# American Institute of Mathematical Sciences

March  2020, 2(1): 55-80. doi: 10.3934/fods.2020004

## Bayesian inference of chaotic dynamics by merging data assimilation, machine learning and expectation-maximization

 1 CEREA, joint laboratory École des Ponts ParisTech and EDF R & D, Université Paris-Est, Champs-sur-Marne, France 2 Nansen Environmental and Remote Sensing Center, Bergen, Norway, and Sorbonne University, CNRS-IRD-MNHN, LOCEAN, Paris, France 3 Departement of Meteorology, University of Reading and NCEO, United-Kingdom, and Mathematical Institute, Utrecht University, The Netherlands 4 Nansen Environmental and Remote Sensing Center, Bergen, Norway

* Corresponding author: Marc Bocquet

Published  March 2020

The reconstruction from observations of high-dimensional chaotic dynamics such as geophysical flows is hampered by (ⅰ) the partial and noisy observations that can realistically be obtained, (ⅱ) the need to learn from long time series of data, and (ⅲ) the unstable nature of the dynamics. To achieve such inference from the observations over long time series, it has been suggested to combine data assimilation and machine learning in several ways. We show how to unify these approaches from a Bayesian perspective using expectation-maximization and coordinate descents. In doing so, the model, the state trajectory and model error statistics are estimated all together. Implementations and approximations of these methods are discussed. Finally, we numerically and successfully test the approach on two relevant low-order chaotic models with distinct identifiability.

Citation: Marc Bocquet, Julien Brajard, Alberto Carrassi, Laurent Bertino. Bayesian inference of chaotic dynamics by merging data assimilation, machine learning and expectation-maximization. Foundations of Data Science, 2020, 2 (1) : 55-80. doi: 10.3934/fods.2020004
##### References:

