Geography Reference
In-Depth Information
This data model (Fig.
3a
) is efficient to produce web services such as the one
presented in Fig.
2
. If we want to go further in the analysis and representation of
pollution fields, we need to enrich it and to include the concept of critical zone.
A
critical zone
is an area where the phenomenon lasts a certain time and reaches
values which are dangerous or unpleasant for some profiles. For example babies and
old people are more sensitive to chemical pollution than other people; but old
people are less sensitive to noise than babies. The detection of critical zones
requires the analysis of 3D Pollution field series. Compare with Fig.
3a
, we need
thus to store the 3D pollution field in a Pollution Data Base to make appropriate
analysis, and not only 2D-DPMs.
Some cut, aggregation, or categorisation are applied on the pollution DB to
detect critical zones or to map specific
focused areas.
These methods are classical
in spatial analysis. They can be implemented thanks to existing methods on R for
example. We did not represent these methods but they would be used for example
on a set of 3D-P-Grids to detect Critical zones. The
focused-area
is defined as an
attribute of a
view
.
As a consequence, the data model should be enriched (see Fig.
3b
). It contains
3D pollution grids
(3D-P-grid, also named the 3D pollution field),
critical zone
, and
a class describing the
phenomenon
, as well as objects that allow a better represen-
tation (thanks to objects in the classes
View
,
3D-V-Grid
and
2D-V-Zone
).
The
phenomenon
(like the noise or the chemical pollution) holds information
which describes what a dangerous (or unpleasant) value is according to the value of
the phenomenon and at least the human profile. If dangerous values are described
on a scale going from 0 to 6, each value is (at least) a function depending on the
value(s) of the phenomenon (e.g. the DB and frequency for noise; the number of
gr/m
3
for Benzene), the
human profile
, the
duration
and eventually other criteria
such as the hour in a day (for noise). These functions are essentials to go from
quantitative to qualitative values, as only the expert can interpret quantitative
values. We suppose that one function (named
f-dangerous-value
in Fig.
3b
)is
defined for each phenomenon. It allows computing critical zones. Thus
critical
zones
are computed by analysing the 3D-P-Grid through the filter of the information
given by f-dangerous-value.
A
Critical zone
object is a 3D area. It can be mapped in 2D by means of a
2D-V-
Zone
object (see below), but it holds specific information describing the profile for
which the area is critical and time information (when the zone is critical).
At least an object
View
is described by a focused-area and as defined previously
information to categorise values (a method and a number of categories) as well as a
family-colour
.A
view
is composed by a set of
2D-V-Zone
objects or
3D-V-Grids
objects. The type of representation describes if the representation of pollution is
done by means of zones (objects from
2D-V-Zone
) or grids (objects from
3D-V-
Grid
). We add the attribute
for efficiency purposes, as it is some time
more efficient to map the pollution on a focused area only and not on the entire
geographical area. Focused-area is used mainly for 3D-V-Grid objects (see below).
A
2D-V-Zone
object is a Visual object (V for visual) that is flat but could be
horizontal, vertical or whatever orientation. In Fig.
2
the
2D-V-Zone
objects are
focused area
'
'
