Data Assimilation Meetings at Reading

Date Meeting type Speakers
30 May 2012 Special double DARC seminar 11-11.50: Alberto Carrassi (IC3 Barcelona)
Accounting for model error in data assimilation
We present the progress and outcomes achieved in the framework of the Belgian Science Policy Project "A new approach to data assimilation accounting for the dynamics and statistics of the model error". A new approach to deal with model error due to uncertain parameters and to the presence of unresolved scales is discussed. An alternative formulation of the extended Kalman filter, referred to as Short-Time Augmented Extended Kalman Filter (ST-AEKF), for state and parameter estimation is presented first. In this algorithm, the evolution of the model error generated by the uncertain parameters is described using a truncated short-time Taylor expansion within the assimilation interval. This allows for a simplification of the forward propagation of the augmented error covariance matrix with respect to the classical state augmented approach. The model error due to unresolved scales is then considered. By treating it as a deterministic fully correlated process an equation for the model error covariance required in the Kalman filter update is derived along with an approximation suitable for application with large scale dynamical systems typical in environmental modeling. A similar approach is then implemented in the context of variational data assimilation procedure and a new formulation of the weak-constraint 4D-Var is discussed.

Page navigation