## The Annals of Applied Probability

### On the stability and the uniform propagation of chaos properties of Ensemble Kalman–Bucy filters

#### Abstract

The ensemble Kalman filter is a sophisticated and powerful data assimilation method for filtering high dimensional problems arising in fluid mechanics and geophysical sciences. This Monte Carlo method can be interpreted as a mean-field McKean–Vlasov-type particle interpretation of the Kalman–Bucy diffusions. In contrast to more conventional particle filters and nonlinear Markov processes, these models are designed in terms of a diffusion process with a diffusion matrix that depends on particle covariance matrices.

Besides some recent advances on the stability of nonlinear Langevin-type diffusions with drift interactions, the long-time behaviour of models with interacting diffusion matrices and conditional distribution interaction functions has never been discussed in the literature. One of the main contributions of the article is to initiate the study of this new class of models. The article presents a series of new functional inequalities to quantify the stability of these nonlinear diffusion processes.

In the same vein, despite some recent contributions on the convergence of the ensemble Kalman filter when the number of sample tends to infinity very little is known on stability and the long-time behaviour of these mean-field interacting type particle filters. The second contribution of this article is to provide uniform propagation of chaos properties as well as $\mathbb{L}_{n}$-mean error estimates w.r.t. to the time horizon. Our regularity condition is also shown to be sufficient and necessary for the uniform convergence of the ensemble Kalman filter.

The stochastic analysis developed in this article is based on an original combination of functional inequalities and Foster–Lyapunov techniques with coupling, martingale techniques, random matrices and spectral analysis theory.

#### Article information

Source
Ann. Appl. Probab., Volume 28, Number 2 (2018), 790-850.

Dates
Revised: January 2017
First available in Project Euclid: 11 April 2018

https://projecteuclid.org/euclid.aoap/1523433625

Digital Object Identifier
doi:10.1214/17-AAP1317

Mathematical Reviews number (MathSciNet)
MR3784489

Zentralblatt MATH identifier
1356.60065

#### Citation

Del Moral, P.; Tugaut, J. On the stability and the uniform propagation of chaos properties of Ensemble Kalman–Bucy filters. Ann. Appl. Probab. 28 (2018), no. 2, 790--850. doi:10.1214/17-AAP1317. https://projecteuclid.org/euclid.aoap/1523433625

