Dan Crisan is professor of stochastic analysis at Imperial College London. He is internationally recognised for his contributions to non-linear filtering theory and published a number of books on the subject. Dan has recently been involved in several large projects on the role of mathematics in the address of global environmental, social and economic problems. Recently he has been editor of the book Mathematics of Planet Earth. A Primer, which collects some of the result of this joint efforts. Dan has been awarded the ERC Synergy Grant on Stochastic Transport in Upper Ocean Dynamics (STUOD), together with Etienne Mémin (Inria), Darryl Holm (Imperial College of London), and Bertrand Chapron (Ifremer). For more information on Dan's many activities, here is the link to his webpage.
Dan Crisan is a collaborator in the project STORM Stochastics for Time-Space Risk Models at the Mathematics Department, UiO. Within this project he is now holding this series of lectures on particle filters and data assimilation.
The lectures are held from Monday to Thursday from 10:00-12:00 on Zoom.
The course is complemented with various exercises to aid understanding of the theoretical material. The solutions of the exercises will be discussed during the tutorial sessions running alongside the lectures from Monday to Wednesday 14:00-16:00 on Zoom. These classes are held by Alex Lobbe, PhD student (UiO) with Dan Crisan and Salvador Ortiz-Latorre, working in the area of filtering and numerical stochastics.
All people interested in the subject are welcome! The course is particularly suitable for PhD students and Master students with interests in the area. All Master students may particularly benefit of the first section of this course.
IMPORTANT MESSAGE: Information about the course and material has been sent to the registered participants on Friday 22nd. Some emails addresses have not been recognised. If you have not received the welcome email with information, please fill in the registration form again with the correct due details.
Registration to the course is free, but mandatory to receive further details and material. Please follow the link.
Please note that Room 720 in NHA's building is booked for broadcasting the lectures at UiO.
Abstract for the Course
In this short course we discuss the application of particle filters as a data assimilation method both from a theoretical and applied perspective.
The course starts with the rigorous introduction of particle filtering theory in discrete time including Bayes' recursion formula and the abstract state space model framework. Additionally, the core particle filtering algorithm is derived and extensions, like model reduction methods, will be explained.
In the final part of the course the emphasis lies on the application of particle filters in computational geophysical fluid dynamics where we focus on recent numerical studies highlighting the use of particle filters to assimilate observational data into fluid dynamical SPDE simulations.