Geoscience Reference
In-Depth Information
Depending on the sophistication of the system being used, further control over appearance may
be possible via additional visual variables such as the following:
•
Transparency
: As a single value, usually referred to as alpha (α), that effectively allows the
user to see through the symbols and overlays present (Roth et al., 2010)
•
Material properties
: To describe the light reflectance and absorption characteristics of the
objects in a scene
•
Animation scripts
: To govern when a symbol appears and how its visual variables may
change over time
In many visualisation systems, it is now possible to extend the visual library by adding in new sym-
bols and other geometries that may suit a particular purpose or audience. In addition, many SciViz
environments provide further control over the entire scene, including the position from which the
scene is observed (
viewpoint
) and the location and characteristics of light sources to control how
the scene is illuminated. These scene properties can be used as a basis for interaction. For example,
a fly-through or walk-through can be simulated by incrementally changing the viewpoint. Many
3D solid-earth visualisation environments (
block models
) also support the slicing and dissection of
solid
structures, such as geological formations, or in climate circulation models to reveal the interior
structure of solid objects that might otherwise be hidden. Scene lighting can be used to highlight or
focus attention on
interesting
trends in the data.
Compared to SciViz environments, the degree of control over visual appearance available in
many GIS, web mapping and InfoViz environments is rather limited, since the latter are often
restricted to a small number of pre-defined symbols, with few visual variables available to the
user. Instead of concentrating on the quality and realism of rendered objects, these systems focus
more on supporting abstract graph displays that translate unfamiliar information into a more
common and uniform visual format, such as a map or a scatterplot. In doing so, the complexity
of each individually rendered symbol is greatly reduced, in exchange for the clarity of a well-
designed and more consistent layout. Choosing between these two approaches can be compli-
cated; SciViz systems provide much more visual flexibility and so offer the possibility of creating
a single scene that encodes many data attributes so that subtle relationships may become appar-
ent. But using this flexibility to advantage is a challenge - it is easy to present too much visual
information to a user, and the more complex the display, the harder it becomes to
read
. On the
other hand, presenting a large number of variables in scatterplots and maps involves dividing
up data attributes among several displays, which may not help in recognising subtle multivari-
ate associations (Gahegan, 1998; Griffin and Robinson, 2010). The circumstances in which one
approach might be better than the other are not well understood at this time and would make an
excellent research topic.
5.3.2 V
iSual
d
iSPlayS
Some of the visualisation displays commonly used in GeoViz are choropleth mapping (including
bivariate mapping), scatterplots, cartograms, parallel coordinate plots and star maps; coordinating
multiple displays together is also a common feature. An example session in
GeoViz Toolkit
show-
ing some of these tools in coordinated use appears in Figure 5.5. Ward et al. (2010) provide a thor-
ough description of many of the more widely used visualisation displays; Slocum et al. (2008) and
Gahegan (2008) provide summaries of displays commonly used in GeoViz.
5.3.2.1 More Complex Visual Metaphors
Since GeoViz currently most closely follows InfoViz, the symbols and displays used in GeoViz
tend to be simple and abstract and employ consistent graphing metaphors and strong perceptual