Data Assimilation Meetings at Reading

Date Meeting type Speakers
16 May 2012 Informal DARC meeting David Livings (UoR)
How to Turn an Ensemble Kalman Filter into a Particle Filter
In recent years the Ensemble Kalman Filter (EnKF) has become widely-used in both operational and research data assimilation systems. The particle filter is an alternative ensemble-based algorithm that offers the possibility of improved performance in non-linear and non-Gaussian problems. The subject of this seminar is how to turn an existing EnKF into a particle filter as simply as possible. It will start by reviewing the particle filter algorithm with special emphasis on the materials that are required to implement such an algorithm. Different formulations of the EnKF will then be examined in turn and it will be shown how they can furnish these materials. By examining the algorithms closely, shortcuts will be identified that increase both the simplicity and the efficiency of the resulting particle filter in comparison with a naive implementation.

Page navigation