Geoscience Reference
In-Depth Information
In the majority of the cases handled, there are issues concerning the referential,
both in terms of processing and interpretation of the spatio-temporal phenomenon at
stake: in relation to what can the change be described and interpreted? The issue is
different in the case where it concerns following in time the overall organization of a
spatial system (first part of Table 3.1) or the entities making up this system (second
part of Table 3.1):
- The four cases dealt with in the second part of Table 3.1 reveal a variety of
links to the referential. In the first case, which gives rise to two examples, the
referential chosen refers to the whole period, and the evolution is interpreted
accordingly through the system's average evolution. This approach allows us to
focus on the specific characteristics of certain entities (in the two examples in
section 3.3.1, the concern is about cities) or groups of entities registering a rhythm
of change slower or faster than the general tendency of the system. In the second
case, the referent is absolute: in the example of section 3.3.2, it relates to the first
date of the considered period, as is often the case in applications (in this example,
the employment centers in 1999, regarding their existence and their spatial
inscription). The point of focus then specifically concerns the change that has
occurred from a situation considered as initial, which has been observed at a given
time, and that served as a basis for comparison in the representation of change. In
the third case (land cover, section 3.3.3), the referential is the intitial situation /
transition matrix couple. Indeed this matrix, at the heart of the method, directly
represents the manner in which the process of change operates.
- The five cases relating to the following-up of the overall organization of a
spatial system illustrate another issue concerning the referential. In the first case
(section 3.2.1), where the purpose is descriptive the analysis does not depend on a
referential. A gradual representation of differentials is simply represented and
interpreted by the user. In the other two cases referred to in section 3.2.1, the
referential is constituted by a statistical law or a spatial analysis model, the rank-size
law in one case, the Clark's laws and the gravity model in the other. In the case of
explanatory models dealt with in section 3.2.2, there is again no referential, the
variable “to be explained” representing a measurement of change.
These different ways of understanding change in a spatial system are naturally
very complementary. Although proposing different insights, they are all oriented
toward an objective of representation and description. The question of
how
is at the
center of the questioning. The term process has only been mentioned in
the context of the last example. In order to understand the involved processes, the
underlying mechanisms of change have to be identified, and then the reason
why
a
change has taken place in such a way, why a system has evolved in such a direction
etc. Chapter 4 is dedicated to the modeling of these processes that are essential to
give an idea of the systems' future dynamics.
