Department Systems Analysis, Integrated Assessment and Modelling

Overview

Mathematical models are essential for gaining insights into environmental systems and making predictions about their behaviour. In this summer school, we provide guidance on how to use Bayesian techniques to treat uncertainties in data, model structure, and parameters quantitatively. This course is designed for PhD students, post-doctoral researchers, and senior research scientists who work with mathematical models in all disciplines. However, the presented examples are mostly from hydrology and ecology.

The course consists of lectures and practice sessions with didactical exercises. We will also have time to discuss your individual modeling challenges. Emphasis is on the concepts and applications of methods, not on mathematical derivations. To profit most from the course, you should feel comfortable with basic probability calculus, and reading and writing simple code. For those in need of a refresher, we offer optional lectures on Sunday covering probability calculus fundamentals and providing a brief introduction to R.

Attending the school in person will provide you the greatest benefit. If this is not possible and you want to participate remotely, please get in touch with us. 

The course is equivalent to 2 ECTS points.

Topics

We cover the following topics:

• Model representation and philosophy: Importance of models, sources of uncertainty in models, description of uncertainty, mathematical representation of models.
• Model inference: Sensitivity analysis, model calibration, likelihood construction, prior formulations, Bayesian inference and required numerical algorithm such as Importance Sampling and Markov Chain Monte Carlo simulation.
• Model predictions: Estimation of the uncertainty of model predictions in the Bayesian framework.

For details please have a look at the Program 2022 (Program 2023 will follow).

Location

The 14th Summer School will be held for the first time in Kastanienbaum near Luzern (Switzerland), an idyllic location at the shores of Lake Lucerne. See here for more details.

Accommodation

Basic accommodation is available at the site: single rooms (CHF 36), 3-bed rooms (CHF 20) and dormitories (CHF 15). The rooms will be filled on a first-come-first-serve basis. Alternatively, the location can be reached by public transportation in about 30 min from Lucerne (see here). 

Requirements

The participants should be familiar with basic probability calculus and have some knowledge of R, Python or Julia. The probability calculus exercises should help you to decide whether or not to attend the introductory lectures on Sunday.  If you have any doubts, please do not hesitate to contact us.

Course Fee

The course fee is CHF 800. For participants belonging to an institution of the ETH domain, a reduced fee of CHF 400 applies. This includes documentation, coffee, lunch and dinner, but not accommodation.

Application

Please send your application by email to  Jacqueline Eicher by April 15, 2023. The number of participants is limited to 20. The participants will be considered according to the time of their application.

Lecturers

The course will be taught by Carlo Albert, Peter Reichert, Andreas Scheidegger and Marco Baity Jesi from Eawag Dübendorf and ETH Zurich, Switzerland, and Dmitri Kavetski  from the School of Civil, Environmental and Mining Engineering, University of Adelaide, Australia.