Pierre Goovaerts on
Indicator Kriging
From an email to AI-Geostats:
Date: Fri, 4 Dec 1998 11:03:10 -0500 (EST)
From: Pierre Goovaerts
- discretize the range of variation of your data using
a given number of thresholds, say 5: the first threshold
would be 0% (which is close to the median of your sample
distribution) and 4 other thresholds corresponding to the
0.6, 0.7, 0.8 and 0.9 quantiles of your distribution.
- for each threshold, code each observation into an indicator value
which is zero if the measured percentage is larger than the threshold
and one otherwise.
- Compute and model the 5 corresponding indicator semivariograms,
that is the semivariograms of indicator values.
- Use indicator kriging to interpolate the probabilities
to be no greater than each of the 5 thresholds at the nodes
of your interpolation grid.
- At each location, you can now model the conditional cumulative
distribution function which provides you with the probability
that the unknown percentage value is no greater than any
given threshold. You could use the mean of that distribution
as your estimate and the variance as a measure of uncertainty.
The method is described at length in the geostat literature,
as well as in Chapter VII of my recent book. All the computations
can be carried out using the public domain software Gslib,
you can download the source code from Deutsch's
Website.
[Here is] a review paper that I recently
wrote for soil scientists in which I present (or try to present!)
in layman terms a broad overview of main applications of geostat.
Goovaerts, P. 1998.
Geostatistical tools for characterizing the spatial variability of
microbiological and physico-chemical soil properties.
Biology and Fertility of Soils, 27(4): 315-334.
Website maintained by Andy
Long. Comments appreciated.
aelon@sph.umich.edu