Anzeige
Erschienen in:
2018 | OriginalPaper | Buchkapitel
10. Practical Examples
verfasst von : Eduard Hofer
Erschienen in: The Uncertainty Analysis of Model Results
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
First, the application of a population dynamics model is analysed. The evolution of a fish population (anchovy) as well as that of their main natural predators (guano-birds) is simulated over the coming 20 years and for a given management strategy. The uncertainty analysis considers the combined effect of 37 uncertain data. “Harvestable anchovy biomass” and “guano-bird population” are the analysed main results. Their computed uncertainty is presented by statistical tolerance intervals with their lower and upper limit shown as functions over the given time. Three measures of uncertainty importance are computed over this time and are compared. The main uncertainty contributors are identified and presented together with the conclusions drawn from the analysis results.
The second example requires the separation of variability and epistemic uncertainty. A carcinogenic substance was accidentally released many years ago, and it is required to know the percentage of the exposed population with dose value above a given decision threshold. In total, over 3000 uncertain data contribute to the epistemic uncertainty of the computed percentage. The correlation between measured values and measurement errors is estimated, and its state of knowledge is taken into account following the method suggested in Chap. 3. Neglecting this correlation shifts the statistical tolerance intervals for the percentage of individuals with dose above the threshold to higher values. Additionally, the individual dose values and their uncertainty need to be known in order to decide about compensation. Uncertainty importance measures indicate where the state of knowledge needs to be improved.