show all references

##### References:
From top to bottom: representation of the flow rate $\boldsymbol \phi_ \mathbf{A}$ with a NN, integration of the flow rate into $\mathbf{f}_ \mathbf{A}$ using an explicit integration scheme (here a second-order Runge Kutta scheme), and ${N_\mathrm{c}}-$fold composition up to the full resolvent $\mathbf{F}_ \mathbf{A}$. $\delta t$ is the integration time step corresponding to the resolvent $\mathbf{f}_ \mathbf{A}$
On the left hand side: Properties of the surrogate model obtained from full but noisy observation of the L96 model in the nominal configuration ($L = 4$, $K = 5000$, $\sigma_y = 1$, ${N_\mathrm{y}} = {N_\mathrm{x}} = 40$). On the right hand side: Properties of the surrogate model obtained from full but noisy observation of the L05Ⅲ model in the nominal configuration ($L = 4$, $K = 5000$, $\sigma_y = 1$, ${N_\mathrm{y}} = {N_\mathrm{x}} = 36$). From top to bottom, are plotted the FS (NRMSE as a function of lead time in Lyapunov time), the LS (all exponents), and the PSD (in log-log-scale). A total of $10$ experiments have been performed for both configurations. The curves corresponding to each member are drawn with thin blue lines while the mean of each indicator over the ensemble are drawn in thick dashed orange line
Same as Figure 2 but for several values of the training window length $K$. Each curve is the mean over $10$ experiments with different sets of observations. The LS and PSD of the reference models are also plotted for comparison
On the left hand side: Properties of the surrogate model obtained from full but noisy observation of the L96 model in the nominal configuration ($L = 4$, $K = 5000$, ${N_\mathrm{y}} = {N_\mathrm{x}} = 40$ and with several $\sigma_y$). On the right hand side: Properties of the surrogate model obtained from full but noisy observation of the L05Ⅲ model in the nominal configuration ($L = 4$, $K = 5000$, $\sigma_y = 1$, ${N_\mathrm{y}} = {N_\mathrm{x}} = 36$ and with several $\sigma_y$). From top to bottom, are plotted the FS (NRMSE as a function of lead time in Lyapunov time) and the PSD (in log-log-scale), averaged over an ensemble of $10$ samples
On the left hand side: Properties of the surrogate model obtained from partial and noisy observation of the L96 model in the nominal configuration ($L = 4$, $K = 5000$, $\sigma_y = 1$, ${N_\mathrm{x}} = 40$) where ${N_\mathrm{y}}$ is varied. On the right hand side: Properties of the surrogate model obtained from partial and noisy observation of the L05Ⅲ model in the nominal configuration ($L = 4$, $K = 5000$, $\sigma_y = 1$, ${N_\mathrm{x}} = 36$) where ${N_\mathrm{y}}$ is varied. From top to bottom, are plotted the mean FS (NRMSE as a function of lead time in Lyapunov time), the mean LS (all exponents), and the mean PSD (in log-log-scale). A total of $10$ experiments have been performed for both configurations
Scalar indicators for nominal experiments based on L96 and L05Ⅲ. Key hyperparameters are recalled. The statistics of the indicators are obtained over $10$ samples
 Model ${N_\mathrm{y}}$ $\sigma_y$ $K$ $L$ $\pi_ \frac{1}{2}$ $\sigma_q$ $\lambda_1$ L96 $40$ $1$ $5000$ $4$ $4.56 \pm 0.06$ $0.08790 \pm 2\, 10^{-5}$ $1.66 \pm 0.02$ L05Ⅲ $36$ $1$ $5000$ $4$ $4.06 \pm 0.21$ $0.07720 \pm 2\, 10^{-5}$ $1.03 \pm 0.05$
 Model ${N_\mathrm{y}}$ $\sigma_y$ $K$ $L$ $\pi_ \frac{1}{2}$ $\sigma_q$ $\lambda_1$ L96 $40$ $1$ $5000$ $4$ $4.56 \pm 0.06$ $0.08790 \pm 2\, 10^{-5}$ $1.66 \pm 0.02$ L05Ⅲ $36$ $1$ $5000$ $4$ $4.06 \pm 0.21$ $0.07720 \pm 2\, 10^{-5}$ $1.03 \pm 0.05$
Scalar indicators for L96 and L05Ⅲ in their nominal configuration, using either the full or the approximate schemes. The statistics of the indicators are obtained over $10$ samples
 Model Scheme $\pi_ \frac{1}{2}$ $\sigma_q$ $\lambda_1$ L96 Approximate $4.56 \pm 0.06$ $0.08790 \pm 2\, 10^{-5}$ $1.66 \pm 0.02$ L96 Full $4.24 \pm 0.07$ $0.09152$ $1.66 \pm 0.02$ L05Ⅲ Approximate $4.06 \pm 0.21$ $0.07720 \pm 2\, 10^{-5}$ $1.03 \pm 0.05$ L05Ⅲ Full $3.97 \pm 0.17$ $0.08024$ $1.03 \pm 0.04$
 Model Scheme $\pi_ \frac{1}{2}$ $\sigma_q$ $\lambda_1$ L96 Approximate $4.56 \pm 0.06$ $0.08790 \pm 2\, 10^{-5}$ $1.66 \pm 0.02$ L96 Full $4.24 \pm 0.07$ $0.09152$ $1.66 \pm 0.02$ L05Ⅲ Approximate $4.06 \pm 0.21$ $0.07720 \pm 2\, 10^{-5}$ $1.03 \pm 0.05$ L05Ⅲ Full $3.97 \pm 0.17$ $0.08024$ $1.03 \pm 0.04$
