Geography Reference
In-Depth Information
hours of active periods, followed by much slower movements of higher turning
angles during the middle hours (when the possum was feeding in gum trees), then a
return to short intervals of rapid and directed 'commuting' movements at the end of
periods of activity as the animal returned to its favoured denning site.
Each segment of the timelines is coded with a single colour in Fig.
5
a and b.
With transparency several colours can be associated with the same segment of a
timeline to explore multiple variables (Fig.
5
c).
Figure
5
c depicts the same information expressed in Fig.
5
a and b in a single
image by overlaying wider semi-transparent colour 'masks' of habitat on top of
behaviour timelines. While this promotes closer examination of the relationships
between habitat and behaviour, the rich information displayed may challenge
human abilities in visual information-processing. Interactively switching on and
off the 'masks' may help reduce the challenge.
Figure
5
d shows an aggregated pattern of the timelines, i.e. flows. We can see
that a 'co-incidence in space and time pattern'
3
(Andrienko and Andrienko
2007
;
Dodge et al.
2008
) occurs mostly at 4-8 h after sunset to the north-west of her den
(see green to dark blue colours). By comparing Fig.
5
a and d, we see that the
concentration of locations in the gum trees depicted in Fig.
5
a reflects the main co-
incidence patterns seen in Fig.
5
d, indicating that the gum trees might be a key
driver of the possum's co-incidence patterns.
Timelines offer an elegant and intuitive means of expressing patterns of move-
ment, by bringing space, time, activity and geographic context together as a whole.
However, timelines only can effectively represent a small amount of data in a single
diagram, and they are subject to inherent issues associated with perspective, such as
difficulties in relating timelines to exact locations in space and time. Additional
approaches providing a compact and flexible means of exploring rich and complex
patterns of movement and their geographic context are needed. It is also necessary to
develop a geovisualisation toolset which coordinates multiple representations,
derived from various approaches and scales, that complement and augment each
other, and provide further insights (Zhao et al.
2006
). Our next section applies the
innovative geovisual analytics method—the Ringmap to movements of possums.
3.3 Visualising Movement Patterns of Possums
with Modified Ringmaps
Ringmaps were originally developed to visualise 24-h cyclic patterns of human
movement and activities (Zhao et al.
2008
). However, the movement patterns of
possums and the data used to describe them are substantially different from those
of humans: (1) The Halifax data recorded activity and movement of thousands of
3
Co-incidence in space and time pattern refers to the same spatial locations are attained at the
same time or after a time delay.
