How is Pareto mode used in exploring predictive uncertainty?
Exploration of predictive uncertainty using PEST’s Pareto mode is a lot like hypothesis-testing. Remember that the scientific method is based on the testing of hypotheses. If a hypothesis contradicts the data, then it can be rejected. If it does not, then it is retained (but not necessarily accepted as a good description of reality because other hypotheses may also be consistent with all available data). In environmental management, decisions are often made on the basis of avoiding unwanted future occurrences. Can a model say for sure that a certain occurrence will not happen? Often not, because there is too much uncertainty associated with predictions of future environmental behaviour. However we can use the model to test the hypothesis that the unwanted happening will occur. To do this we “observe” its occurrence through including its occurrence as a member of the calibration dataset, added to the normal membership of the calibration dataset comprised of historical measurements of sy