#### References

• [1] Abou-Kandil, H., Freiling, G., Ionescu, V. and Jank, G. (2003). Matrix Riccati Equations in Control and Systems Theory. Systems & Control: Foundations & Applications. Birkhäuser, Basel.
• [2] Allen, J. I., Eknes, M. and Evensen, G. (2002). An Ensemble Kalman filter with a complex marine ecosystem model: Hindcasting phytoplankton in the Cretan Sea. Ann. Geophys. 20 1–13.
• [3] Anderson, B. D. O. (1971). Stability properties of Kalman–Bucy filters. J. Franklin Inst. 291 137–144.
• [4] Anderson, J. L. (2001). An ensemble adjustment Kalman filter for data assimilation. Mon. Weather Rev. 129 2884–2903.
• [5] Anderson, J. L. (2003). A local least squares framework for ensemble filtering. Mon. Weather Rev. 131 634–642.
• [6] Baras, J. S., Bensoussan, A. and James, M. R. (1988). Dynamic observers as asymptotic limits of recursive filters: Special cases. SIAM J. Appl. Math. 48 1147–1158.
• [7] Bernstein, D. S. (1988). Inequalities for the trace of matrix exponentials. SIAM J. Matrix Anal. Appl. 9 156–158.
• [8] Berry, T. and Harlim, J. (2014). Linear theory for filtering nonlinear multiscale systems with model error. Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. 470 20140168, 25.
• [9] Bishop, A. N. and Del Moral, P. (2016). On the stability of Kalman–Bucy diffusion processes. Available at arXiv:1610.04686.
• [10] Bittanti, S., Laub, A. J. and Willems, J. C. (1991). The Riccati Equation. Communications and Control Engineering Series. Springer, Berlin.
• [11] Bolley, F., Guillin, A. and Malrieu, F. (2010). Trend to equilibrium and particle approximation for a weakly selfconsistent Vlasov–Fokker–Planck equation. M2AN Math. Model. Numer. Anal. 44 867–884.
• [12] Bougerol, P. and Fakhfakh, S. (1996). A note on the stability of the Kalman–Bucy filter with randomly time-varying parameters. J. Math. Sci. 78 28–33.
• [13] Brockett, R. W. (2015). Finite Dimensional Linear Systems. Classics in Applied Mathematics 74. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA. Reprint of the 1970 original.
• [14] Bucy, R. S. (1965). Nonlinear filtering theory. IEEE Trans. Automat. Control 10 198–198.
• [15] Bucy, R. S. (1967). Global theory of the Riccati equation. J. Comput. System Sci. 1 349–361.
• [16] Burgers, G., van Leeuwen, P. J. and Evensen, G. (1998). Analysis scheme in the ensemble Kalman filter. Mon. Weather Rev. 126 1719–1724.
• [17] Cattiaux, P., Guillin, A. and Malrieu, F. (2008). Probabilistic approach for granular media equations in the non-uniformly convex case. Probab. Theory Related Fields 140 19–40.
• [18] Coppel, W. A. (1978). Dichotomies in Stability Theory. Lecture Notes in Mathematics 629. Springer, Berlin–New York.
• [19] Costa, E. F. (2008). On the stability of the recursive Kalman filter for linear time-invariant systems. In Proceedings of the IEEE 2008 American Control Conference 1286–1291, Seattle, WA.
• [20] Del Moral, P. (2004). Feynman–Kac Formulae. Genealogical and Interacting Particle Systems with Applications. Probability and Its Applications (New York). Springer, New York.
• [21] Del Moral, P. (2013). Mean Field Simulation for Monte Carlo Integration. Monographs on Statistics and Applied Probability 126. CRC Press, Boca Raton, FL.
• [22] Del Moral, P. and Tugaut, J. (2013). Uniform propagation of chaos and creation of chaos for a class of nonlinear diffusions. https://hal.archives-ouvertes.fr/hal-00798813.
• [23] Dyda, B. and Tugaut, J. (2017). Exponential rate of convergence independent of the dimension in a mean-field system of particles. Probab. Math. Statist. 37 145–161.
• [24] Einicke, G. (2012). Continuous-time minimum-variance filtering. In Smoothing, Filtering and Prediction—Estimating the Past, Present and Future. InTech, London.
