Stochastic Collocation Methods for Polynomial Chaos: Analysis and Applications

Size: px
Start display at page:

Download "Stochastic Collocation Methods for Polynomial Chaos: Analysis and Applications"

Transcription

1 Stochastic Collocation Methods for Polynomial Chaos: Analysis and Applications Dongbin Xiu Department of Mathematics, Purdue University Support: AFOSR FA (Computational Math) SF CAREER DMS (Computational Math) DOE DE-FC52-8A28617 (PSAAP)

2 Overview Generalized polynomial chaos Stochastic collocation Lagrange interpolation Pseudo spectral gpc Applications Bayesian inverse problem Data assimilation

3 Stochastic PDE: Uncertainty Quantification via gpc u nz (, txz, ) = L( u ), (, T ] D R t B( u) =, [, T] D R u = u( x, Z), { t = } D R th -order gpc expansion: nz + u (,, ) ˆ t x Z uk(, t x) Φ k( Z), # of basis= n k = uˆ = E[ u( Z) Φ ( Z)] = u( Z) Φ ( Z ) ρ( Z ) dz,, k k k k n n z z z Orthogonal basis: E Z ( Z) ( Z) Φ Φ Φ ( Z) Φ ( Z) ρ( Z) dz = δ i j i j ij Optimality: u u = inf nz u Ψ L 2 2 ρ Z Ψ Π L ρ Z ( ) ( )

4 Generalized Polynomial Chaos (gpc) Basis functions: Hermite polynomials: seminal work by R. Ghanem Global orthogonal polynomials (Xiu & Karniadakis, 2) Wavelet basis (Le Maitre et al, 4) Piecewise basis (Babuska et al 4, Wan & Karniadakis, 5) Implementations: Stochastic Galerkin Stochastic collocation Properties: Rigorous mathematics High accuracy, fast covergence Curse-of-dimensionality

5 Stochastic Collocation Collocation: To satisfy governing equations at nodes Sampling: (solution statistics only) Random (Monte Carlo) Deterministic (lattice rule, tensor grid, cubature) Stochastic collocation: To construct polynomial approximations Lagrange interpolation Can not be constructed for any given nodes Interpolation error hard to control Pseudo spectral Utilize gpc polynomial basis Becomes multivariate integration Response surface method Multivariate interpolation Many ad hoc approaches

6 Stochastic Collocation Lagrange Interpolation odal set: Lagrange interpolation: Θ = Q i Q nz { Z } R 1 i= Q Q j j u ( Z) u( Z ) Lj( Z) Li( Z ) = δij, 1 i, j Q j= 1 Solution: for j=1,,q, u (, txz j, ) = L( u ), in (, T ] D, t B( u) =, [, T] D, j u = u ( x, Z ), { t = } D Tensor product: i1 n ( U U i z ) Sparse grid (Smolyak): q + 1 i q 1 q i q i i i 1 ( 1) ( U U nz ) (Xiu & Hesthaven, SIAM J. Sci. Comput., 5)

7 Stochastic Collocation: Pseudo Spectral Approach th -order gpc projection u (,, ) ˆ t x Z = uk(, t x) Φk( Z), uˆ = u( Z) Φ ( Z) ρ ( Z) dz. k k gpc-collocation approximation w (, t x, Z) = wˆ (, t x) Φ ( Z), k = k j j j k k k j= 1 k Q wˆ = u( t, x, Z ) Φ ( Z ) α u( Z) Φ ( Z) ρ( Z) dz k = wˆ (, t x) uˆ (, t x), Q k k Aliasing Error: ε u w Q L 2 ( Z ) ρ (Xiu, Comm. Comput Phys, vol. 2, 7)

8 gpc-collocation: Algorithm j j Q { Z α } j = 1 nz 1. Choose a nodal set, in R Deterministic solver 2. Solve for each j = 1,, Q, u (, txz j, ) = L( u ), in (, T ] D, t B( u) =, [, T] D, j u = u ( x, Z ), { t = } D 3. Evaluate the approximate gpc expansion coefficient Q j k k j j α k j= 1 wˆ = u( t, x, Z ) Φ ( Z ), ; th 4. Construct the -order gpc approximation w ( t, x, Z) = wˆ Φ ( Z). k = 1 k k Post-process Error bound (Xiu, 7): ( ) ε u w ε + ε + M ε C Δ 2 L ( Z) Q Q ρ 1/2 Error Finite-term projection error + aliasing error + umerical error

9 Parameter Estimation: Bayesian Inverse Approach Stochastic PDE: u nz (, txz, ) = L( u ), (, T ] D R t B( u) =, [, T] D R u = u( x, Z), { t = } D R n n z z Solution: u utxz (,, ):[, T] D R n n z R Prior distribution: n z π ( z) = π ( z ) Z i i i= 1 Estimation of the prior distribution Requires direct measurements of the parameters o/not enough direct measurements? (Use experience/intuition ) How to take advantage of measurements of other variables?

