Geoscience Reference
In-Depth Information
places is proportional to the masses of places and inversely proportional to the
distance separating them as stipulated in the theory (challenge 3). It can also be
mobilized as a method for estimating the value of the exchanges during the stage of
data construction (challenge 1). Still, it can operate as a filter that allows
highlighting pairs of specific places that exchange more or less than the model
would let us suppose (challenge 2). Finally, it can be introduced in a simulation
model to account for the dynamics of the exchanges between places (challenge 4).
In the remainder of this chapter, these four challenges are examined in more
detail, with an anchor in the operationalization. Then, they will be expounded in a
series of examples, thematically very different but presented following the same
framework, defined by the succession of these four challenges. The purpose of this
part is to explain these various challenges and to illustrate their combinations. In
particular, the links of dependency that they maintain will be shown. The choices
made to overcome a given challenge have, as a matter of fact, implications on the
processing that follows as well as on the type of models referred to at each stage.
2.1.1.
Building the spatio-temporal objects from the empirical observations
(challenge 1)
This challenge relies on the conceptual formalization of spatio-temporal objects
considered in Chapter 1. That stage has given rise to a first model, conceptually
based. Its operationalization gives rise to a physical data model, suitable for a
computing environment in which information will be stored and processed. The
various categories presented in Chapter 1 will be used as a formal framework for
the presentation of examples. It may appear complex to systematically force the
different grids and proposed categorizations to cohabit (object/field, bona fide/fiat,
simple/composite and endurant/perdurant). The benefit resides in that the questions
that this type of specification leads to facilitate a coherent construction of data. This
construction evolves through round-trips between the initial thematic issue, the
different levels of conceptual questions and the existing empirical “bricks”
3
. Thus,
the researcher/practitioner is lead to question himself/herself concretely: what
conception of space is the most suitable to its study? Is there a need to design
objects, or is it more pertinent to apprehend space as a continuous field? For
example, in the study of the glaciers' evolution, should a “glacier” object be defined
and its delimitation characterized (fiat object) in order to follow its evolution? Or is
it preferable to adopt a field-based approach by interpolating a glacier mass index at
3 We will denote as “brick” the raw, elementary and observable material. The empirical
objects will be constructed from these bricks. In some cases, there will be a bijection between
the objects and bricks (for example, the cadastral land parcels), in other cases the objects will
be a composition of bricks (for example, a parcel's land occupation will require the
aggregation of the polygons of land use, which will be called elementary bricks).
