Modeling forest growth I. Canonical approach

The paper proposes a canonical modeling approach as a useful tool for the analysis of complex phenomena in forestry. In contrast to most traditional models in forestry, this approach offers the advantage that no assumptions about the mechanisms of the underlying physiological processes and their translation into mathematical functions are necessary. Instead, the structure of the describing equations is derived directly from qualitative information about the processes that govern the phenomenon with methods of general approximation theory. All parameters of the model have a clearly defined meaning and are, at least in principle, measurable from standard experiments. The companion paper illustrates the canonical modeling process with a specific, quantitative analysis of biomass budgets of Scots pine (Pinus sylvestris). It shows how information about biomass budgets is translated into a specific model, how the model parameters are derived from experimental data on compartment sizes and fluxes, and what kinds of questions can be analyzed. It is demonstrated that none of the typical assumptions about growth rates, relationships between roots and shoots, or allometry are made, but rather that all these phenomena are produced by the model as output.