10 Bayesian Inference Data: n d = G( Z) + e, e R d is i.i.d. Posterior distribution: d π ( d Z) π ( Z) π ( Z) π( Z d) = π( d Z) π( Z) dz Likelihood function: d LZ ( ) π( d Z) = π ( d G( Z)) n i= 1 e i i otes: Difficult to manipulate Classical sampling approaches can be time consuming (MCMC, etc) GPC (Galerkin) based approach: (Marzouk, ajm, Rahn, JCP, 7) gpc approximation: Properties: d L ( Z) π ( Z) π ( Z) = L ( Z) π ( Z) dz d L ( Z) π ( d Z) = π ( d G ( Z)) ei i, i i= 1 Allows direct sampling in term of Z with arbitrarily large samples (Virtually) no additional computational cost forward problem solver only Convergence seems natural n i

11 Convergence of gpc Bayesian Inference Kullback-Leibler divergence: π1( z) D( π1 π2) π1( z)log dz π ( z) 2 Observation error: 2 e (, σ I), i.i.d. ormal Theorem π d 2. If the gpc expansion G converges to G in Lπ, then the posterior density d converges to π in the sense Moreover, if G( Z) G ( Z) i i, 2 π z then for sufficiently large, L d d ( ) D π π,. D α C, 1 i n, α >, C independent of, d d 2 ( π π ) otes: Fast (exponential) convergence rate is retained Factor of 2 in the convergence rates α. d Z (Marzouk & Xiu, Comm. Comput. Phys. 8)

12 Parameter Estimation: Supersensitivity Example Burgers equation : Boundary conditions : 2 u u u + u = ν, x 1,1 2 t x x [ ] u( 1) = 1 + δ( Z); u(1) = 1; < δ<< 1 Deterministic results with no uncertainty (Xiu & Karniadakis, Int. J. umer. Eng., 4) 1% uncertainty in left BC

13 25 exact posterior gpc, p=4 gpc, p=8 δ true 2 p(δ data ) Prior distribution is uniform Measurement noise: e~(,.5 2 ) δ

14 1 2 error 1 3 D(π π ) G G L Factor = 2.1 (theory = 2)

15 Parameter Estimation: Step Function Assume the forward model is a step function Posterior distribution is discontinuous Gibb s oscillations exist Slow convergence with global gpc basis functions exact posterior gpc posterior.8 2 G(z) or G (z).6.4 π d (z) exact forward solution gpc approximation z Forward model and its approximation z Posterior distribution and its approximation

16 D(π π ) 1.1 G G L error p Factor = 1.99 (theory = 2)

17 Kalman Filter for Data Assimilation t m True state (unknown): u R, m 1 Forecast: Observation: Analysis: f du f (, tz) = F(, tu ), t (, T] dt f u (, Z) = u t m d = Hu + ε R, H : R R a f f u = u + K( d Hu ) K = P H ( HP H + R) f T f T 1 (Kalman gain matrix) P = E ( u u )( u u ) f f t f t T R E εε T = Properties: Straightforward for linear dynamic equations Extension to nonlinear equations: Extended KF (EKF) Optimal for Gaussian Explicit calculation of covariance can be costly

18 Ensemble Kalman Filter (EnKF) Ensemble: f f i ( u ) u (, ), = 1,, i tz i M ( d) = d+ (), ε i i = 1,, i a f f ( ) ( ) e i ( ) u = u + K ( ), i = 1,, M i i d H u i f T f T K = P H ( HP Η + R ) e e e e 1 f f P f = f f T f e ( u u )( u u ) P R εε T R e = Properties: nonlinear dynamics sampling errors Measurement. Can be eliminated by square-root filter (EnSRF) Solution states. Computational cost is of great concern