• [25] Eknes, M. and Evensen, G. (2002). An Ensemble Kalman filter with a 1-D marine ecosystem model. JMS 36 75–100.
• [26] Evensen, G. (1994). Sequential data assimilation with a non-linear quasi-geostrophic model using Monte Carlo methods to forecast error statistics. J. Geophys. Res. 99 10143–10162.
• [27] Evensen, G., Hove, J., Meisingset, H. C., Reiso, E. and Seim, K. S. (2007). Using the EnKF for assisted history matching of a North Sea reservoir model. SPE 106184.
• [28] Givens, C. R. and Shortt, R. M. (1984). A class of Wasserstein metrics for probability distributions. Michigan Math. J. 31 231–240.
• [29] Gottwald, G. and Majda, A. J. (2013). A mechanism for catastrophic filter divergence in data assimilation for sparse observation networks. Nonlinear Process. Geophys. 20 705–712.
• [30] Harlim, J. and Hunt, B. (2005). Local Ensemble Transform Kalman Filter: An Efficient Scheme for Assimilating Atmospheric Data. Preprint. Available at https://www.atmos.umd.edu/ekalnay/pubs/harlim_hunt05.pdf.
• [31] Harlim, J. and Majda, A. J. (2010). Catastrophic filter divergence in filtering nonlinear dissipative systems. Commun. Math. Sci. 8 27–43.
• [32] Houtekamer, P. and Mitchell, H. L. (1998). Data assimilation using an ensemble Kalman filter technique. Mon. Weather Rev. 126 796–811.
• [33] Ilchmann, A., Owens, D. H. and Prätzel-Wolters, D. (1987). Sufficient conditions for stability of linear time-varying systems. Systems Control Lett. 9 157–163.
• [34] Johns, C. J. and Mandel, J. (2008). A two-stage ensemble Kalman filter for smooth data assimilation. Environ. Ecol. Stat. 15 101–110.
• [35] Kalnay, E. (2003). Atmospheric Modeling, Data Assimilation, and Predictability. Cambridge Univ. Press, Cambridge.
• [36] Kelly, D., Majda, A. J. and Tong, X. T. (2015). Concrete ensemble Kalman filters with rigorous catastrophic filter divergence. Proc. Natl. Acad. Sci. USA 112 10589–10594.
• [37] Krause, J. M. and Kumar, K. S. P. (1986). An alternate stability analysis framework for adaptive control. Systems Control Lett. 7 19–24.
• [38] Kreisselmeier, G. (1985). An approach to stable indirect adaptive control. Automatica J. IFAC 21 425–431.
• [39] Kushner, H. J. (1964). On the differential equations satisfied by conditional probablitity densities of Markov processes, with applications. J. Soc. Indust. Appl. Math. Ser. A Control 2 106–119.
• [40] Kwakernaak, H. and Sivan, R. (1972). Linear Optimal Control Systems. Wiley, New York.
• [41] Lancaster, P. and Rodman, L. (1995). Algebraic Riccati Equations. Oxford Science Publications. The Clarendon Press, New York.
• [42] Lawson, J. and Lim, Y. (2007). A Birkhoff contraction formula with applications to Riccati equations. SIAM J. Control Optim. 46 930–951.
• [43] Le Gland, F., Monbet, V. and Tran, V.-D. (2011). Large sample asymptotics for the ensemble Kalman filter. In The Oxford Handbook of Nonlinear Filtering 598–631. Oxford Univ. Press, Oxford.
• [44] Lisaeter, K. A., Rosanova, J. and Evensen, G. (2003). Assimilation of ice concentration in a coupled ice-ocean model, using the Ensemble Kalman filter. Ocean Dyn. 53 368–388.
• [45] Luenberger, D. G. (1966). Observers for multivariable systems. IEEE Trans. Automat. Control AC-11 190–199.
• [46] Majda, A. J. and Harlim, J. (2012). Filtering Complex Turbulent Systems. Cambridge Univ. Press, Cambridge.
• [47] Malrieu, F. (2001). Logarithmic Sobolev inequalities for some nonlinear PDE’s. Stochastic Process. Appl. 95 109–132.
• [48] Mandel, J., Cobb, L. and Beezley, J. D. (2011). On the convergence of the ensemble Kalman filter. Appl. Math. 56 533–541.
• [49] McKean, H. P. Jr. (1966). A class of Markov processes associated with nonlinear parabolic equations. Proc. Natl. Acad. Sci. USA 56 1907–1911.
