Models in Policy Formulation and Assessment: The WadBOS Decision-Support System - Environmental Modelling: Finding Simplicity in Complexity - page 336

with the decision-theory methods, analytical techniques,
and software instruments required to work in an effective
manner with the domain models and the data. Each of
the four components has a complex internal structure.

21.3 Decision-support systems

Decision-support systems (DSS) are computer-based
information systems developed to assist decision makers
to address semi-structured (or ill-defined) tasks in a spe-
cific decision domain. They provide support of a formal
nature by allowing decision-makers to 'access' and use
'data' and appropriate 'analytic models' (El-Najdawi and
Stylianou, 1993). The term 'semi-structured' in this def-
inition refers to the fact that DSS are typically applied
to work on problems for which the scientific knowledge
is missing to solve them in an unambiguous manner, or
problems for which there is insufficient consensus rela-
tive to the values, criteria, or norms to define them in
an unambiguous manner (van Delden, 2000). The term
'appropriate' refers to the fact that use is made of the
best available analytical solution methods to approximate
heuristically the unique answer. Thus, the DSS provides
the decision-maker with a suit of domain specific ana-
lytic models considered appropriate for representing the
decision domain. Integrated models play a key role in any
DSS in the sense that their constituting sub-models are
covering, at the least in part, the (sub)domains related to
the decision problem, but also because integrated models
explicitly include the many complex linkages between
the constituting models and related domains. Thus, they
provide immediate access to very rich and operational
knowledge of the decision domain.
The usefulness, richness and scope of the DSS are
predominantly determined by the spectrum and appro-
priateness of the models available from its model base
(1) (Figure 21.1). Three more components fulfil specific
tasks within the DSS (Engelen et al ., 1993): (2) a user inter-
face , the vehicle of interaction between the user and the
system, (3) a database containing the raw and processed
data of the domain and the area at study; and (4) a toolbase

21.4 Building the integrated model

In line with the stated objectives, the approach taken
in WadBOS was clearly bottom-up. It was based on a
reasonable understanding of the characteristic processes
and problems typifying the Wadden Sea, and based on
expertise, a fair amount of complementary knowledge,
and modelling material available from a large number
of organizations. In this way an integrated model was
designed and constructed. This integrated WadBOS-
model is the core element of the modelbase of the
DSS. It consists of linked submodels and represents
the Wadden system as completely as possible with a
view to facilitate the design and evaluation of integrated
policies.
TheWaddenSea is an extensively studied coastal system
and an amazing amount of data, information and knowl-
edge is available. However, this material is very diverse
in nature and spread out among the very many organiza-
tions and people that produced it. With a view to carry out
a gradual systems analysis and to gather, structure, and
link the available material, three consecutive knowledge-
acquisition/modelling sessions were organized (see, for
example, Firley and Hellens, 1991; Gonzalez and Dankel,
1993; de Kok et al ., 1997). Prior to these sessions, a
preliminary requirements analysis had been carried out
focussing strongly on the use and user of the system.
The intended use and functions of WadBOS have been
described in Section 21.2. The envisaged end users are
coastal zone managers and policy-makers. Their profile is
best described as: high-level technicians actively involved in
the design and evaluation of coastal zone management poli-
cies. They perform policy work of a formal/analytic nature
in support of the administrator or politically appointed
person responsible for taking the actual decision and start-
ing the actual policy-implementation process . Thus, this
policy-maker is a technician and not a politician.

User interface

Tool-
base

Model-
base

Data-
base

21.4.1 Knowledgeacquisitionandsystems
analysis

In the first series of knowledge-acquisition sessions, an
extensive groupof potential endusers, all actively involved

Figure 21.1 Basic functional components of the WadBOS
decision-support system.