19 Error Analysis of the EnKF Assimilation step size: Δ T = t t n+ 1 n Lemma (local error): e M ε + ΔK ε + ΔK H ε f n+ 1 Δt Δt p ( Δ, σ ) O t α M = I KH Δ K = Ke K Theorem (global error): n E E + e exp Λ t ( ) n k n k= 1 Λ ΔT 1 ote the inverse dependence on assimilation step size (Li & Xiu, vol. 197, CMAME 8)

20 EnKF Example: Linear Wave Equation Model description: Linear advection equation; Periodic domain of length L=1; Wave speed = 1; grid spacing=1; time step = 1; True states are sampled from a Gaussian process, with zero mean and unit variance, and a spatial decorrelation length of 2. The dimension of random space is n z =5. Four measurements uniformly in space are made every 5 time units. Measurement variance is.1. o model error ( xz, ) R t=5 t=1 t= R 5 Long-term Wave propagation

21 Error Behavior of EnKF T=1 T=5.15 T=1 T=5 6.1 log (Error) Error log () w.r.t. ensemble size Standard Deviation of Measurements w.r.t. data noise level

22 EnKF Error Behavior EnKF EnSRF qensrf 2 qensrf 3 8 log(error) log(δ T) qensrf: EnSRF combined with deterministic sampling using optimal cubature

23 GPC Collocation based Kalman Filter Errors of assimilated results T=1 T=5 T=1, T=1,5 EnKF (=1) EnKF (=1 3 ) EnKF (=1 4 ) gpc-kf (=51) gpc-kf (=1) dimensional random space (Li & Xiu, vol. 197, CMAME 8)

24 Accuracy Improvement of EnKF via gpc Use cubature equally weighted Use pseudo-spectral gpc Analytical expression in Z f f u (, tz ) ˆ u () t ( Z k Φk ), k = Statistics ( ) u = uˆ, P = uˆ uˆ f f f f k k < k T (Li & Xiu, J. Comput. Phys. 8)

25 GPC Based Ensemble Kalman Filter Ensemble: f f i ( ) k k ( ) u = u ˆ () t Φ Z, i = 1,, M, M 1 i k = Square-root update: f f f ' a a a ( ) ( ) ( ) ( ) u = u + u, u = u + u, i = 1,, M i i i i Mean state update: u a f ( f = u + K d Hu), Perturbation update: a f f ( u) ( u) K H( u) ' ' ' = +, i = 1,, M i i i ' f T f T K = P H ( HP Η + R) 1 T 1 f T f T f T = ( ) ( ) K PH HPH R HPH R R 1

26 Example: onlinear Population Dynamics du dt f f u f f = r 1 u, u () = u A True state Measurement Estimate Time

27 Error Convergence Mean Standard deviation 2 Mean Standard deviation log(error) log(error) Degree of the gpc expansion () umber of quardrature points (Q) =8, Q=1 is sufficient

28 Comparison: gpc-kf vs EnKF 2 Mean (gpc EnSRF) Std (gpc EnSRF) Mean (EnSRF) Std (EnSRF) 4 log(error) log(ensemble size)

29 onlinear System Example: Lorenz Equations dx dt dy dt dz dt = σ ( y x) = ρ x y xz = xy β z σ = 1, ρ = 28, β = 8/ 3 ( x, y, z ) = (1.5887, , ) 2 X Time 5 Y Time 6 4 Z Time Small deviation in initial condition (.1 in x ) causes large deviation

30 Qualitative Comparison: gpc EnKF vs EnKF =2 (gpc EnSRF) =3 (gpc EnSRF) EnSRF Estimate True state Time GPC EnKF: Q=5 3 =125 EnKF: ensemble size = 14

31 Summary Point selection is crucial for the efficacy of stochastic collocation GPC expansion is much more than a forward UQ method Bayesian inverse (Marzouk & Xiu, Comm. Comput. Phys, 8) Kalman filter for data assimilation (Li & Xiu, CMAME 8; JCP 8)

A Stochastic Collocation Approach to Bayesian Inference in Inverse Problems

A Stochastic Collocation Approach to Bayesian Inference in Inverse Problems COMMUNICATIONS IN COMPUTATIONAL PHYSICS Vol. 6, No. 4, pp. 86-847 Commun. Comput. Phys. October 9 A Stochastic Collocation Approach to Bayesian Inference in Inverse Problems Youssef Marzouk, and Dongbin

