Dynamically learning the parameters of a chaotic system using partial observations

Motivated by recent progress in data assimilation, we develop an algorithm to dynamically learn the parameters of a chaotic system from partial observations. Under reasonable assumptions, we rigorously establish the convergence of this algorithm to the correct parameters when the system in question is the classic three-dimensional Lorenz system. Computationally, we demonstrate the efficacy of this algorithm on the Lorenz system by recovering any proper subset of the three non-dimensional parameters of the system, so long as a corresponding subset of the state is observable. We also provide computational evidence that this algorithm works well beyond the hypotheses required in the rigorous analysis, including in the presence of noisy observations, stochastic forcing, and the case where the observations are discrete and sparse in time.

[1]  Animikh Biswas,et al.  Continuous Data Assimilation for the Three Dimensional Navier-Stokes Equations , 2020, SIAM J. Math. Anal..

[2]  Adam Larios,et al.  Approximate continuous data assimilation of the 2D Navier-Stokes equations via the Voigt-regularization with observable data , 2018, Evolution Equations & Control Theory.

[3]  Charles R. Doering,et al.  On the shape and dimension of the Lorenz attractor , 1995 .

[4]  Edriss S. Titi,et al.  A Computational Study of a Data Assimilation Algorithm for the Two-dimensional Navier-Stokes Equations , 2015, 1505.01234.

[5]  Jiguo Cao,et al.  Parameter Estimation of Partial Differential Equation Models , 2013, Journal of the American Statistical Association.

[6]  James C. Robinson.  Infinite-Dimensional Dynamical Systems: An Introduction to Dissipative Parabolic PDEs and the Theory of Global Attractors , 2001 .

[7]  Patrick Gallinari,et al.  Learning Dynamical Systems from Partial Observations , 2019, ArXiv.

[8]  Edriss S. Titi,et al.  Uniform-in-Time Error Estimates for the Postprocessing Galerkin Method Applied to a Data Assimilation Algorithm , 2018, SIAM J. Numer. Anal..

[9]  Leo G. Rebholz,et al.  Global in time stability and accuracy of IMEX-FEM data assimilation schemes for Navier–Stokes equations , 2018, Computer Methods in Applied Mechanics and Engineering.

[10]  Holger R. Dullin,et al.  Extended Phase Diagram of the Lorenz Model , 2005, Int. J. Bifurc. Chaos.

[11]  J. Hoke,et al.  The Initialization of Numerical Models by a Dynamic-Initialization Technique , 1976 .

[12]  Benjamin Pachev,et al.  Concurrent multi-parameter learning demonstrated on the Kuramoto-Sivashinsky equation , 2021, ArXiv.

[13]  Hakima Bessaih,et al.  Continuous data assimilation with stochastically noisy data , 2014, 1406.1533.

[14]  Traian Iliescu,et al.  Continuous data assimilation reduced order models of fluid flow , 2019, Computer Methods in Applied Mechanics and Engineering.

[15]  E Weinan,et al.  Model Reduction with Memory and the Machine Learning of Dynamical Systems , 2018, Communications in Computational Physics.

[16]  A. M. Stuart,et al.  Accuracy and stability of the continuous-time 3DVAR filter for the Navier–Stokes equation , 2012, 1210.1594.

[17]  A. Stuart,et al.  Analysis of the 3DVAR filter for the partially observed Lorenz'63 model , 2012, 1212.4923.

[18]  Charles R. Doering,et al.  Maximal transport in the Lorenz equations , 2014 .

[19]  Animikh Biswas,et al.  Higher-order synchronization for a data assimilation algorithm for the 2D Navier–Stokes equations , 2017 .

[20]  A. Mazzino,et al.  Inferring flow parameters and turbulent configuration with physics-informed data assimilation and spectral nudging , 2018, Physical Review Fluids.

[21]  Igor Kukavica,et al.  The Lorenz equation as a metaphor for the Navier-Stokes equations , 2001 .

[22]  Leo G. Rebholz,et al.  Continuous data assimilation applied to a velocity-vorticity formulation of the 2D Navier-Stokes equations , 2020, Electronic Research Archive.

[23]  Edriss S. Titi,et al.  Uniform in Time Error Estimates for a Finite Element Method Applied to a Downscaling Data Assimilation Algorithm for the Navier-Stokes Equations , 2018, SIAM J. Numer. Anal..

[24]  Eran Treister,et al.  jInv-a Flexible Julia Package for PDE Parameter Estimation , 2016, SIAM J. Sci. Comput..

