Dynamical Systems Perspective of Observing System Design for Lagrangian Data Assimilation

Kayo Ide
Institute of Geophsycis and Planetary Physics, University of California, Los Angeles

Christopher K.R.T. Jones
Department of Mathematics, University of North Carolina at Chapel Hill

Hayder Salman
Department of Mathematics, University of North Carolina at Chapel Hill

We present the Lagrangian data assimilation (LaDA) method due to Ide and collaborators (Ide et al 2002, Kuznetsov et al 2003). We invoke an ensemble Kalman filter in order to estimate and forecast the (ocean) state using the shallow-water model (Salman et al, 2005). Based on the augmented state representation, the LaDA eliminates the need for any conventionally used approximation in assimilating the Lagrangian information. This augmentation also allows us to use dynamical systems theory for the design of a comprehensive observing system. We show how deploying drifters in the flow near the (Lagrangian) saddle point enhances the information content of the (Eulerian) flow dynamics extracted from the Lagrangian data using LaDA.

Last Update: September 29, 2005