The method of fractional factorial numerical experimentation is described which permits identification of important model factors (i.e. parameters plus variables) and, perhaps more importantly, the potential interactions between two or more factors - not possible with traditional sensitivity testing. Factorial experiments are then reviewed for a number of environmental modelling areas: inclusion of land surface schemes into global climate models, atmospheric plume rise and lake stratification.
|Number of pages||5|
|Publication status||Published - 1996|
- Factorial experiments
- Sensitivity analysis