3D management system on a GSDI, preliminary study of dams and ...

WWW.ABACOGROUP.EU
ISO 9001:2008 Certified Quality System
ABACO srl
C.so Umberto I, 43 - 46100 Mantova, Italy
tel. +39 (0)376 222181 fax +39 (0)376 222182
<strong>3D</strong> <strong>management</strong> <strong>system</strong> on a GSDI,
preliminary study of dams and
drainage basins
Fabio Slaviero
Jean Jacques Mezzarobba
ABACO Srl
Geo Tunis 2009 conference
Tunis – December 16‐20, 2009

WWW.ABACOGROUP.EU
ISO 9001:2008 Certified Quality System
ABACO srl
C.so Umberto I, 43 - 46100 Mantova, Italy
tel. +39 (0)376 222181 fax +39 (0)376 222182
Abaco is active since 1990 in the R&D of edge
technologies that can be now applied to GIS
solutions.
In addition to the usual GIS information, we have
now solutions to take advantage of spatial data
analysis, captured with different acquisition
technologies.

<strong>3D</strong> on a SDI
• A <strong>3D</strong> model allows to view and analyse
complex situations
• Having <strong>3D</strong> in a Spatial Data Infrastructure is
not expensive, most of the existing
information is re-used
• Must have, possibly, a continuum on the
territory

<strong>3D</strong> on a SDI
• The <strong>3D</strong> model must be accessible for
customised calculation
• It must be able to represent existing 2D
information (images, landscape, seismic, soil
types, etc.)
• It must help doing preliminary analysis (before
doing real projects)
• It should allow to see the environmental
impact of big instalments, like dams, levees,
bridges, etc.

Dams, levees and basins
• Projects require a multi-disciplinary team,
even for the feasibility study and risk analysis
• <strong>3D</strong> GIS allows to pre-analyse a bigger number
of suitable places
• Immediate view of
• Watershed delineation
• Storage Capacity
• Affected territory
• Environmental Impact

From 2D to <strong>3D</strong>
• A <strong>3D</strong> representation of land requires the
existing GIS information plus a DEM:
• Aerial or Satellite imagery
• Digital Elevation Model
• ….plus any other existing layers

From 2D to <strong>3D</strong>

From 2D to <strong>3D</strong>

DbMAP Flyer Pro
• A <strong>3D</strong> model browsing tool
• An open architecture to accept analysis
methods
• Measuring tools, among which:
• Dams
• Water Basins

Dams

Dams

Dam analysis
• Mathematically and computationally simple:
• Lowest elevation of the barrier
• Determines the set of points at a lower elevation
• Similar to calculating a water source spilling
• It’s enough to consider only the elevation (and not
the function derivatives)

Dam analysis

Drainage Basins

Drainage Basins

Drainage basin analysis
• Mathematically and computationally complex:
• Each cell is considered a water source
• Water moves following “deflux” direction, when its
path crosses the basin it becomes part of it
• Typical model of rainfall
• Calculating each path is computationally not
applicable

Drainage basin analysis
• We needed to find a better way:
• Determining homogeneous sub-regions
• Flow of each cell is followed until it reaches a cell
already calculated, or until possible, starting from
cells at higher elevation

Deflux direction
• 8 surrounding cells
• Flow to lesser elevation (Jenson and Domingue
1988) – simple and computationally cheap, but it is
discrete
• Flow following maximum slope (discrete gradient
providing a direction) – complex, but representing
the reality.

First step: Deflux Regions
Using maximum slope (discrete derivative)

Second step: Merge Regions
Points with lesser elevation in a region
Flowing into another region

Second step: merge regions
• Still fragmented, also due to depressions
• Frequent problem in DEMs due to
imperfections
• Eliminating imperfections on DEM (Jenson
and Domingue ’88, Martz and Garbrecht J
’98) introduces arbitrariness (especially on flat
land)
• Better to eliminate depressions using a
flooding model (Martz and de Jong ’88)

Second step: Merge Regions
Simplification using flooding

Conclusions
An new tool to:
• Build <strong>3D</strong> models which are representing the
reality, which can be easily queried;
• Run analysis, re‐using existing studies and
enhancing them;
• Computationally affordable;
• Avoid spending money for detailed project
studies, when a preliminary analysis is possible.

Thanks
Q&A