 [1] Xin Guo, Lexin Li, Qiang Wu. Modeling interactive components by coordinate kernel polynomial models. Mathematical Foundations of Computing, 2020, 3 (4) : 263-277. doi: 10.3934/mfc.2020010 [2] Ying Lin, Qi Ye. Support vector machine classifiers by non-Euclidean margins. Mathematical Foundations of Computing, 2020, 3 (4) : 279-300. doi: 10.3934/mfc.2020018 [3] Aihua Fan, Jörg Schmeling, Weixiao Shen. $L^\infty$-estimation of generalized Thue-Morse trigonometric polynomials and ergodic maximization. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 297-327. doi: 10.3934/dcds.2020363 [4] Xin Guo, Lei Shi. Preface of the special issue on analysis in data science: Methods and applications. Mathematical Foundations of Computing, 2020, 3 (4) : i-ii. doi: 10.3934/mfc.2020026 [5] Marion Darbas, Jérémy Heleine, Stephanie Lohrengel. Numerical resolution by the quasi-reversibility method of a data completion problem for Maxwell's equations. Inverse Problems & Imaging, 2020, 14 (6) : 1107-1133. doi: 10.3934/ipi.2020056 [6] Dan Zhu, Rosemary A. Renaut, Hongwei Li, Tianyou Liu. Fast non-convex low-rank matrix decomposition for separation of potential field data using minimal memory. Inverse Problems & Imaging, , () : -. doi: 10.3934/ipi.2020076 [7] Peizhao Yu, Guoshan Zhang, Yi Zhang. Decoupling of cubic polynomial matrix systems. Numerical Algebra, Control & Optimization, 2021, 11 (1) : 13-26. doi: 10.3934/naco.2020012 [8] Ilyasse Lamrani, Imad El Harraki, Ali Boutoulout, Fatima-Zahrae El Alaoui. Feedback stabilization of bilinear coupled hyperbolic systems. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020434 [9] Felix Finster, Jürg Fröhlich, Marco Oppio, Claudio F. Paganini. Causal fermion systems and the ETH approach to quantum theory. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020451 [10] Xiyou Cheng, Zhitao Zhang. Structure of positive solutions to a class of Schrödinger systems. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020461 [11] Simon Hochgerner. Symmetry actuated closed-loop Hamiltonian systems. Journal of Geometric Mechanics, 2020, 12 (4) : 641-669. doi: 10.3934/jgm.2020030 [12] Javier Fernández, Cora Tori, Marcela Zuccalli. Lagrangian reduction of nonholonomic discrete mechanical systems by stages. Journal of Geometric Mechanics, 2020, 12 (4) : 607-639. doi: 10.3934/jgm.2020029 [13] Yuri Fedorov, Božidar Jovanović. Continuous and discrete Neumann systems on Stiefel varieties as matrix generalizations of the Jacobi–Mumford systems. Discrete & Continuous Dynamical Systems - A, 2020  doi: 10.3934/dcds.2020375 [14] João Marcos do Ó, Bruno Ribeiro, Bernhard Ruf. Hamiltonian elliptic systems in dimension two with arbitrary and double exponential growth conditions. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 277-296. doi: 10.3934/dcds.2020138 [15] Awais Younus, Zoubia Dastgeer, Nudrat Ishaq, Abdul Ghaffar, Kottakkaran Sooppy Nisar, Devendra Kumar. On the observability of conformable linear time-invariant control systems. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020444 [16] Shiqi Ma. On recent progress of single-realization recoveries of random Schrödinger systems. Electronic Research Archive, , () : -. doi: 10.3934/era.2020121 [17] Sergey Rashkovskiy. Hamilton-Jacobi theory for Hamiltonian and non-Hamiltonian systems. Journal of Geometric Mechanics, 2020, 12 (4) : 563-583. doi: 10.3934/jgm.2020024 [18] Touria Karite, Ali Boutoulout. Global and regional constrained controllability for distributed parabolic linear systems: RHum approach. Numerical Algebra, Control & Optimization, 2020  doi: 10.3934/naco.2020055 [19] Maoding Zhen, Binlin Zhang, Vicenţiu D. Rădulescu. Normalized solutions for nonlinear coupled fractional systems: Low and high perturbations in the attractive case. Discrete & Continuous Dynamical Systems - A, 2020  doi: 10.3934/dcds.2020379 [20] Zedong Yang, Guotao Wang, Ravi P. Agarwal, Haiyong Xu. Existence and nonexistence of entire positive radial solutions for a class of Schrödinger elliptic systems involving a nonlinear operator. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020436

Impact Factor: