Geography Reference
In-Depth Information
waterline. Because a decrease in future water levels for the Great Lakes is possible,
the conceptualization of the Lake Level Viewer as a
visualization is
inappropriate. Therefore, different symbolization is needed for newly exposed
land versus newly inundated land.
Many of the visualization tools encode the flood depth (a numerical variable)
using an additional visual variable. Six (24 %) tools use color hue to represent water
depth, two (8 %) use transparency, and one (4 %) tool uses a combination of color
value and color saturation. Drawing from semiotics, the use of value + saturation
and transparency are predicted to be effective solutions for representing a numerical
variable, while the use of color hue is not (MacEachren 1995 ).
Finally, two unique representation solutions are worth noting. First, the Surging
Seas visualization loads basemap layers of different detail for areas within the flood
extent (satellite imagery) versus beyond the flood extent (a generalized vector map)
(Fig. 1b ). This solution allows for impacted areas to be viewed in more detail
without a flood symbol obfuscating the area of interest, a limitation of other tools.
Unfortunately, this solution may not be as useful for representing a declining water
level in the Great Lakes, as imagery is not available for areas currently inundated.
Second, the Sea Level Trends tool makes use of the visual variable orientation to
represent water level change at a small cartographic scale, with the amount of
change represented redundantly using color hue and size (Fig. 1c ). The Lake Level
Viewer may benefit from such an
flood
'
'
(Shneiderman 1996 ), as land flooding
or exposure on the Great Lakes is confined to a relatively small area along the coast
that is viewable at large cartographic scales only.
overview
'
'
Representation of the Certainty of the Waterline/Flood Extent
Prediction
The term uncertainty describes any cause for a mismatch between reality and the
user
s understanding of reality (Roth 2009 ) and may be considered as a series of
filters within the reality
'
!
variable-definition
!
data-collection
!
information-
assembly
knowledge-construction pipeline (Longley et al. 2005 ). Effective
uncertainty representation is essential to the design of visualizations that support
decision making (Agumya and Hunter 2002 ). In GIScience, information un-
certainty is considered multifaceted, exhibiting at least three components: (1) accu-
racy/error , or the correctness of a measurement or estimate, (2) precision/
resolution , or the exactness of a measurement or estimate, and (3) trustworthiness ,
or the confidence that the user has in the information (MacEachren et al. 2012 ).
Trustworthiness typically is conceptualized as a
!
category that includes
aspects of the currency, completeness, internal consistency, credibility, subjecti-
vity, interrelatedness, and lineage of the represented information (MacEachren
et al. 2005 ).
'
catch-all
'
Search WWH ::




Custom Search