Bayesian Modeling for Environmental Health

Course 2

Date

18 – 20 January 2027

Faculty

Dr. Garyfallos Konstantinoudis
Grantham Institute for Climate Change and the Environment, Imperial College London, UK

Dr. Robbie M. Parks
Columbia University’s Mailman School of Public Health, USA

Venue

Wengen, Switzerland

Course description

In environmental and climate epidemiology in particular but also in many other disciplines, spatial and spatiotemporal methods are increasingly used with high-resolution environmental and health data, allowing more detailed investigation of associations between exposures and health outcomes. In this context, Bayesian methods provide a natural and flexible framework for incorporating uncertainty and prior knowledge, borrowing information across neighbouring units, and developing hierarchical structures tailored to complex data. However, the appropriate application of these methods requires careful consideration and technical understanding, which can be challenging for researchers who are new to Bayesian modelling. This course offers an advanced (yet accessible) introduction to Bayesian methods in environmental and climate epidemiology, with particular emphasis on spatial and spatiotemporal applications using high-resolution environmental and health data. It covers the foundations of Bayesian inference, hierarchical modelling, prior specification, Bayesian distributed lag non-linear models, and the use of spatial, temporal, and spatiotemporal conditional autoregressive priors. Participants will also be introduced to widely used software for Bayesian analysis, including NIMBLE and INLA, and will explore applied case studies quantifying the impacts of environmental and climatic factors on infectious and chronic diseases. Through a combination of lectures, hands-on computer labs, and discussion of participants’ own research questions, the course provides a stimulating learning environment in which methodological concepts are directly linked to contemporary challenges in environmental and climate health research.

Course objectives

By the end of our course, participants will be able to:

• Apply Bayesian methods using the context of environmental and climate epidemiology  to analyse associations between environmental exposures and health outcomes, particularly in spatial and spatiotemporal settings.
• Develop and interpret Bayesian hierarchical models using appropriate prior specification, including conditional autoregressive priors and Bayesian distributed lag non-linear models, for complex environmental and health data.
• Use Bayesian software tools such as NIMBLE and INLA to conduct applied analyses and critically evaluate findings from real-world case studies in environmental and climate health research

Course audience

PhD students, early career, or advanced researchers in the areas of epidemiology, public health, or climate modelling or any researchers who want to develop new skills in spatial and spatiotemporal methods. 

The participants should:

• Know the basics of R and RStudio (e.g., installing packages).
• Be familiar with spatial/temporal data and common distributions (e.g., normal, Poisson), helpful but not required.

Course outline

The course runs over three days and consists of short lectures, computer demos, and practical sessions with real-data analysis.

Credits

1.0 ECTS

Course materials

Students should bring their own portable computers with the latest version of R and RStudio.
We strongly recommend only bringing computers you have administration rights for to the course.

Onsite University of Bern IT staff provides support upon e-mail () request.

Course book

We provide course materials: presentation slides, documents illustrating demos, and R script and data.

Course fee

PhD Bern:	CHF 600
PhD other:	CHF 800
Academic:	CHF 1’000
Industry:	CHF 2’000

Registration

Go to registration information

Accommodation

Book your accommodation separately. Please see recommendations for special prices.