[R-sig-Geo] Universal Block Kriging covariate definition

[Bruin, Sytze de]

On 14/01/16 12:35, Bruin, Sytze de wrote:
> I believe the residual variogram should then be computed using the covariate data at block support.
>
> Sytze de Bruin
> Wageningen University
> Laboratory of Geo-Information Science and Remote Sensing

why?
--
Edzer Pebesma
Institute for Geoinformatics  (ifgi),  University of Münster
Heisenbergstraße 2, 48149 Münster, Germany; +49 251 83 33081
Journal of Statistical Software:   http://www.jstatsoft.org/
Computers & Geosciences:   http://elsevier.com/locate/cageo/
Spatial Statistics Society http://www.spatialstatistics.info<http://www.spatialstatistics.info/>

That believe was a mistake. The residual variogram computed from the covariate data at point support should be used. It doesn't matter that the covariate data are first aggregated over the blocks. The order of the operations is unimportant for the kriging predictions while the residual variogram at point support is the correct model for the computations of point-block and point-point covariances (or semivariances).  Sorry for any confusion caused by my previous post.

Sytze de Bruin
Wageningen University
Laboratory of Geo-Information Science and Remote Sensing


	[[alternative HTML version deleted]]