An EIS should combine all necessary tools for
managing and analyzing data in the central database.
All clients have the possibility to retrieve all kinds of
data from different sources (quality data, simulation
results, climate data, etc.) from this database. An
important aspect towards interoperability is that
spatial related data are stored in a database which
supports Open GIS technology defined by the Open
Geospatial Consortium (OGC) (OGC, 2005).
All data in an EIS are related to spatial temporal
information, thus all monitoring stations, catchment
areas, river reaches etc. can be visualized and
analyzed in a GIS.
Since a MU has a spatial reference which can be
static (online measurement unit) or even dynamic
(mobile measurement unit) it can be visualized on a
map using spatial services. A river network consists
of nodes and edges and relations to catchments or
reservoirs. A node, for, instance, can be a junction or
diversion, whereas an edge can be a river reach, a
channel, etc. In Table 1 different possible spatial
representations of a MU are shown. In a river
network it is evident that each network node, edge or
region has the possibility to store time series data.
Table 1: Different kinds of spatial representations
Representation Objects in a river network
point source, junction, diversion, etc.
polyline river segment, channel, etc.
polygon catchment, reservoir, etc.
3 SPATIAL DECISION SUPPORT
SYSTEM
Basically, a Spatial Decision Support System
(SDSS) attempts to provide the water-resources
managers with analytical assistance based on spatial
information in making rational choices based on
objective assessment, thereby reducing the element
of subjective opinion (Gunatilaka, 2001),
Malczewski, 1999). This requires a broader
approach, which is otherwise limited within the
narrow realms of hydrology and water resources.
For the decision making process there is the need to
include spatial and quantitative information
wherever possible on economical and environmental
considerations (Clemen, 1996). Therefore, an SDSS
can be regarded as form of artificial intelligence in
which computers are used not only to predict, what
is likely to happen given various assumptions, but
mainly to supplement management experience in
decision-making.
3.1 Spatial Modeling
The first step towards a SDSS is to describe
processes and data by means of hydrological,
hydraulic, sediment transport, meteorological and
ecological models. These models have to be
integrated into general decision making approaches.
Integrated mathematical computer models
comprising hydrological models, hydraulic models,
flood forecasting models, water balance models,
water resources management and reservoir
optimization models as well as water quality models
are in themselves tools that support decision making.
In order to transform the outputs from these models
into practical decisions, they need to be combined
with other type of information, such as details of
infrastructure, possibilities for control, spatial
information etc. In an SDSS these tools combined
with spatial data can be integrated in a GIS
environment. A model can be an internal model,
which runs on the same machine in the decision
support environment, or an external model, which is
an external application like HEC-1 or HEC-HMS
(Cunge, 1992). Important for an external model is
that a preprocessor prepares the data necessary for
the external model and a post processing task which
retrieves the result data back to the central database
(Fürst, 2005).
Figure 2: Modeling Workspace.
A model can be controlled by means of model
parameters. A Model Parameter Set (MPS) contains
all attributes necessary to initialize and control a
model calculation. As input data for a model
calculation all time series data in the EIS are valid.
After successful calculation of a model the results
are stored again as time series data. One model can
be defined as network and can have one or more
predecessor and one or more successor models.
Using this principle, different models are using time
series data from the EIS. All connected models
together with the MPS’s build up a Modeling
Workspace (MW), which is depicted schematically
in Figure 2.
SPATIAL APPROACH IN RIVER BASIN MANAGEMENT USING DECISION MAKING STRATEGIES
321