Non-stationary kriging factors?

[RyanConway91]

I would like to develop a k field with anisotropy that follows a meandering buried alluvial channel, and preferably calibrate the degree of and anisotropy and the bearing. Is there any way to do non-stationary kriging in pyemu/PEST? Or would the approach be to be krige the channel in little chunks and combine the data together in the end?

[rhugman]

Checkout 'pypestutils' (integratable with pyemu; see examples in the repo) and/or J.Doherty's 'alluvsim' utility (pesthomepage.org). Example applications of the latter in one of the first GMDSI worked examples.