Report on the Advanced Statistics Training for Climate Research

The “Advanced Statistics Training for Climate Research” course took place at the Geophysics Institute at the University of Bergen from 26-29 October 2021. The course was delivered by Prof. David Stephenson and Dr. Stefan Siegert from the Department of Mathematics at the University of Exeter.
The course aims to give climate scientists a deeper appreciation of statistical concepts and modelling, and awareness of some relevant areas of modern statistics, and the practical ability to apply statistical methods and interpret them using the powerful free statistical language R. The course consists of a mixture of lectures with two practical hands on sessions using R each day. Topics covered included probability and inference, linear modelling, multivariate statistics, time series modelling, spatial statistics and extreme value modelling. Students were also encouraged to bring along their own research problems for discussion.
The course was attended by 9 participants from UiB, UiT and NERSC. The students were all actively involved in a diverse range of climate-related research areas such as paleoclimate, wind power, and hydrometeorology. An initial discussion revealed that despite regularly using statistics in their research, none had really received any training in more advanced statistical modelling methods useful for climate scientists.
In addition to learning about modern statistical reasoning and developing the ability to use the R language, some of the students gained additional insights into how to analyse their own data and problems e.g. extracting signals from Formanifer experiments and records. By the end of the week, heavy rainfall led to the October monthly rainfall total exceeding the previous record and so this was used as a good data example in the extreme value session.

Text: David Stephenson, photo: Mandy Kong