ARTICULATION, ACCURACY AND EFFECTIVENESS OF MATHEMATICAL-MODELS - A REVIEW OF FRESH-WATER WETLAND APPLICATIONS

Mathematical models [87] of freshwater wetlands and shallow water bodies were classified by wetland type, location, and degree of nonlinearity, and rated by 3 new indices: articulation; accuracy; and effectiveness. Articulation measures the size and complexity of the model in the 3 modes of components, space, and time. Accuracy combines measures of goodness-of-fit in each mode. Effectiveness measures explanatory power as a combination of articulation and accuracy. For the models reviewed accuracy was seen to fall with increasing articulation, probably as a result of increasing complexity and cost. Effectiveness, however, rose to a maximum at intermediate articulation and then fell, reflecting the fact that highly accurate models tended to be low in articulation (they said much about little), while highly articulate models tended to be low in accuracy (they said little about much). These methods for ranking models may prove useful for further analysis and the results of this analysis may provide a useful guide to model builders concerned with maximizing the effectiveness of their models using limited resources.