|Statistical questions relating to the validation of air quality simulation models /
Brier, Glenn W.
|Brier (Glenn W.), Fort Collins, Colo.;National Environmental Research Center, Research Triangle Park, N.C. Meteorology Lab.
| U.S. Environmental Protection Agency, National Environmental Research Center ; National Technical Information Service,
|EPA 650/4-75-010; EPA-ROAP-21ADO
Air quality--United States--Statistical methods ;
Air quality management--United States ;
Air pollution ;
Statistical analysis ;
Normal density functions ;
Analysis of variance ;
Regression analysis ;
Mathematical prediction ;
Atmospheric motions ;
|Received from HQ
|Some EPA libraries have a fiche copy filed under the call number shown.
|21 p. : ill.
This study examines some of the statistical problems that arise in the validation and evaluation of air quality models. It considers the various scores or indices that can be used in measuring the predictive accuracy of a model and shows how the verification statistics are affected by errors in the input and output data and imperfections in the model. Suggestions are made regarding the major problem of separating input-output data errors from those introduced by a poor mathematical representation of the physical and chemical processes, and recommendations are made regarding validation procedures to be followed as the RAPS data base becomes available.