• [50] Méléard, S. (1996). Asymptotic behaviour of some interacting particle systems; McKean–Vlasov and Boltzmann models. In Probabilistic Models for Nonlinear Partial Differential Equations (Montecatini Terme, 1995). Lecture Notes in Math. 1627 42–95. Springer, Berlin.
• [51] Moler, C. and Van Loan, C. (2003). Nineteen dubious ways to compute the exponential of a matrix, twenty-five years later. SIAM Rev. 45 3–49.
• [52] Naevdal, G., Johnsen, L. M., Aanonsen, S. I. and Vefring, E. H. (2005). Reservoir monitoring and continuous model updating using ensemble Kalman filter. SPE Journal 10 66–74.
• [53] Ocone, D. and Pardoux, E. (1996). Asymptotic stability of the optimal filter with respect to its initial condition. SIAM J. Control Optim. 34 226–243.
• [54] Ott, E., Hunt, B. R., Szunyogh, I., Zimin, A. V., Kostelich, E. J., Corazza, M., Kalnay, E., Patil, D. and Yorke, J. A. (2004). A local ensemble Kalman filter for atmospheric data assimilation. Tellus, Ser. A Dyn. Meteorol. Oceanogr. 56 415–428.
• [55] Poubelle, M.-A., Petersen, I. R., Gevers, M. R. and Bitmead, R. R. (1986). A miscellany of results on an equation of Count J. F. Riccati. IEEE Trans. Automat. Control 31 651–654.
• [56] Rosenbrock, H. H. (1963). The stability of linear time-dependent control systems. J. Electron. Control 15 73–80.
• [57] Seiler, A., Evensen, G., Skjervheim, J.-A., Hove, J. and Vabø, J. G. (2011). Using the ensemble Kalman filter for history matching and uncertainty quantification of complex reservoir models. In Large-Scale Inverse Problems and Quantification of Uncertainty. Wiley Ser. Comput. Stat. 247–271. Wiley, Chichester.
• [58] Skjervheim, J.-A., Evensen, G., Aanonsen, S. I., Ruud, B. O. and Johansen, T. A. (2005). Incorporating 4D seismic data in reservoir simulation models using ensemble Kalman filter. SPE 12 95789.
• [59] Snyder, D. A. (2016). On the Relation of Schatten Norms and the Thompson Metric. Available at ArXiv:1608.03301.
• [60] Sontag, E. D. (1998). Mathematical Control Theory: Deterministic Finite-Dimensional Systems, 2nd ed. Texts in Applied Mathematics 6. Springer, New York.
• [61] Stratonovič, R. L. (1960). Conditional Markov processes. Teor. Verojatnost. i Primenen. 5 172–195.
• [62] Su, J., Li, B. and Chen, W.-H. (2015). On existence, optimality and asymptotic stability of the Kalman filter with partially observed inputs. Automatica J. IFAC 53 149–154.
• [63] Sznitman, A.-S. (1991). Topics in propagation of chaos. In École D’Été de Probabilités de Saint-Flour XIX—1989. Lecture Notes in Math. 1464 165–251. Springer, Berlin.
• [64] Tong, X. T., Majda, A. J. and Kelly, D. (2016). Nonlinear stability and ergodicity of ensemble based Kalman filters. Nonlinearity 29 657–691.
• [65] Van Loan, C. (1977). The sensitivity of the matrix exponential. SIAM J. Numer. Anal. 14 971–981.
• [66] Veselić, K. (2011). Damped Oscillations of Linear Systems: A Mathematical Introduction. Lecture Notes in Math. 2023. Springer, Heidelberg.
• [67] Wen, X.-H. and Chen, W. H. (2005). Real-time reservoir model updating using ensemble Kalman filter SPE-92991-MS. In SPE Reservoir Simulation Symposium, the Woodlands, Texas.
• [68] Wonham, W. M. (1968). On a matrix Riccati equation of stochastic control. SIAM J. Control 6 681–697.
• [69] Yang, X. (2000). A matrix trace inequality. J. Math. Anal. Appl. 250 372–374.
• [70] Yang, X. M., Yang, X. Q. and Teo, K. L. (2001). A matrix trace inequality. J. Math. Anal. Appl. 263 327–331.
• [71] Zakai, M. (1969). On the optimal filtering of diffusion processes. Z. Wahrsch. Verw. Gebiete 11 230.
• [72] Zelikin, M. I. (1991). On the theory of the matrix Riccati equation Mat. Sb. 182 970–984. Available at http://www.mathnet.ru/links/5052aa2d9c921f105d61c94a204912c1/sm1336.pdf.
• [73] Ziebur, A. D. (1970). On determining the structure of $A$ by analyzing $e^{At}$. SIAM Rev. 12 98–102.