This talk presents decision analysis methodology for decisions based on data from geographic information systems. The consequences of a decision alternative are modeled as distributions of outcomes across a geographic region. We discuss conditions that may conform with the decision maker’s preferences over a specified set of alternatives; then we present specific forms for value or utility functions that are implied by these conditions. Decisions in which there is certainty about the consequences resulting from each alternative are considered first; then probabilistic uncertainty about the consequences is included as an extension. Two hypothetical urban planning decisions are modeled: one on water use and temperature reduction in regional urban development and one on fire coverage across a city.

One of the remaining challenges is the need for a cardinal preference function that incorporates the spatial nature of the outcomes. We explore preference conditions that will yield the existence of spatial measurable value and utility functions. Such measurable preference functions allow simpler assessment procedures and a strength of preference interpretation of the results. We present a simple example on household freshwater usage across regions.

The talk is based upon:
L. Robin Keller, Jay Simon, “Preference Functions for Spatial Risk Analysis”, Risk Analysis, (forthcoming in print).Version of Record online: 7 SEP 2017 | DOI: 10.1111/risa.12892, http://onlinelibrary.wiley.com/doi/10.1111/risa.12892/full, online in early view.
Jay Simon, Craig W. Kirkwood, L. Robin Keller, “Decision Analysis with Geographically Varying Outcomes: Preference Models and Illustrative Applications”, Operations Research, 62(1), Jan.-Feb. 2014, 182-194. Printed paper: opre 2013 1217-2au_copy; Supplement Simon et al. 2014 supplement. http://pubsonline.informs.org/doi/abs/10.1287/opre.2013.1217