Hypothesis Testing by Simulation: An Environmental Example

Fedra, K. (1981). Hypothesis Testing by Simulation: An Environmental Example. IIASA Working Paper. IIASA, Laxenburg, Austria: WP-81-074

Preview

Text WP-81-074.pdf Download (1MB) | Preview

Abstract

The study of environmental systems as ecological, physico-chemical as well as socio-economic entities requires a high degree of simplifying formalism. However, a detailed understanding of a systems function and response to various changes for the explicit purpose of systems management and planning, still requires complex hypotheses, or models, which can hardly be subjected to rigorous tests without the aid of computers. Systems simulation is a powerful tool when subjecting complex hypotheses to rigorous tests of their logical structure, as well as a possible means for rejecting or corroborating the underlying hypotheses.

The complexity and variability of environmental systems, the scarcity of appropriate observations and experiments, problems in the interpretation of empirical data, and the lack of a well established theoretical background make it difficult to test any possible conceptualization, or hypothesis, describing a given system. A formal approach to hypothesis testing, based on numerical simulation, which explicitly considers the above constraints, is proposed.

Based on a data set from the North Sea, a series of hypotheses on the structural relations and the dynamic function of the pelagic food web is formulated in terms of numerical models. Hypotheses of various degrees of aggregation and abstraction are tested by comparing singular statements (predictions) deduced from the proposed hypotheses (the models) with the observations. The basic processes of primary production, consumption, and remineralization, driven by light, temperature, and advection/diffusion, are described in systems models ranging in complexity from two compartments to many compartments and species groups. With each of the proposed models, yearly cycles of the systems behavior are simulated. A comparative analysis of the response of each of the models allows conclusions to be drawn on the adequacy of the alternative hypotheses. This analysis also allows one to reject inadequate constructs, and provides some guidance on how to improve a certain hypothesis, even in the presence of a high degree of uncertainty.

Actions (login required)

View Item