Geography Reference
In-Depth Information
2.2 Constructing Travel Paths of Individual Movement
The case study data of WCTFS records the tourists' activities only at stops of
known location, not during movement, although each respondent's survey map
contains a sketch of the trip and stops. These were used only in validating the
written answers. A WCTFS stop location is typically a named location such as a
town or attraction. Spatial coordinates are assigned to these locations using a
centroid measure or local landmark. Un-named locations are assigned coordinates
from positions as shown on the survey's map. Accuracy of the stops in most cases
is 250-500 m but with map-marked locations this is harder to ascertain and may be
higher.
Presence outside of those times occurs during movement, which we can record
as the activity between stops. In the original data this activity is not present, only
inferred in both time and space. To infer the missing temporal activity and location
a travel path has to be built into the dataset. The critical part of this is to identify
the path that the tourist takes and then identify the time at different places along
the route. Routes can be roads, tracks, railways, rivers but are predominantly
thoroughfares suited for vehicles.
The technique we use for this is a basic minimum-cost path algorithm run on a
network comprising various grades of sealed and unsealed roadways and a number
of walking tracks. The coverage represents all of the roads on the West Coast and
major roads outside the area. Analysis was run using network tools available in
ArcInfo (ESRI
2008
) and analyzed in ArcMap (ESRI
2008
). This approach rep-
resents roads as a series of segments, each one representing the location of part of
the road. The length in meters can be calculated from this and other information on
sections can be stored with the sections in order to provide a means to represent
attributes such as nature of the road pavement, travel time, typical speed,
restrictions and other variables. Movement between two stops is represented as a
number of segments. A series of such groups can be combined to represent the
tourist's itinerary.
The network analysis tools provide a means to calculate the optimal path
between any two points on the network, i.e. for any trip that a tourist records. This
allows the trip to be represented as a path along the roads and tracks (rather than as
just two points) and for a most likely path to be identified. Furthermore, each
section of the network can be calibrated in terms of travel time, with more or less
sophistication, allowing individual trips to be expressed in minutes. Significant
benefits of this are:
• Travel time estimates can be calculated for any intervening location on the road;
• Movement can be related to the geography and places that are passed through by
a tourist;
• Travel times between survey stops can be compared with recorded survey travel
times;
• The position of an individual at any time can be estimated, so that a dense point
set can be created to allow fast query and visualization of movement;
