Bayesian Methods for Environmental Modelling

Mathematical models are essential for gaining insights into environmental systems and predicting their behavior. This summer school provides guidance on how to use Bayesian techniques to treat uncertainties in data, model structure, and parameters quantitatively.

The course consists of lectures and practice sessions with didactical exercises with solutions in R, Julia, and Python. Emphasis is on the concepts and applications of methods, not on mathematical derivations.

Course leader

Carlo Albert, Dmitri Kavetski, Andreas Scheideggers and guest lectures.

Target group

The course is designed for PhD students, post-doctoral researchers, and senior research scientists from all disciplines who utilize mathematical models in their work, with a focus on examples from hydrology and ecology.

Course aim

We cover the following topics:

- Model representation and philosophy: Importance of models, sources of uncertainty in models, description of uncertainty, mathematical representation of models.

- inference and required numerical algorithm such as Importance Sampling, Markov Chain Monte Carlo simulation, and Machine Learning based inference.

- Model predictions: Estimation of the uncertainty of model predictions in the Bayesian framework.