Geoscience Reference
In-Depth Information
FIGURE 5.4
A screenshot of GeoVISTA
Studio
in use, the left panel depicts (and saves) the workflow used
to create the visualisation - an example of visual programming. The right panel shows the resulting map, with
the colour choices supplied by
ColorBrewer
.
Janikas, 2006) and
GeoViz Toolkit
(Hardisty and Robinson, 2011). Slocum et al. (2008) provide a
useful overview of a variety of several research-based tools and rank their ease of use and utility.
Some excellent examples of the range of possibilities for map-based GeoViz are provided by Demaj
and Field (2013). Figure 5.4 shows an example session with the
GeoVISTA Studio
visualisation
system, where the workflow used to create the scene is also visualised (on the left).
The website at http://www.infovis-wiki.net/index.php?title=Toolkit_Links contains links to
many actively developed visualisation systems, some of which contain specific displays for GeoViz.
A very useful comparison of systems based on JavaScript (for use in a web browser) is available at
http://kraskniga.blogspot.co.at/2012/06/comparison-of-javascript-data.html.
5.3
DECOMPOSING THE VARIOUS ASPECTS OF A VISUALISATION
5.3.1 V
iSual
V
ariaBleS
As noted previously, visualisation systems provide a degree of control over the visual variables that
together define the appearance of the symbol sets and surfaces to be rendered. For example, the fol-
lowing visual variables may be separately configurable for a given symbol:
•
Position
: Usually in two or three dimensions.
•
Colour
: Often as three separate variables representing hue, saturation and intensity (or red,
green and blue); more sophisticated systems may offer perceptually graduated colour tools.
•
Size
: Usually as a scaling up or down of a symbol.
Most visualisation environments use a transformation function to assign quantities to each visual
variable, from values derived from a data attribute, that is, to visually encode the data. It is also
typical in many GeoViz systems to choose a small number of classes by which to group the data val-
ues (often via a binning classifier of some kind), so that a user can easily differentiate between the
resulting small number of different colours or sizes that are then used in the display. That is to say,
it is often helpful to keep a clearly perceptible visual difference between the settings used for the
visual variables (e.g. Brewer, 2003). Slocum et al. (2008: Chapter 13) provide a much fuller account
of the subtleties of visual classification.
