680 Lab 5 - Disjunctive / Indicator kriging

Objective: The lab's main objective is check out (ArcGIS Geostatistical Analyst's) facilities for disjunctive and indicator kriging. (You can try R or Surfer instead if you like.)

Data: Just one data set this time (like the W&O example):


  1. Look for indicator, disjunctive and probability kriging in the interface and software documentation. ("Hermite Polynomials" is not indexed but there is an option for "Gaussian kernel" in step 2 of the G-wizard that might be acomplishing the same thing.) Looking around in there should give you the "lay of the land". What are the apparent differences?
  2. Then, take the facilites for a spin using the threshold that W&O suggest, 25 mg/l. Try all three indicator, disjunctive and probability kriging. What outputs do they give you?
  3. Compare a prediction map result with a contoured (or reclassifyed) surface from ordinary kriging w.r.t. the critical value. Do you see any difference?
  4. Report what you've found.

Deliverable: A brief (maximum 5 pages) report addressing your assessment of the technique.

Due: in two weeks.