breathtakingly detailed models in many respects, their detailed incorporation of feedback mechanisms into their outputs--a task that is impossible for EBMs and met by individual EMICs only for their narrow domains of application (if it is met at all). Since CGCMs are characterized as a group by their melding of atmospheric, oceanic, and land-based models, let’s begin by considering a representative sample of an important feedback mechanism from each of these three domains.
While feedback mechanisms are not definitive of complex systems like the climate, they are frequently the sources of non-linear behavior in the natural world, and so are often found in real-world complex systems. It’s not difficult to see why this is the case; dynamically complex systems are systems in which interesting behavioral patterns are present from many perspectives and at many scales (see Chapter Three), and thus their behavior is regulated by a large number of mutually interacting constraints. [1] Feedback mechanisms are a very common way for natural systems to regulate their own behavior. Dynamically complex systems, with their layers of interlocking constraints, have ample opportunity to develop a tangled thicket of feedback loops. Jay Forrester, in his 1969 textbook on the prospects for developing computational models of city growth, writes that “a complex system is not a simple feedback loop where one system state dominates the behavior. It is a multiplicity of interacting feedback loops [the behavior of which is] controlled by nonlinear relationships.[2]” The global climate is, in this respect, very
- ↑ The fact that a particular complex system exhibits interesting behavior at many scales of analysis implies this kind of inter-scale regulation: the features of a given pattern in the behavior of the system at one scale can be thought of a constraint on the features of the patterns at each of the other scales. After all, the choice of a state space in which to represent a system is just a choice of how to describe that system, and so to notice that a system’s behavior is constrained in one space is just to notice that the system’s behavior is constrained period, though the degree of constraint can vary.
- ↑ Forrester (1969), p. 9
178
