Re: A fundamental question on semivariogram

[Edzer J. Pebesma]

Kanishka Perera wrote:
> 
> Dear gstat users:
> 
> Could someone differentiate between semivariance and covariance? My
> understanding is semivariance is determined from large number of data
> couples whereas covariance is just for one such couple. Please let me
> know.
> 
Not true. 

Semivariance: 0.5 E(Z(x)-Z(x+h))^2
Covariance:       E(Z(x)-m)(Z(x+h)-m)

Please read any book on geostatistics for further information. 
It might even be in the gstat manual.
--
Edzer