[25]  Ming-Cheng Shiue,et al.  Analysis and computation of continuous data assimilation algorithms for Lorenz 63 system based on nonlinear nudging techniques , 2021, J. Comput. Appl. Math..

[26]  O. Knio,et al.  Efficient dynamical downscaling of general circulation models using continuous data assimilation , 2019, Quarterly Journal of the Royal Meteorological Society.

[27]  Leo G. Rebholz,et al.  Simple and efficient continuous data assimilation of evolution equations via algebraic nudging , 2018, Numerical Methods for Partial Differential Equations.

[28]  M. U. Altaf,et al.  Downscaling the 2D Bénard convection equations using continuous data assimilation , 2015, Computational Geosciences.

[29]  Rudolph van der Merwe,et al.  The square-root unscented Kalman filter for state and parameter-estimation , 2001, 2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.01CH37221).

[30]  Edriss S. Titi,et al.  Continuous Data Assimilation for a 2D Bénard Convection System Through Horizontal Velocity Measurements Alone , 2016, J. Nonlinear Sci..

[31]  Edriss S. Titi,et al.  Data Assimilation algorithm for 3D B\'enard convection in porous media employing only temperature measurements , 2015, 1506.08678.

[32]  L. Biferale,et al.  Synchronization to Big Data: Nudging the Navier-Stokes Equations for Data Assimilation of Turbulent Flows , 2019, Physical Review X.

[33]  E. Lunasin,et al.  An efficient continuous data assimilation algorithm for the Sabra shell model of turbulence. , 2021, Chaos.

[34]  Adam Larios,et al.  Sensitivity Analysis for the 2D Navier–Stokes Equations with Applications to Continuous Data Assimilation , 2020, Journal of Nonlinear Science.

[35]  Alan C. Hindmarsh,et al.  Description and use of LSODE, the Livermore Solver for Ordinary Differential Equations , 1993 .

[36]  Edriss S. Titi,et al.  Continuous data assimilation for the three-dimensional Navier-Stokes-α model , 2016, Asymptot. Anal..

[37]  E. Titi,et al.  On the Charney Conjecture of Data Assimilation Employing Temperature Measurements Alone: The Paradigm of 3D Planetary Geostrophic Model , 2016, 1608.04770.

[38]  Animikh Biswas,et al.  Data Assimilation for the Navier-Stokes Equations Using Local Observables , 2020, SIAM J. Appl. Dyn. Syst..

[39]  E. Titi,et al.  Determining modes and Grashof number in 2D turbulence: a numerical case study , 2008 .

[40]  Ling Xu,et al.  Application of the Newton iteration algorithm to the parameter estimation for dynamical systems , 2015, J. Comput. Appl. Math..

[41]  Dylan J. Foster,et al.  Learning nonlinear dynamical systems from a single trajectory , 2020, L4DC.

[42]  M. Jolly,et al.  Continuous Data Assimilation with Blurred-in-Time Measurements of the Surface Quasi-Geostrophic Equation , 2018, Chinese Annals of Mathematics, Series B.

[43]  Yuan Pei,et al.  Continuous data assimilation for the 2D magnetohydrodynamic equations using one component of the velocity and magnetic fields , 2017, Asymptot. Anal..

[44]  D. Stensrud,et al.  The Ensemble Kalman Filter for Combined State and Parameter Estimation , 2009 .

[45]  Edriss S. Titi,et al.  Abridged Continuous Data Assimilation for the 2D Navier–Stokes Equations Utilizing Measurements of Only One Component of the Velocity Field , 2015, 1504.05978.

[46]  Edriss S. Titi,et al.  Downscaling data assimilation algorithm with applications to statistical solutions of the Navier-Stokes equations , 2017 .

[47]  Benjamin Peherstorfer,et al.  Lift & Learn: Physics-informed machine learning for large-scale nonlinear dynamical systems , 2019, Physica D: Nonlinear Phenomena.

[48]  Eric Olson,et al.  Data assimilation using noisy time-averaged measurements , 2017, Physica D: Nonlinear Phenomena.

[49]  Edriss S. Titi,et al.  Continuous Data Assimilation Using General Interpolant Observables , 2013, J. Nonlinear Sci..

[50]  Emine Celik,et al.  Spectral Filtering of Interpolant Observables for a Discrete-in-Time Downscaling Data Assimilation Algorithm , 2018, SIAM J. Appl. Dyn. Syst..

