- Hello everybody,

I have several (maybe too simple) questions on geostatistics:

1. I would like to know where I can find any kind of detailed interpretation

guide for the nugget effect. I know how to describe the nugget effect in my

words, but I don�t know how do it in scientific english.

2. I do have count data from a regular sampling grid which is in my eyes to

sparse. Can I refer to the nugget effect to demonstrate this ?

3. I would like to know more about the sensitivity of kriging for non-normal

distributed datasets. I have distributions from my above mentioned count

data which have more or less 80% zeros. Is this problematic when working

with kriging (and why???).

4. Is ist really practicable to transform the above mentioned datasets with

many zeros in it into a lognormal distribution (log = err !!!) ?

Hopefully anybody out there can help me understanding this better...

Peter Pinn

_________________________________________________________________

The new MSN 8: smart spam protection and 2 months FREE*

http://join.msn.com/?page=features/junkmail

--

* 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 > Hello everybody,

nugget effect in my words, but I don´t know how do it in scientific

>

> I have several (maybe too simple) questions on geostatistics:

>

> 1. I would like to know where I can find any kind of detailed

> interpretation guide for the nugget effect. I know how to describe the

english.>

mentioned count data which have more or less 80% zeros. Is this

> 2. I do have count data from a regular sampling grid which is in my eyes

> to sparse. Can I refer to the nugget effect to demonstrate this ?

>

> 3. I would like to know more about the sensitivity of kriging for

> non-normal distributed datasets. I have distributions from my above

problematic when working with kriging (and why???).

From an empirical point of view: i often meet data with lots of zeros when

i do gs from samples taken by ships during marine ecology surveys. I have

done the analysis (both the structural part and kriging) with and without

the zeros without much of a difference.

> 4. Is ist really practicable to transform the above mentioned datasets

There is a generalization of the lognormal distribution which accounts for

> with many zeros in it into a lognormal distribution (log = err !!!) ?

the presence of zeros. It is called the Delta distribution. IIRC, the

maximum likelihood estimate of the lognormal mean (l) is related to the

mle of the delta mean (d) by the simple relation d=l*(m/n) where m is the

number of non zero observations and n is the total number of observations.

Here you have two useful references:

Aitchison J. 1955. On the distribution of positive random variables having

a discrete probability mass at the origin. J. American Statistical

Association 50:901. (Aitchison and Brown wrote the first basic book on the

lognormal).

Pennington M. 1983. Efficient estimators of abundance, for fish and

plankton surveys. Biometrics 39:281. (this work deals specifically with

the problem of large number of zeros and the lognormal pdf).

Good luck.

Ruben

--

* 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