Companion Notebook: Cokriging with Unequal Supports

Resource models often draw on multiple data sources with different sample quality or support, such as different sampling vintages or drilling types. Cokriging with unequally sampled data integrates these data sets while honoring their spatial relationships.

The Geostatistics Lesson by Luis Davila and Clayton Deutsch introduces the theory. Luis Davila and Ryan Barnett developed a Python companion notebook that applies a full cokriging workflow to blasthole and drill hole data and compares the results with ordinary kriging and cokriging with an intrinsic model of coregionalization, showing which workflow simplifications should and should not be made.