes me
still, and I could not find a complete treatment of issues concerning the
I am kriging rainfall data for the West African Sahel (n=45) - a rather
sparse data set. The data are not normally distributed, and do not appear
to be log normal either. Stratified kriging is not a good idea because of
the small sample size. Taking the square root seems to fix the
problem. However, from what I gather from previous discussions,
back-transforming the kriged data is *not* a good idea. One way to
overcome this, as I understand, is to normal-score transform the raw data,
perform the kriging, and then back transform from normal scores. How is
this different than back-transforming from a log or square root
transformation : are there still problems with bias in the esimtates?
Dept. of Physical Geography
S-223 62 Lund
phone: 46 46 222 4262
fax: 46 46 222 4011
* To post a message to the list, send it to ai-geostats@...
* 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 majordomo@... 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