Geoscience Reference
In-Depth Information
Exploration: models for learning
Besides prediction, models are vehicles for learning about the 'real' world. This is
particularly true of simulation models. Recently, simulation models have become
seen as systems that are open to examination in similar ways to other 'traditional'
experimental systems (e.g., see Humphreys, 1995/96; Dowling, 1999; Winsberg,
2001; 2003; Peck, 2004). Certainly, the application of simulation modelling in some
disciplines falls between traditional theorising and experimentation (Humphreys,
1995/96; Dowling, 1999). This approach opens up the possibility that following
Dowling (1999), simulation models provide a means of 'experimenting on theories'.
'Experimental' simulation modelling seeks to mimic systems in silico 2 . The in silico
form has the advantage that it can be manipulated in ways the 'real' world cannot;
global climate change models are obvious examples of this (Frigg and Hartmann,
2006). Using models in this manner is a two-step process: we learn about the model
and then transfer knowledge about the model to the target system. In practice,
however, analysis often concentrates predominantly on the model. Nevertheless, it
must be remembered that the model is a tool designed to help understand the real
world; the (often understated) diffi culty with detailed models is maintaining that
connection (O'Sullivan, 2004; Frigg and Hartmann, 2006).
Models for integration: adaptive and participatory approaches
Models have become important tools for aiding in the decision-making process (e.g.,
forecasts of air quality are used to inform decisions about public health). Such
modelling has often been viewed as the domain of the 'expert' and has been isolated
from the rest of the decision-making process. Recently, this has begun to change as
models are seen as integrative tools. Adaptive environmental management and
assessment (AEMA) is an iterative process of structured learning through modelling,
fi eld experimentation and system monitoring (Walters, 1986). AEMA uses models
to aid in the synthesis and integration of data and understanding, and to identify
and reduce uncertainty. For example, Walters et al. (2000) used a series of concep-
tual and simulation models to fi lter various alternatives for restoring the fl ow regime
affected by the Glen Canyon Dam in the Grand Canyon. Their models considered
multiple spatio-temporal scales from localised algal responses to long-term patterns
of sedimentation. They were used to: (i) highlight key areas of uncertainty in
the system, and (ii) identify components of the system potentially amenable to
controlled fi eld experimentation. Model outcomes demonstrated the potential
inability of the current monitoring framework to detect ecosystem responses to
either experiment or management. Thus, models form(ed) part of an iterative and
adaptive process, in which knowledge and understanding are constantly refi ned and
management practices adapted to refl ect this.
Models are also used to facilitate communication both between researchers in
different disciplines and between the various stakeholders involved in environmental
decision making. Castella et al. (2005) provide an interesting example of this
approach. Castella et al. used a range of tools including a narrative model, an ABM,
a role-playing game (derived from the ABM) and a GIS in an attempt to understand
human-environment interactions and LUCC following Vietnam's doi mois eco-
nomic reforms of the 1980s. The ABM explicitly considered: (i) farmers' decision-
making strategies, (ii) the institutions that control resource use and access, and (iii)
Search WWH ::




Custom Search