Model evaluation methodology applicable to environmental assessment models [electronic resource].
- Oak Ridge, Tenn. : Oak Ridge National Laboratory, 1979.
Oak Ridge, Tenn. : Distributed by the Office of Scientific and Technical Information, U.S. Dept. of Energy.
- Physical Description:
- Pages: 53 : digital, PDF file
- Additional Creators:
- Oak Ridge National Laboratory and United States. Department of Energy. Office of Scientific and Technical Information
- Restrictions on Access:
- Free-to-read Unrestricted online access
- A model evaluation methodology is presented to provide a systematic framework within which the adequacy of environmental assessment models might be examined. The necessity for such a tool is motivated by the widespread use of models for predicting the environmental consequences of various human activities and by the reliance on these model predictions for deciding whether a particular activity requires the deployment of costly control measures. Consequently, the uncertainty associated with prediction must be established for the use of such models. The methodology presented here consists of six major tasks: model examination, algorithm examination, data evaluation, sensitivity analyses, validation studies, and code comparison. This methodology is presented in the form of a flowchart to show the logical interrelatedness of the various tasks. Emphasis has been placed on identifying those parameters which are most important in determining the predictive outputs of a model. Importance has been attached to the process of collecting quality data. A method has been developed for analyzing multiplicative chain models when the input parameters are statistically independent and lognormally distributed. Latin hypercube sampling has been offered as a promising candidate for doing sensitivity analyses. Several different ways of viewing the validity of a model have been presented. Criteria are presented for selecting models for environmental assessment purposes.
- Report Numbers:
- E 1.99:ornl-5507
- Published through SciTech Connect.
- Funding Information:
View MARC record | catkey: 14148403