Geoscience Reference
In-Depth Information
between the municipalities. Sixty-seven employment centers have thus been
identified and delimited (Figure 3.12).
The follow-up of composite entities, that is to say whose construction is the
result of a sequence of processes, as is the case here, raises the problem of the
identification of the temporal referential. This issue has been raised in section 3.1.3:
how to detect change, when change at the level of the observations may also induce
change at the level of the parameters involved? How can we interpret change when
observations and means of observations change. Is it possible to construct
employment centers in 2006 on the basis of the same methodology as that used in
1999 and to compare them? Are we sure we are working on the change of the
objects? Or is there a deeper functional change of the metropolitan system involved
that makes the comparison at the level of the centers impossible?
Faced with the difficulty of resolving such an issue, the choice was made to use a
classification method based on decision tree logic. Six types of theoretical change
have thus been identified
a priori
:
1) disappearence of an employment center;
2) appearance of an employment center;
3) spatial extension of a previously existing employment center;
4) decrease of the spatial extent of a previously existing employment center;
5) splitting of a center into two new employment centers;
6) fusion of two employment centers.
The type of categories of change presented in Chapter 2 in the case of simple
objects is encountered once more. Since composite objects are dealt with here, the
change must be identified with criteria similar to those that have served to construct
the objects. The point is then to develop the criteria that will allow us to assess the
most probable change. To identify the cases of “appearance”, a score has been
calculated for all the municipalities that did not belong to a center in 1999, on their
ability to attract. Then, for each center existing at the first date, four scores have
been calculated to assess: (1) its capability to polarize; (2) its propensity to spread
over the neighborhood; (3) the intensity of its internal cohesion; (4) its propensity to
interact with other neighboring centers. A decision tree was then built by
establishing for each type of change, the criterion or criteria to analyze, and the
associated thresholds. The score on the capability to polarize has for example been
called up on all the municipalities to identify a possible “appearance” (high score) or
on all the centers to identify, on the contrary, a “disappearance” (low score). The
score measuring the strength of the internal cohesion has, for its part, been used to
identify the disappearance, the retraction, or the scission of a center existing at the
