Hauhs, M; Lange, H: Foundations for the simulation of ecosystems in Lenhard, J., Küppers, G., Shinn, T.: Yearbook Sociology of the Sciences, Simulation, Pragmatic Constructions of Reality, Springer, 25, 57-77 (2006), doi:10.1007/1-4020-5375-4_4 [Link] | |
Abstract: Modelling is a general activity from which the concepts of computation and simulation are derived depending on context and task. Modelling can be performed under two paradigms: a traditional one based on dynamic systems theory originating from physics and an interactive one developed in computer science. The first modelling paradigm is appropriate in areas where a formal theory is available; rigorous applications to specific situations are characterised as computation, with physics and meteorology as examples. In other areas, where practical experience derived from past interactions with the system is often based on heuristics, one finds the term simulation. Under the traditional modelling paradigm simulation implies a less rigorous approach than computation. This is different under the second modelling paradigm: Here, interactive simulation is more powerful than algorithmic computation and adds an extra dimension to modelling. Ecosystem management serves in this case as the example. The underlying assumptions implicit in the two modelling approaches are translated into three different definitions for ecosystems: two of which used in the context of algorithmic, computational models (in earth- and bio-sciences) and one version is suited for interactive simulation in ecosystem management. We discuss the limitations that may occur among these various perspectives at ecosystems and how interactive simulation models offer novel ways to characterise or overcome them. Under the traditional modelling paradigm, some problems appear complex, but tractable in principle while under the new interactive paradigm some of the former elusive problems become tractable. Other formerly complex tasks appear now as intractable in principle. A more refined concept of simulation models may thus change the perception on ecosystems. The impact of computer science on ecosystem research is that these differences can be expressed and tested. |