Abstract
Dynamical models are used routinely throughout various branches of geology and decisions are often based on the results of such models. Awareness of some fundamental limitations of any model used is therefore vital in order to avoid decision making based on highly uncertain results. Computer models are based on mathematical models which describe essential features of processes or system behaviour and a systematic approach to investigate and understand model limitations can therefore be applied.
Sensitivity analysis provides a feel for the system response to uncertainties in both assumptions and observations. However, sensitivity analyses do not provide a feel for the level of confidence one should assign modelling results. A probabilistic approach to evaluation of uncertainties and modelling results is therefore demonstrated. Two cases are used as examples for the method and it is demonstrated how the combined sensitivity analyses and probabilistic evaluation greatly improve the use of even uncertain modelling results.
Get full access to this article
View all access options for this article.