Visualizing Uncertainty

This chapter focuses on approaches for visualizing the uncertainty associated with thematic map displays. Three basic approaches for depicting uncertainty are introduced: (1) individual maps are shown for both an attribute and its uncertainty ( maps are adjacent ); (2) the attribute and its uncertainty are displayed on the same map (maps are coincident); and (3) dynamic displays allow users to either interact with or animate the attribute and its uncertainty. Two basic categories of visual variables for depicting uncertainty are covered: those that are intrinsic to the display and those that are extrinsic to the display (they involve adding objects to the display). Several approaches are covered that involve depicting uncertainty on choropleth maps, including the coefficient of variation, the use of confidence levels to determine class separability, and maximum likelihood estimation. Examples of applications outside the realm of choropleth maps include visualizing the uncertainty associated with the following: climate change, global water balance models, and forecasted hurricane paths. The chapter introduces both the potential and the limitations of using interactivity and animation to visualize uncertainty. The chapter concludes by considering the potential of using sound to represent data uncertainty, and thus offload uncertainty information from the visual channel to the aural channel.

Complex Systems and Decision Making · Computer Science