Geography Reference
In-Depth Information
been made convincingly elsewhere (see for example, Haining et al. 1998;
Anselin 1999) and would be strong in our case with the CityCycle dataset,
where no simple descriptive statistic can be easily defined that captures the
complex dynamics of an origin-destination matrix.
3.2. Using Flow Mapping
The flow map is a well-established cartographic technique. It can assist in
our understanding of these matrices by mapping transitions between spatial
units. Here, lines depicting the transitions between spatial units are typically
appended with arrows to indicate flow direction and the width of the line is
used to indicate the volume of flow. While the flow map can be used to
develop an understanding of origin-destination matrices, it does not readily
allow the incorporation of other variables such as weather and calendar events
as a component of the visual output. In such circumstances, the analysis of
bivariate spatial data can be analysed using a technique termed the comap in
which plots with overlapping subsets of data are selected using non-spatial
variables. These plots of raw data can be used to give a sense of how spatial
relationships change conditional on one or more external variables (for
example, how bicycle trips vary spatially according to how windy it was at the
time the particular trips were made).
Flow mapping, a visual analytical tool to depict spatial interaction and
movement, has a long history dating back to 1869 where Charles Minard first
used the technique to depict Napoleon's army's advancement towards
Moscow (Minard 1869). The development of computerised flow mapping
tools, however, only commenced in the late 1980s with Waldo Tobler's flow
mapper program. This program was designed to visualise discrete node-to-
node movements (Glennon & Goodchild 2005). Since then, a number of
significant efforts, particularly in computer science, have been made to
develop more advanced tools to visualise flow information in various
standalone applications (see for example, Phan et al. 2005; Guo 2009;
Boyandin et al. 2010; Boyandin et al. 2011).
The flow map, on its own, only conveys pertinent information of an
origin-destination matrix where the objective of the investigation is to study
the matrix in isolation. In circumstances where there is a need to investigate
the extent to which the origin-destination matrix changes as a function of other
variables (for example, under certain weather conditions), another technique
