Travel Reference
In-Depth Information
U
i
=(
M
i
,
A
i
,
C
i
)
(7)
where
U
i
is the use of the
i
th battlefi eld park,
M
is the park 's location relative to the population
of potential users, A is its location relative to other national battlefi elds, and
C
is a vector of
recreational facility characteristics variable across the parks.
Although the competing destinations model has had its critics (Ubøe
et al.
, 2008), particu-
larly with respect to achieving its aim of removing the map pattern from distance decay
parameters, it is nevertheless recognised as being superior to previous models with respect to
both reproducing the interaction fl ows and giving behavioural explanation to distance decay
parameters (Pingzhao and Pooler, 2002). In addition, the spatial information processing
approach has proven infl uential on research on tourist destination choice (Lin and Morais,
2008) as well as tourist shopping behaviour (Kemperman
et al.
, 2009).
Spatial modelling of tourism related mobility
The modelling of tourist mobility and the movement of tourism-related populations has long
been of interest to geographers (e.g. Oppermann, 1995; Flognfeldt, 1999; Bell and Ward,
2000; Forer, 2002a; McKercher and Lew, 2004; Hall and Page, 2006; Lew and McKercher,
2006). Model development has generally taken three, often related, forms. First, there is the
use of mathematical models, such as the spatial interaction models discussed above. Second,
there is the generation of visual representations of spatial data, in the form of maps and
models. Third, there are descriptive models developed from empirical analyses and case
studies.
One of the most infl uential frameworks for describing tourism mobility has been that of
time geography (see also Shoval,
Chapter 22
in this volume). Originally developed by
Hägerstrand (1970) and colleagues at Lund University, Sweden (Pred, 1981), time geography
focuses on the movement and interaction of individuals in time and space and has been
extremely infl uential in recent years as a way of conceiving tourism and leisure-related
mobility (Baerenholdt
et al.
, 2004; Coles
et al.
, 2004; Haldrup, 2004, 2010; Hall, 2005b,
2005c, 2008a; Coles and Hall, 2006; Axhausen, 2007). However, there are signifi cant differ-
ences in the way the concept is applied, with some researchers utilising it more as a conceptual
framework (Baerenholdt
et al.
, 2004; Larsen
et al.
, 2007) than as a formal analytical tool
(Zillinger, 2007). Building on the basis of personal mobility biographies (Frändberg and
Vilhelmson, 2003; Frändberg, 2006, 2009), Frändberg (2008) has used a time-geographical
form of notation to represent transnational mobility as paths in time and space, and to demon-
strate how such representations can contribute to explaining some of the dynamics of long-
distance mobility. An advantage of using time-space paths is that several aspects of an
individual's travel biography can be represented in a single image (Frändberg, 2008).
GIS has been recognised as an excellent platform with which to model the space-time
patterns of individuals, including tourists (van der Knaap, 1999), and their actual and poten-
tial activity paths (Miller, 1991, 1999, 2005b; Miller and Wu, 2000). For example, using GIS
software, in a study of Hong Kong McKercher and Lau (2008) identifi ed a total of 78 discrete
movement patterns, which were categorised into eleven movement styles. The large number
of movement patterns was a refl ection of the interaction of six factors: territoriality, the
number of journeys made per day, the number of stops made per journey, participation in a
commercial day tour, participation in extra-destination travel and observed patterns of multi-
stop journeys. In addition, GIS can be combined with employment and economic data to
generate an improved understanding of labour market mobility, business development and
