Crash Course on Data Assimilation – Theoretical foundations and advanced applications with focus on ensemble methods @ Nansen Center (NERSC), Bergen
May 5 – May 8 all-day

The 4-days course aims at PhD-level students and early-stage scientists intending to apply data assimilation as part of their research. The course should also be useful for students with beginner or little notions of data assimilation. The crash course will cover the basic concepts of data assimilation, focusing on ensemble methods, illustrated with real-scale / operational applications and with practical exercises. It will also include notions of Machine Learning of relevance for data assimilation.

The event is organised by NERSC and NORCE in the framework of the projects DIGIRES and REDDA from the Norwegian Research Council, and supported by the climate research school CHESS (travel costs of CHESS members will be covered by CHESS).

For more details and application, please see their webpage:

Deadline for application: 15 February

Fundamentals of Ocean/Atmosphere Data Analysis @ Oslo Research Park
May 11 – May 15 all-day

Responsible: Joe LaCasce / UiO
International lecturer: Jonathan Lilly / Theiss Research
Max. no. of participants: 12 CHESS students (total participants: 20)
Credit points: 2 ECTS
Registration form here. Deadline: 15 March
Submitted applicant list

Course description:  This course introduces students to essential statistical and conceptual tools for analyzing any type of dataset from oceanography, atmospheric science, or climate.

In this course, the students will learn how to use our creativity together with simple statistical tools to delve into datasets, uncovering whatever information they may contain, and how to shape that information into stories.  In particular, a powerful method called “distributional data analysis” allows us to deconstruct potentially large, multivariate datasets by examining their statistics in two-dimensional slices.  Careful attention is given to the variance ellipse, the fundamental second-order statistical quantity for bivariate data such as velocity.  Data organization and manipulation techniques, visualization strategies, and healthy coding habits are all also addressed.  Finally, the course provides innovative training in the essential mental factors of curiosity, imagination, and objectivity.

Students apply techniques to datasets of their own choosing using the Matlab programming language, and learn further through homework problems and group exercises.

This will be the fourth time a version of this course is offered in Oslo, and the second time focusing on a greatly expanded version of the “low-tech” methods that form the foundation of the data analyst’s toolbox.

This course is strongly recommended for all students wishing to participate in a more advanced time series analysis course with the same instructor, to be offered over two weeks in Fall 2020 at the Alfred Wegener Institute in Bremerhaven, Germany.

Learning modules/structure:  There will be two hours of lectures in the morning sessions and a two-to-three hour lab session in the afternoons. Lectures will be given in the mornings and lab sessions in the afternoons, allowing the students to apply the methods directly to data. The students will also complete a final project on data of their choice. The students employ the statistical and time series analysis toolbox jLab developed by the instructor (  Course notes are available online at (specifically chapters 1-8).

Learning outcomes: At the end of the course, students will be well-prepared to begin efficiently analyzing any dataset they might encounter, while avoiding common pitfalls.  Students gain practical experience through hands-on demonstrations and exercises in Matlab.

Prerequisites:  Students must have a fully functioning version of Matlab with jLab already installed at the start of the course.  Students are expected to bring a dataset of any type that they would like to analyze for a course project.  Multivariate datasets are encouraged.  Model output is also acceptable.