Assessment of Conceptual Uncertainty in Hydrosystem Modelling
Accounting for the uncertainty in model choice in a Bayesian framework
Post-Doc Researcher: Anneli Guthke (geb. Schöniger)
Approach
Hydrosystem models are subject to uncertainty in model structure, parameters, forcing terms and data. To specifically address conceptual uncertainty (the uncertainty in model choice), Bayesian model averaging (BMA) is chosen as an integrated modeling framework. BMA is a formal statistical approach that rests on Bayesian probability theory. Posterior model weights are obtained for a set of alternative conceptual models based on their individual skill in reproducing the observed data and the principle of parsimony. These weights can be used for model ranking, model selection or model averaging. The conceptual uncertainty within the set of considered models can be quantified as so-called between- model variance.
Despite the rigorous procedure, a number of obstacles exist(ed) to the wide-spread use of BMA in scientific research and practice. Some of these have been addressed in the course of my PhD project. Several options for evaluating the BMA equations efficiently and accurately have been compared (numerical methods vs. information criteria). Further, the shift in justified model complexity with an increasing amount of available data has been investigated, and a statistical concept to assess the robustness of model ranking against noisy input or output data has been proposed.
The focus of this post-doctoral project is on further extending the applicability of BMA by handling combinations of physically-based models with statistical error models. Error models are introduced to reduce the bias in predictions. The optimal complexity of such error models can be identified within the BMA framework. The goal is to infer improved model structures with realistic parameters that yield most accurate and reliable prognoses.
Peer-Reviewed Articles
- Nowak, W., & Guthke, A. (2016), Reducing model choice uncertainty by collecting most informative data: A rigorous, entropy-based approach to optimal experimental design, Entropy (submitted)
- Lötgering-Lin, O., Schöniger, A., Nowak, W. & Gross, J. (2016), Bayesian model selection helps to choose objectively between thermodynamic models: A demonstration of selecting a viscosity-model based on entropy scaling, Industrial & Engineering Chemistry Research (in press)
- Schöniger, A., Wöhling, T. & Nowak, W. (2015), A statistical concept to assess the uncertainty in Bayesian model weights and its impact on model ranking, Water Resources Research, 51, 7524–7546, doi: 10.1002/2015WR016918
- Schöniger, A., Illman, W. A., Wöhling, T. & Nowak, W. (2015), Finding the right balance between groundwater model complexity and experimental effort via Bayesian model selection, Journal of Hydrology, 531(1), 96-110, doi: 10.1016/j.jhydrol.2015.07.047
- Wöhling, T., Schöniger, A., Gayler, S. & Nowak, W. (2015), Bayesian model averaging to explore the worth of data for soil-plant model selection and prediction, Water Resources Research, 51, 2825–2846, doi:10.1002/2014WR016292
- Schöniger, A., Wöhling, T., Samaniego, L. & Nowak, W. (2014), Model selection on solid ground: Rigorous comparison of nine ways to evaluate Bayesian model evidence, Water Resources Research, 50, 9484–9513, doi:10.1002/2014WR016062
- Schöniger, A., Nowak, W. & Hendricks Franssen, H.-J. (2012), Parameter estimation by ensemble Kalman filters with transformed data: Approach and application to hydraulic tomography, Water Resources Research, 48, doi:10.1029/2011WR010462
Conference Talks & Posters
- Guthke, A.: A Bayesian view on conceptual uncertainty: Analysis and optimal design tools for model evaluation, selection and averaging. - Invited talk at Model Selection Workshop KU Leuven, Sep 2016 (planned)
- Schöniger, A., Illman, W. A., Wöhling, T. & Nowak, W.: Which level of model complexity is justified by your data? A Bayesian answer. - Talk at EGU GA Vienna, Apr 2016
- Lötgering-Lin, O., Hopp, M., Gross, J., Schöniger, A. & Nowak, W.: Prediction of pure component and mixture viscosities using PCP-SAFT and entropy scaling. - Talk at SAFT 2015, Houston, May 2015
- Schöniger, A., Wöhling, T., Samaniego, L. & Nowak, W.: On the various (good and bad) ways to evaluate Bayesian model weights. - Talk at EGU GA Vienna, Apr 2015
- Nowak, W., Wöhling, T. & Schöniger, A.: Lessons learned from a past series of Bayesian model averaging studies for soil/plant models. - Talk at EGU GA Vienna, Apr 2015
- Schöniger, A.: Multi-model approaches to quantify conceptual uncertainty in environmental modelling. - Talk at International Conference on Integrated Hydrosystem Modelling, Tübingen, Apr 2015
- Nowak, W., Schöniger, A., Wöhling, T. & Samaniego, L.: Model selection on solid ground: Comparison of techniques to evaluate Bayesian evidence. - Talk at AGU Fall Meeting San Francisco, Dec 2014
- Schöniger, A., Wöhling, T. & Nowak, W.: How to address measurement noise in Bayesian model averaging. - Poster at AGU Fall Meeting San Francisco, Dec 2014
- Schöniger, A., Wöhling, T. & Nowak, W.: How reliable is Bayesian model averaging under noisy data? Statistical assessment and implications for robust model selection. - PICO-Presentation at EGU GA Vienna, Apr 2014
- Schöniger, A., Nowak, W. & Wöhling, T.: Do Bayesian model weights tell the whole story? New analysis and optimal design tools for maximum-confidence model selection. - Talk at AGU Fall Meeting San Francisco, Dec 2013
- Wöhling, T., Schöniger, A., Geiges, A., Nowak, W. & Gayler, S.: Evaluating experimental design for soil-plant model selection using a Bootstrap Filter and Bayesian model averaging. - Talk at AGU Fall Meeting San Francisco, Dec 2013