[R-sig-Geo] Risk mapping: Parametric vs Non-Parametric Conditional Simulation approaches?
zua3 at cornell.edu
Fri Dec 18 17:00:51 CET 2009
I like to compare areas not exceeding of particular threshold values
between parametric and non-parametric approaches of 1000 realizations.
For non-parametric approach, I did IK for 1000 realizations (conditional
simulation) and calculated the area not exceeding (0.5 probability) of
the threshold value for each of 1000 realizations. Now, I want to
compare these areas with areas calculated from parametric approach. For
this, I did OK conditional simulation of 1000 realizations. My
question is: Is it possible to create probability maps not exceeding a
threshold value of each of 1000 realizations. I know, I need variance
of each prediction location of each realization to get the probability.
Please consider my limited knowledge about theory of geostatistics.
Any suggestion will be appreciated.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 281 bytes
Desc: not available
More information about the R-sig-Geo