Geoscience Reference
In-Depth Information
Integrated Assessment Modelling (IAM)
IAM is often mischaracterised as seeking to create tightly specifi ed predictive models.
Most practitioners are keenly aware of the limitations of models and the fundamen-
tal uncertainty involved in forecasting the future. They insist, however, that not
every future is equally possible in biogeophysical terms, and thus, that the goal of
IA ought to be to identify and assess 'not implausible' futures (Yohe et al., 1999).
While prediction may still represent a distant but ideal goal, there are many inter-
mediate goals that are also extremely important in seeking to inform policy responses
to the challenge of climate change.
Rotmans and Dowlatabadi (1998) identify four central benefi ts of IAM. First,
IA can illuminate the feedbacks and relationships between linked systems all too
often studied in isolation. For instance, IAM can represent interactions between
climate change and other global issues such as ozone depletion (which affects CO2
uptake in the Antarctic Ocean), desertifi cation (which can transfer more dust into
the atmosphere), and acid rain. IAMs have also been used to represent the impacts
of a range of non-climate driving forces that exist alongside increasing greenhouse
gas emissions; for instance, land use change contributes directly to climate change
and can also exacerbate the vulnerability of human populations and ecosystems.
Second, IA can represent the contingency of global environmental systems; human
choices about development pathways have fundamental impacts on the viability of
adaptation and mitigation strategies and on the vulnerability of populations. The
entire IPCC process is driven by a scenario based approach where distinct develop-
ment pathways over a century generate large differences in emissions, land use,
energy use, population etc. IAMs can be used to represent the high level interac-
tions between development pathway choices, their impacts on the global climate,
and then in turn the feedbacks of those climate changes on development
pathways.
Third, IA can be used for exploratory purposes, or as a kind of hypothesis
machine. The development of plausible integrated models allows researchers to
develop hypotheses based on their observations of the order of magnitudes of fl uxes
and responses and the sensitivity of the model to variations in input choices. Often
these models simplify more complex disciplinary models to reduced form models
that allow for more successful integration. Integrated models allow researchers to
explore the dependencies between natural systems and to identify critical uncertain-
ties about relationships between them. For instance, Sigman and Boyle's (2000)
classic study of the role of the southern oceans in drawing down carbon dioxide
levels during interglacial periods basically synthesises a large number of disciplinary
studies, connecting up plausible fl uxes and stocks into a simple mass balance frame-
work. As a result, they were able to suggest that the traditional explanations for
the decline in CO2 during the interglacial periods (temperature changes, changes in
salinity, changes in photosynthesis) are not of a suffi cient magnitude to explain a
drop in atmospheric CO2 from roughly 280ppm to 80-100 ppm. Instead they
discuss changes to the acidity of oceans, complex geochemical reaction in the deep
waters of the southern oceans and physical changes resulting from changes in the
distribution of sea ice. Drawing together evidence about the order of magnitude of
effects necessary to change carbon dioxide concentrations, this integrated modelling
exercise generated a number of hypotheses that can be tested through other
research.
