Abstract
This paper explores in detail a new and effective metaphor for visualising choropleth map uncertainty. The “level-of-detail” metaphor has been shown here to communicate attribute uncertainty, but also spatial uncertainty as a secondary expression. The metaphor is delivered to the map viewer via the regular tessellated output of the Hexagonal or Rhombus (HoR) quadtree spatial data structure, as a semi-transparent map layer that lies on top of the choropleth (termed the trustree when used in this manner). For testing, multiple images were created with differing resolution levels of output from the trustree and superimposed on a New Zealand 2001 census choropleth map of Dunedin City. An Internet survey was designed and run, to reveal the visual metaphors that the trustree communicates uncertainty through. The choice of metaphor offered was (1) a level of detail (or resolution) metaphor, where less detail (i.e. coarser resolution cells) represents more uncertainty (i.e. uncertainty is sketchy), or (2) a metaphor of clutter, where the data structure output can be sufficiently dense so as to cover spatial information, in effect hiding uncertain areas (i.e. uncertainty is a barrier). In this case the finer resolution cells indicate more uncertainty. Also, the survey aimed to determine a usabletrustree tessellation resolution level to express uncertainty information. The results showed the trustree tessellation was more effective when representing a metaphor of detail and that attribute and spatial uncertainty can be effectively expressed, depending on the tessellation level used.