Hi Daniel
First of all consider, if you are working with a layered 3D
architecture, that a full 3D approach could be not reasonable, but it could
be easier to subdivide your data set in horizontal strata.
Are you interpolating lithologies?
In regard to gslib the related book is very exaustive, but you can
also calculate a 3d variogam with gslib and then compare it with sage.
Finally, there was also the sgem project (based in gslib) that could
be good for you.
Hope this is useful.
Bye
Sebastiano
At 11.37 18/11/10, Daniel Nienstedt wrote:
dear users,
currently i'm working on my master thesis. the aim is to reconstruct
hydrogeologic parameters via sisim in gslib on a contaminated site.
afterwards a simulation with feflow will be done. my input consists
of 35 borehole profiles with a total amount of 1200 datapoints
consisting of 4 indicators. these points are set 10 cm before and
after a layer is changing and in the middle of each layer (more
points if the interval between points will be over 1 m).
i'm relatively new to geostatistics and i'm experiencing some
problems doing the indicator variograms. currently i use sage2001
(<http://www.isaaks.com>http://www.isaaks.com). the output are 3
rotations, 3 azimuths and 3 dips for the anisotropy. i'm not exactly
sure how to put these into the sisim routine in gslib. the azimuths
are the angles 1-3 for each indicatorvariogram but the rotations? and dips?
for the search ellipsoid, which are the ranges i need to enter - the
biggest and smallest of all variogram ranges?
another problem i got is that my simulation results matches my pdf
input only by ~60 %. will these results get better with exacter
variogram inputs?
thanks for any feedback and advices,
daniel
