Varying measurement error in Kriging predictions

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Varying measurement error in Kriging predictions

alexiadis.ant
Hello, my question is how can I implement known varying measurement error
in my
Kriging predictions?

I have two separate datasets: the first one is a 2-D dataset that includes
the observations and
the second is the a 2-D dataset that includes the measurement error in each
specific location.
The measurement error dataset follows a structure, it's not random (which
can be characterized via the use of the experimental variogram).

I have searched a lot online on how to incorporate the measurement
uncertainty in
Kriging predictions but it seems to still be an open question in the forums
(a paper solving this issue has been developed by William F. Christensen
titled as: Filtered Kriging for Spatial Data with Heterogeneous Measurement
Error Variances). I have tried to use gstat and incorporate the variances
using the weights functionality but after I do the kriging and visualize
the predicted variance field, even though it qualitatively resembles the
defined one, the values of the variances are magnitudes lower than the ones
proposed.

Does anyone have an idea on how to solve it, or aware of some
software-package that
has already implemented this functionality? (It's quite tough to understand
the stated paper already, rather having to program
its contents) .


Thank you.

Regards,
Antonios.

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Varying measurement error in Kriging predictions

edzer
Please read p 74, "Kriging data with known measurement errors", of
http://gstat.org/gstat.pdf . It refers to a method published by Delhomme
(1978). The software was written long before the Christensen paper you
mention, to which I don't have access.


On 11/6/18 3:55 PM, Antonis Alexiadis wrote:

> Hello, my question is how can I implement known varying measurement error
> in my
> Kriging predictions?
>
> I have two separate datasets: the first one is a 2-D dataset that includes
> the observations and
> the second is the a 2-D dataset that includes the measurement error in each
> specific location.
> The measurement error dataset follows a structure, it's not random (which
> can be characterized via the use of the experimental variogram).
>
> I have searched a lot online on how to incorporate the measurement
> uncertainty in
> Kriging predictions but it seems to still be an open question in the forums
> (a paper solving this issue has been developed by William F. Christensen
> titled as: Filtered Kriging for Spatial Data with Heterogeneous Measurement
> Error Variances). I have tried to use gstat and incorporate the variances
> using the weights functionality but after I do the kriging and visualize
> the predicted variance field, even though it qualitatively resembles the
> defined one, the values of the variances are magnitudes lower than the ones
> proposed.
>
> Does anyone have an idea on how to solve it, or aware of some
> software-package that
> has already implemented this functionality? (It's quite tough to understand
> the stated paper already, rather having to program
> its contents) .
>
>
> Thank you.
>
> Regards,
> Antonios.
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>
--
Edzer Pebesma
Institute for Geoinformatics
Heisenbergstrasse 2, 48151 Muenster, Germany
Phone: +49 251 8333081

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo

pEpkey.asc (2K) Download Attachment