A Qualitative Approach to Understanding the Role of Geographic Information Uncertainty during Decision Making
Roth, Robert E., Cartography and Geographic Information Science
"We actually made a map of the country on the scale of a mile to a mile!" said Mein Herr.
"Have you used it much ?" I enquired.
"It has never been spread out, yet," said Mein Herr.
"The farmers objected: they said it would cover the whole country and shut out the sunlight/So we now use the country itself as its own map, and I assure you it does nearly as well." (Carroll 1893, 1982, p. 726.)
The passage from Lewis Carroll, well known among cartographers and GIScientists, illustrates a central concern with the representation of geographic information. In order to be understandable and usable, a map must abstract reality, removing unnecessary or less important details while maintaining, and therefore accentuating, features of interest. As noted in the Carroll passage, a map that does not abstract carries little semiotic advantage and is no more useful than interacting with the world itself. By abstracting reality, however, information is removed that may be requisite for a clear and comprehensive understanding of geographic phenomena or processes. Tasks employing the representation or decisions informed by the representation can no longer be accomplished with complete certainty. This is the cartographic problematic: in order to create an abstraction of reality that makes complex geographic information understandable and usable, uncertainty is introduced into the representation (and into the knowledge constructed from the representation) as a necessary compromise.
Because uncertainty is inherent to all geographic information, and therefore decisions based on this information, research in GIScience should focus on better ways to manage and use uncertainty during decision-making rather than attempting to purge it from all geographic information (Couclelis 2003; Deitrick and Edsall 2008). There are at least three challenges to coping with geographic information uncertainty: (1) determining the current involvement of uncertainty at different stages in the geographic information life cycle (i.e., is it collected and represented for use in a decision?); (2) identifying the many forms that uncertainty can take in this process (i.e., do decision-makers need to consider multiple categories of uncertainty and, if so, what are they?); and (3) understanding the influence these forms have on the use of geographic information (i.e., are all uncertainties considered as equal when making a decision?).
Unfortunately there are currently few empirically based solutions for these three challenges. Although there is a strong emphasis in the GIScience literature on developing new methods for collecting and representing uncertainty, information there is rarely a follow-up component to these studies investigating if and how the suggested techniques are applied in practice (challenge #1). However, many scholars have attempted to partition the broad concept of uncertainty and define its components (challenge #2), but with little agreement. A nine-category typology of uncertainty first offered by Thomson et al. (2005) then restated and elaborated upon by MacEachren et al. (2005) is examined in this research because it is specific to uncertainties that influence information analysts. Finally, despite the large amount of writing on GIScience uncertainty typologies, authors have only speculated on the relative influence of these categories on the decision-making process (challenge #3), instead examining decision-making influence with a broad definition of uncertainty.
This research addresses the aforementioned three challenges using the case study domain of floodplain mapping. Floodplain maps are a formalized cartographic tool used for the evaluation of flood liabilities. Urban planners use these maps to assess flood hazards before developing an area. Similarly, emergency agencies and private insurance firms use floodplain maps to evaluate the vulnerability of features that are already in the built landscape, whose construction cannot now be avoided. …