1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
ecological goal functions, ecological modelling, exergy, fisher information, information theory, structurally dynamic modelling
Start Date
1-7-2002 12:00 AM
Abstract
Ecosystems are open, dynamic systems changing both their structure and their function over time. A model, which is an abstraction of the system manifest at the time of its construction, can project future dynamics so long as the fundamental structure of the system does not change. A primary modelling challenge arises when the system structure changes, i.e., internal reorganization, as a result of external perturbation. Ideally, one could include algorithms in the model to anticipate such changes, but that would entail knowing a great deal more about the system and disturbances to it than is possible. Other methods to anticipate, recognize, and adapt to changes are needed. A new modelling approach, structurally dynamic models, has been developed to address this need. Several methods have been used to implement structurally dynamic modelling; one is by optimising an ecological goal function. An ecological goal function is a macroscopic property of the system such as total system throughflow, total system cycling, residence time, or exergy. Structurally dynamic models can be constructed such that the model’s parameters are continually updated –within accepted empirical bounds– to reflect values that generate the optimum value of the goal function. For example, a goal function for maximizing an exergy index has successfully modelled population dynamics. Another proposed goal function is an information index. In this paper, exergy and information indices are applied to a food web model to demonstrate and compare the usefulness of the new information metric for structurally dynamic modelling.
Exergy and Information Indices: A Comparison for Use in Structurally Dynamic Models
Ecosystems are open, dynamic systems changing both their structure and their function over time. A model, which is an abstraction of the system manifest at the time of its construction, can project future dynamics so long as the fundamental structure of the system does not change. A primary modelling challenge arises when the system structure changes, i.e., internal reorganization, as a result of external perturbation. Ideally, one could include algorithms in the model to anticipate such changes, but that would entail knowing a great deal more about the system and disturbances to it than is possible. Other methods to anticipate, recognize, and adapt to changes are needed. A new modelling approach, structurally dynamic models, has been developed to address this need. Several methods have been used to implement structurally dynamic modelling; one is by optimising an ecological goal function. An ecological goal function is a macroscopic property of the system such as total system throughflow, total system cycling, residence time, or exergy. Structurally dynamic models can be constructed such that the model’s parameters are continually updated –within accepted empirical bounds– to reflect values that generate the optimum value of the goal function. For example, a goal function for maximizing an exergy index has successfully modelled population dynamics. Another proposed goal function is an information index. In this paper, exergy and information indices are applied to a food web model to demonstrate and compare the usefulness of the new information metric for structurally dynamic modelling.
