2. 3D SPATIAL REPRESENTATION O
ECOSYSTEM
Pelagic marine features are characterize boundaries, dynamic, and full 3D structure
1995; Shyue and Tsai, 1996. These restrictive for data acquisition, as well
modelling and representation. A furth oceanographic data sets is the frequent anis
of data, due primarily to logistics and co expensive sampling at sea. Development of
3D geospatial modelling of the marine pe consequently challenging. This might explai
analysis of oceanographic phenomena is tr environment, limited to static cutting plane
vertical sections either contoured or colou various parameters Head et al., 1997.
Oceanic physical parameters, such as tempe define distinct water masses with more or les
Vertical and horizontal distribution of th influences the oceanic carbon cycles, whi
important role in regulating global cli objective of the Malina oceanographic cam
better understanding of these interactions. spatial analyses in a 3D geospatial model o
can then be of great value. Such a model co use in a resource management or conservatio
A summary of some common and specialise this work are listed in Table 1. These ha
according to criteria for their suitability modelling of the pelagic environment. Th
Criteria ArcGIS 10
Commercial GIS
3D Interpolation -
3D raster representation -
3D vector representation Visualization cuts
- Visualization iso-surfaces
- Visualization volumes
- 3D statistical analyses
- 3D spatial analyses
- Table 1. Review of five geospatial m
3. 3D GEOSPATIAL SOLUTI
REPRESENTATION OF WATER DISTRIBUTION: MALINA CAS
In the south-eastern Beaufort Sea, several typ can be identified, such as the nutrient
Halocline Water UHW. The fractional pre obtained for each of the 243 sampling
accordingly to a method described by Lans geospatial voxel model of this water mass w
Paradigm GOCAD, a scientific visualization 3D geological spatial modelling. This spati
by a grid of 150 x 75 x 100 voxels in x, compressed vertically between water surfa
surface. UHW values were attributed to each through a 3D interpolation of sampling po
kriging.
OF THE PELAGIC
ized by their fuzzy re Gold and Condal,
e characteristics are ell as for geospatial
rther problem with nisotropic distribution
costs associated with of tools available for
pelagic ecosystem is lain why conventional
traditionally in a 2D anes in horizontal and
lour-coded to present perature and salinity,
less fuzzy boundaries. these water masses
hich in turn play an climate. One partial
campaign was to gain ns. Visualization and
l of these phenomena could also be of good
tion perspective. lised tools reviewed in
have been evaluated ty for 3D geospatial
This review indicates that recent efforts of 3D develop
mostly focused on object centred co structures for example: ArcGIS
Indeed, several research teams h adequate 3D marine GIS Arsenau
2009. We consider that an adequ pelagic continuous phenomena
volumetric field representations. T more developed in geomodelling t
the use of 3D raster-based mode voxel VOlume piXEL structures
beginning of 1990, they are still GIS the GIS open-source GRA
exception to this generality, offerin It is also worth noticing that the m
Voronoï tessellations, whose adva various academic works Beni et
2008, are absent so far in comm tools. A draw-back with the geomo
paper is its limited ability for im imperative in oceanographic resea
use of remote sensing. Another ess marine geospatial modelling tool
static cuts in vertical direction. E might seem trivial and does not re
method, this operation is not poss and very limited with EnterVol
ArcGIS that permits volumetric Finally, none of the tools eval
consideration the dynamic natur pelagic phenomena or to assess a
of spatial 3D models, such as cross 2007.
Fledermaus HabitatSpace
CTech EnterVol Commercial
marine GIS Academic
prototype GIS Extension to ArcGI
Limited Kriging
Kriging limited, ID Limited
? Limited
- Limited
- ?
- Limited
l modelling tools from commercial and academic GIS as well a
TION TO ER MASSES
ASE STUDY
types of water masses t rich pacific Upper
presence of UHW was ing points x, y, z
nsard et al. 2012. A s was constructed with
ion tool developed for atial model was built
x, y, z direction and rface and bathymetric
ach voxel in the model points with ordinary
Figure 2. Kriging variance of sp Upper Halocline Water in the Beau
sampling p lopment in the GIS field have
conceptual design using vector IS version 10, Fledermaus.
have recognized the lack of ault et al., 2004; Mesick et al.,
quate representation of marine na needs fully developed
That kind of representation is g tools. The general solution is
dels, commonly referred to as res. Although in use since the
ll mostly absent in commercial RASS might however be an
ring limited volume rendering. more dynamic data structures -
vantages have been stressed in et al., 2011; Ledoux and Gold,
mmercial geospatial modelling modelling tool reviewed in this
image treatment and analyses, search considering the common
essential function for an optimal ool is the visualization of 2D
. Even though such a function require a true 3D interpolation
ossible at present with ArcGIS ol, commercial extension to
tric representation Table 1. valuated permit to take into
ture and fuzzy boundaries of a general predictive capability
oss-validation e.g. Foglia et al. ol
Paradigm Gocad cGIS
Geomodelling tools , IDW
Kriging, IDW, other.
ll as from geomodelling.
spatial 3D model for pacific Beaufort Sea. Black dots indicate
points.
7th International Conference on 3D Geoinformation, May 16-17, 2012, Québec, Canada 22 of 24
The kriging variogram’s dependent predictiv in Figure 2. In general, the lower the e
location, the better is the prediction of the sp The final spatial model permits us to visual
volumes as well as cuts in any plane of the
Figure 3. Geospatial representation of Up approximately the upper 500 m of the
constituted of UHW and voxels containing black dots. Vertical exaggeration in fig
4. DISCUSSION AND CONCL