OPTIMIZATION OF EXERGY IN A STRUCTURAL DYNAMIC-MODEL

The hypothesis that ecosystems are optimizing exergy and the effect of applying exergy as optimizing function in modelling are tested in a s tructural dynamical model. The model used was developed to describe th e development of the phytoplankton community in a Danish shallow lake as a result of biomanipulation. Exergy can be seen as the free biogeoc hemica energy of a system compared ot its surroundings. The function h as previously been proposed as a goal function for ecosystem developme nt. In other words, exergy is thought to act as a quality indicator an d to direct and govern the development of parameters in the ecosystem modelled. Traditional mathematical unconstrained optimization was foun d not to be adequate in the case described, since unrealistic values w ere reached during the optimization runs. Constraints on parameters we re introduced by transforming them into a hyperbolic cosine function. The effects of changing optimization intervals, i.e. the frequency of optimization during the simulation, and the possible change per interv al, i.e. the rate of change, were analysed. The first was to mimic a r esponse time of the system illustrating either memory or resilience of the system. The second was thought to reflect the capability of varia tion of the genetic pools in the system. Both optimization interval an d the allowed change per interval affected the rate with which the mod el evolved with time. Decreasing intervals and increase of rate speeds up the system. The sequence in which the parameters were manipulated was in accordance with observed dominance in natural ecosystems. The p arameter most often changed by the optimizer was the maximum growth ra te of the algae.