Representing Human Decision-Making in Environmental Modelling - Environmental Modelling: Finding Simplicity in Complexity - page 284

Figure 18.1 Schematic of the IMAGE 2.1 integrated assessment model structure. (Modified with permission from Alcamo, J.,
Leemans, R. and Kreileman, E. (eds) (1998) Global Change Scenarios of the Twenty-first Century, Results from the IMAGE 2.1
Model, Elsevier Science, Oxford).

environment model ready to calculate land-use changes
in the next model iteration (Figure 18.1).
Thus, the general statements of the SRES storylines
are used to define parameters for models like IMAGE
2.1, which then produce estimates of global carbon bud-
gets and GHG emissions (the scenarios) to be used as
inputs to the sophisticated GCMs. The danger of this
approach is that while the assumptions and uncertainties
enfolded in the resulting scenarios are obvious when one
has worked with the IAMs themselves (e.g. van Asselt
and Rotmans, 2002), it becomes less and less obvious
as these scenarios become hidden deeper and deeper
within the global climate-modelling process. It is likely
that many policy- and decision-makers read the results
of GCMs with very vague understanding of the scenarios
on which their representation of alternative futures are
based. 2 Understanding these scenario inputs is particu-
larly important given the difficulty in modelling human

actions, as highlighted in this chapter (even at much
smaller time and space scales than the global 100-year
emissions scenarios).
Another problem with this approach is the separation
of the detailed simulations of the climate system (the
GCMs) from the human-landscape-atmosphere carbon
budget models (the IAMs). Climatic factors are a major
determinant of carbon and other GHG emissions from
both natural and anthropic sources (as the very structure
of IMAGE 2.1 acknowledges - Figure 18.1). Decoupling
models like this emphasizes computational physics at
the expense of dealing with human-environment inter-
actions explicitly and in detail. These interactions are
critical in controlling the decisions that influence the
composition of the physical climate being modelled in
the first place. Static projections (based on fixed story-
lines) fail to deal explicitly with human adaptations that
will take place in response to a changing environment
(Wainwright, 2008). These shortcomings encourage the
use of model results by policy makers for self-centred gains
and can act to accentuate the very debates that such mod-
els are supposedly trying to overcome (see discussions in
Oreskes et al ., 1994; Demeritt, 2001; Hulme, 2009). The
failure of the United States to take on board the Kyoto

2 Compare the length of the Summary for Policy Makers for the
fourth IPCC report (22 pages) with the entire report (over 2500
pages in three volumes, which themselves only make limited
reference to the generation of the scenarios).