1: \begin{abstract}

2:   We introduce a new multilevel ensemble Kalman filter method (MLEnKF)

3:   which consists of a hierarchy of independent samples of ensemble

4:   Kalman filters (EnKF). This new MLEnKF method is fundamentally

5:   different from the preexisting method introduced by Hoel, Law and

6:   Tempone in 2016, and it is suitable for extensions towards

7:   multi-index Monte Carlo based filtering methods.  Robust theoretical

8:   analysis and supporting numerical examples show that under

9:   appropriate regularity assumptions, the MLEnKF method has better

10:   complexity than plain vanilla EnKF in the

11:   large-ensemble and fine-resolution limits, for weak approximations

12:   of quantities of interest. The method is developed for

13:   discrete-time filtering problems with finite-dimensional state

14:   space and linear observations polluted by additive Gaussian noise.

15: 

16:  \bigskip

17: \noindent \textbf{Key words}: Monte Carlo, multilevel, convergence

18: rates, Kalman filter, ensemble Kalman filter.

19: %   \bigskip

20: 

21: \noindent \textbf{AMS subject classification}: 65C30, 65Y20. 

22: 

23: \end{abstract}

24: