You are here:
WHEN HAS A MODEL BEEN SUFFICIENTLY CALIBRATED AND TESTED TO BE PUT TO EFFICIENT USE?
Citation:
Rossman*, L A. WHEN HAS A MODEL BEEN SUFFICIENTLY CALIBRATED AND TESTED TO BE PUT TO EFFICIENT USE? JOURNAL OF ENVIRONMENTAL ENGINEERING 123(11):1070-1071, (1997).
Impact/Purpose:
information
Description:
The question of what degree of predictive error is acceptable for environmental models is explored. Two schools of thought are presented. The universalist school would argue that it is possible to agree on general acceptance criteria for specific categories of models, particularly those that have a documented track record of performance. The relativist school believes that modeling situations are unique and that a one-size-fits-all criterion would be unreasonable. More attention should be paid to explicitly representing a model's prediction uncertainty and to incorporating testing and acceptance criteria into model development.