[51]  Zhenpo Wang,et al.  State and parameter estimation based on a modified particle filter for an in-wheel-motor-drive electric vehicle , 2019, Mechanism and Machine Theory.

[52]  Feng Ding,et al.  Gradient-Based Iterative Parameter Estimation Algorithms for Dynamical Systems from Observation Data , 2019, Mathematics.

[53]  Edriss S. Titi,et al.  A Discrete Data Assimilation Scheme for the Solutions of the Two-Dimensional Navier-Stokes Equations and Their Statistics , 2016, SIAM J. Appl. Dyn. Syst..

[54]  E. Lorenz.  Deterministic nonperiodic flow , 1963 .

[55]  C. Doering,et al.  Applied analysis of the Navier-Stokes equations: Index , 1995 .

[56]  Edriss S. Titi,et al.  Discrete data assimilation in the Lorenz and 2D Navier–Stokes equations , 2010, 1010.6105.

[57]  A. Stuart,et al.  The Bayesian Approach to Inverse Problems , 2013, 1302.6989.

[58]  Yuan Pei.  Continuous data assimilation for the 3D primitive equations of the ocean , 2018, Communications on Pure & Applied Analysis.

[59]  Edriss S. Titi,et al.  A Data Assimilation Algorithm for the Subcritical Surface Quasi-Geostrophic Equation , 2016, 1607.08574.

[60]  Edriss S. Titi,et al.  Finite determining parameters feedback control for distributed nonlinear dissipative systems - a computational study , 2015, 1506.03709.

[61]  Adam Larios,et al.  Continuous Data Assimilation with a Moving Cluster of Data Points for a Reaction Diffusion Equation: A Computational Study , 2018, Communications in Computational Physics.

[62]  Cecilia F. Mondaini,et al.  Fully discrete numerical schemes of a data assimilation algorithm: uniform-in-time error estimates , 2019, IMA Journal of Numerical Analysis.

[63]  Louis J. Durlofsky,et al.  Error modeling for surrogates of dynamical systems using machine learning , 2017 .

[64]  Yuan Pei,et al.  Nonlinear continuous data assimilation , 2017, Evolution Equations and Control Theory.

[65]  N. Glatt-Holtz,et al.  Parameter estimation for the stochastically perturbed Navier-Stokes equations , 2010, 1006.1952.

[66]  Roberto Barrio,et al.  A three-parametric study of the Lorenz model , 2007 .

[67]  Didier Georges,et al.  State and parameter estimation in 1-D hyperbolic PDEs based on an adjoint method , 2016, Autom..

[68]  J. Nathan Kutz,et al.  Deep learning in fluid dynamics , 2017, Journal of Fluid Mechanics.

[69]  J. S. Wettlaufer,et al.  Maximal Stochastic Transport in the Lorenz Equations , 2015, 1508.03665.

[70]  Leo G. Rebholz,et al.  Continuous data assimilation and long-time accuracy in a C0 interior penalty method for the Cahn-Hilliard equation , 2021, Appl. Math. Comput..

[71]  Michael A. Demetriou,et al.  On-Line Parameter Estimation for Infinite-Dimensional Dynamical Systems , 1997 .

[72]  A. Farhat,et al.  Data Assimilation in Large Prandtl Rayleigh-Bénard Convection from Thermal Measurements , 2019, SIAM J. Appl. Dyn. Syst..

[73]  Edriss S. Titi,et al.  Continuous data assimilation for the three-dimensional Brinkman–Forchheimer-extended Darcy model , 2015, 1502.00964.

[74]  Edriss S. Titi,et al.  Continuous data assimilation for the 2D Bénard convection through velocity measurements alone , 2014, 1410.1767.

[75]  Fabian J. Theis,et al.  Data2Dynamics: a modeling environment tailored to parameter estimation in dynamical systems , 2015, Bioinform..

[76]  Adam Larios,et al.  Parameter Recovery for the 2 Dimensional Navier-Stokes Equations via Continuous Data Assimilation , 2020, SIAM J. Sci. Comput..

[77]  Adam Larios,et al.  The Bleeps, the Sweeps, and the Creeps: Convergence Rates for Dynamic Observer Patterns via Data Assimilation for the 2D Navier-Stokes Equations , 2021, Computer Methods in Applied Mechanics and Engineering.

[78]  Julia Novo,et al.  Error analysis of fully discrete mixed finite element data assimilation schemes for the Navier-Stokes equations , 2019, Advances in Computational Mathematics.