DARC/NCEO Data Assimilation Training course, Univ. of Reading, 2023


Link to the timetable below

Data assimilation is one of the core activities needed to deliver weather forecasts, and to help understand the environment.  This course will provide an introduction to data assimilation methods used in numerical weather prediction, including some of the latest developments. All methods will be derived from a common framework to allow a full overview of their assumptions and applicability. Furthermore, participants will get the opportunity to apply data assimilation methods to small numerical models, in order to obtain hands-on experience with the different methods.

  • When?
    • Tuesday 9th – Friday 12th May 2023.
  • Where?
    • University of Reading, UK (Covid restrictions permitting; an on-line course will be offered if in-person lectures are not possible).
    • A map of the campus is available here.
  • Who is it for?
    • This course is suitable for postgraduate students, post-doctoral researchers, and weather service employees who would like an introduction to, and a deeper understanding of, data assimilation.
  • How will the course be delivered?
    • The course will be a combination of lectures and computer practicals.
  • What topics will be covered? (to be finalised)
    • Building-blocks of Data Assimilation
    • Introduction to Data Assimilation
    • Variational Methods: Theory and practical issues
    • Computer practicals Variational methods
    • Ensemble Kalman Filter Methods: Theory and practical issues
    • Computer practicals Ensemble Kalman filters
    • Non-Gaussian data assimilation
    • Hybrid Methods: Theory and practical issues
    • Computer practicals Hybrid methods
    • When to use which method
    • Applications of Data Assimilation
  • What is required to attend the course?
    • Some knowledge of basic calculus, statistics, and linear algebra (i.e. vectors and matrices).  Mathematical primers are available below.
    • A laptop computer, to run the practical sessions.
  • Who will deliver the course?
    • Lectures will be delivered by members of DARC including: Alison Fowler, Amos Lawless, Jochen Broecker, Ross Bannister, Sarah Dance, and Yumeng Chen.
    • Practical sessions will be directed by the lecturers and PhD students.
  • How much will it cost to attend?
    • The course is free to attend!  Attendees will need to pay for their own travel, accommodation, and subsistence.
  • How will I run the practicals?
    • Practical examples will be accessed via a Jupyter notebook environment accessed from a web browser on your own laptop.
  • How do I apply?
    • To apply for a place, please send an email to Uzma Saeed (u.saeed@reading.ac.uk), with a short description of your current research interests (around 50-100 words) by Friday 7th April 2023.  Applications will be looked at as they arrive, until all the places have been allocated.

Supplementary Material

  • Campus map (web page)
  • TO BE UPDATED: Setting-up for the computer practicals (pdf)
  • TO BE UPDATED: Practicals – variational and ensemble Kalman filter (zip)
  • TO BE UPDATED: Practicals – hydrids (zip)
  • TO BE UPDATED: Practicals – Bayes theorem and particle filters (zip)
  • Maths primer (pdf)
  • Useful linear algebra (pdf)
  • Inner and outer products (pdf)
  • Covariances (pdf)
  • Basic statistical concepts (pdf)
  • Notation (pdf)
  • List of suggested references (pdf)

Course timetable and lecture/practical material

The course will be held mainly in the Brian Hoskins Building (aka Meteorology Building, BH/Met, building 58), but also in the Agriculture Building (Ag, building 59).  These are on the main Whiteknights campus of the University of Reading (campus map). Registration and most lectures will be in room BH/Met 1L61.

Handouts will not be provided by the lecturers, but lectures can be downloaded via the on-line timetable below when they become available.