Geography Reference
In-Depth Information
The right image is a first attempt to introduce a grid structure in the classical 3D
GIS GeOxyg
`
ne. GeOxyg
`
ne3D was initially conceived to map solid objects in 3D
(such as road and building) but not field information. If it can map a DTM and
match 3D objects as buildings on this DTM, as many other 3D GIS like TatukGIS
does, it is a priori not adapted to map the
.
However this empty space
is the space of phenomena
. So on GeOxygene3D we enriched our software with
appropriate data structures to map empty space on demand. It is not so difficult: it
requires an easy and efficient construction of 3D grids, on focused areas, with a
chosen grid threshold. Then we can consider that each node of this grid is an object
that has its own symbolization according to the value of the phenomenon. First
results of our developments are illustrated in Fig.
4
(right).
In order to master this type of representation we add the class
3D-V-Grid
to our
model. A
3D-V-Grid
object is a Visual object (V for Visual) that is a grid defined in
the
focused-area
and that have a certain
grid-threshold
.A
3D-V-Grid
object is
computed from one of several
3D-P-Grid
objects. If it is several objects, it means
that the
3D-V-Grid
object has a time and duration; else it has just a time. A 3D-V-
Grid object is composed of nodes located on the grid structure. Each node is
symbolised according to its value by means of the method
f-colour
(val). The
value of each node of computed by means of specific aggregation and interpolation
methods applied on
3D-P-Grid
objects.
A current project is to re-implement this model and methods on a new 3D open
source GIS software developed by Oslandia Company on top PostGis3D.
'
empty space
'
Conclusion
In this paper we present a diagnostics of difficulties we had to face to map
pollution data. Many solutions exist outside our community mixing a set of
tools and ending with a poor symbolization on image viewers (Fig.
4
, left).
However it is important to map phenomena in a rich way for different
reasons. The geographical context should be well represented because it
contextualizes each phenomenon. This contextualization gives meaning and
allows better understanding. Then the phenomena itself should be represented
in a qualitative way to integrate human profile as well as the impact of the
duration of the phenomena on human. Last but not least, a 3D representation
would help understanding the variation of the impact according to the alti-
tude. This is not nothing. A phenomenon and its impacts vary in altitude. This
variation should be communicated.
Many projects related to life quality in cities are going on. Valente
et al. (
2012
) show a computation of particle matter exposure around and in
a school. This research illustrates the use of geographical data to compute the
dispersion of pollutants and the exposure of children to these pollutants in
map air pollutants over the Spanish city.
(continued)
