Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Modelling of real systems is often complicated by the presence of uncertainties. A quantitative uncertainty/sensitivity analysis can provide conclusions about key sources of uncertainty. This paper attempts to explain how uncertainty is introduced to a model through the parameterisation process, what steps can be undertaken to recognise this uncertainty, and how to incorporate it into model predictions. The paper gives a brief exposition of sensitivity analysis for a simple nonlinear stochastic model namely, a silage maize harvesting model. For this sensitivity analysis a metamodel approach is used. The metamodel parameters are estimated using design of experiment methods. Sensitivity analysis efficiently selects more important and less important model parameters. First-order polynomial metamodels may be sufficient to describe sensitivity characteristics even for nonlinear problems. For the detection of highly nonlinear effects and estimation of parameters interaction sensitivity the second-order polynomial model approach should be recommended. The results of this case study are discussed.