Stochastic Methods for Sequential Data Assimilation in Strongly Nonlinear Systems
Abstract
This paper considers several filtering methods of stochastic nature based on Monte-Carlo drawings, in view of the sequential data assimilation in non linear models. They include some known methods such as the particle filter and the ensemble Kalman filters and some other introduced by us: the second order particle filters and the singular evolutive interpolated filter. The aim is to study their behaviour in the simple non linear chaotic Lorenz system, in the hope of getting some insight on more complex models. It is seen that these filters perform satisfactory but our filters have the clear advantage in term of cost. This is achieved through the concept of second order exact drawing and the selective error correction parallel to the tangent space of the attractor of the system (which is of low dimension- ). We have also introduced the use of the forgetting factor, which could enhance significantly the filter stability in this nonlinear context.