[R-sig-Geo] calculate semivariogram values and perform kriging BY HAND for a simple dataset

[Sarah Goslee]

Hi,

This sounds rather like homework.

On Tue, Mar 19, 2013 at 11:00 AM, ToNoY <tan.afreebird at gmail.com> wrote:
> Can anyone tell me how to calculate semivariogram values by hand for point C

And by hand? How is this an R question?

If you have an R question, by all means ask here. If you have a
homework question, please ask your professor or TA. If you have a
general stats question, StackOverflow is a good choice.

Sarah

> and D considering the distance between point A and B? say I have the
> following data points:
>
> point A: 60 (X-cord), 50 (Y-cord), & 1.2 (Z)
> point B: 30 (X-cord), 20 (Y-cord), & 2.1 (Z)
> point C: 20 (X-cord), 50 (Y-cord), & ? (Z)
> point D: 60 (X-cord), 50 (Y-cord), & ? (Z)
>
> Now, say, the best fit for model for the semivariogram is given by a
> spherical model as:
>
>  y(h) = C0 + C(1-e^(h/a))
>
> where, C0 = nugget effect, h = distance, and a = range. How can I do
> ordinary kriging by hand to estimate the Z value at point C and D. Also, how
> much weight (w) should be given to ZA and ZB while estimating ZC
> (ZC=w1*ZA+w2*ZB) or ZD (ZD=w1*ZA+w2*ZB)? Thanks in advance.
>
>


--
Sarah Goslee
http://www.functionaldiversity.org