More information

Fast Numerical Methods for Stochastic Computations

Fast Numerical Methods for Stochastic Computations Fast AreviewbyDongbinXiu May 16 th,2013 Outline Motivation 1 Motivation 2 3 4 5 Example: Burgers Equation Let us consider the Burger s equation: u t + uu x = νu xx, x [ 1, 1] u( 1) =1 u(1) = 1 Example:

More information

Stochastic Spectral Approaches to Bayesian Inference

Stochastic Spectral Approaches to Bayesian Inference Stochastic Spectral Approaches to Bayesian Inference Prof. Nathan L. Gibson Department of Mathematics Applied Mathematics and Computation Seminar March 4, 2011 Prof. Gibson (OSU) Spectral Approaches to

More information

Data assimilation in high dimensions

Data assimilation in high dimensions Data assimilation in high dimensions David Kelly Courant Institute New York University New York NY www.dtbkelly.com February 12, 2015 Graduate seminar, CIMS David Kelly (CIMS) Data assimilation February

More information

Ergodicity in data assimilation methods

Ergodicity in data assimilation methods Ergodicity in data assimilation methods David Kelly Andy Majda Xin Tong Courant Institute New York University New York NY www.dtbkelly.com April 15, 2016 ETH Zurich David Kelly (CIMS) Data assimilation

More information

Algorithms for Uncertainty Quantification

Algorithms for Uncertainty Quantification Algorithms for Uncertainty Quantification Lecture 9: Sensitivity Analysis ST 2018 Tobias Neckel Scientific Computing in Computer Science TUM Repetition of Previous Lecture Sparse grids in Uncertainty Quantification

More information

A Spectral Approach to Linear Bayesian Updating

A Spectral Approach to Linear Bayesian Updating A Spectral Approach to Linear Bayesian Updating Oliver Pajonk 1,2, Bojana V. Rosic 1, Alexander Litvinenko 1, and Hermann G. Matthies 1 1 Institute of Scientific Computing, TU Braunschweig, Germany 2 SPT

More information

Fast Numerical Methods for Stochastic Computations: A Review

Fast Numerical Methods for Stochastic Computations: A Review COMMUNICATIONS IN COMPUTATIONAL PHYSICS Vol. 5, No. 2-4, pp. 242-272 Commun. Comput. Phys. February 2009 REVIEW ARTICLE Fast Numerical Methods for Stochastic Computations: A Review Dongbin Xiu Department

More information

Data assimilation in high dimensions

Data assimilation in high dimensions Data assimilation in high dimensions David Kelly Kody Law Andy Majda Andrew Stuart Xin Tong Courant Institute New York University New York NY www.dtbkelly.com February 3, 2016 DPMMS, University of Cambridge

More information

Benjamin L. Pence 1, Hosam K. Fathy 2, and Jeffrey L. Stein 3

Benjamin L. Pence 1, Hosam K. Fathy 2, and Jeffrey L. Stein 3 2010 American Control Conference Marriott Waterfront, Baltimore, MD, USA June 30-July 02, 2010 WeC17.1 Benjamin L. Pence 1, Hosam K. Fathy 2, and Jeffrey L. Stein 3 (1) Graduate Student, (2) Assistant

More information

Performance Evaluation of Generalized Polynomial Chaos

Performance Evaluation of Generalized Polynomial Chaos Performance Evaluation of Generalized Polynomial Chaos Dongbin Xiu, Didier Lucor, C.-H. Su, and George Em Karniadakis 1 Division of Applied Mathematics, Brown University, Providence, RI 02912, USA, gk@dam.brown.edu

More information

Beyond Wiener Askey Expansions: Handling Arbitrary PDFs

Beyond Wiener Askey Expansions: Handling Arbitrary PDFs Journal of Scientific Computing, Vol. 27, Nos. 1 3, June 2006 ( 2005) DOI: 10.1007/s10915-005-9038-8 Beyond Wiener Askey Expansions: Handling Arbitrary PDFs Xiaoliang Wan 1 and George Em Karniadakis 1

More information

A Stochastic Collocation based. for Data Assimilation

A Stochastic Collocation based. for Data Assimilation A Stochastic Collocation based Kalman Filter (SCKF) for Data Assimilation Lingzao Zeng and Dongxiao Zhang University of Southern California August 11, 2009 Los Angeles Outline Introduction SCKF Algorithm

More information

Lecture 2: From Linear Regression to Kalman Filter and Beyond

Lecture 2: From Linear Regression to Kalman Filter and Beyond Lecture 2: From Linear Regression to Kalman Filter and Beyond January 18, 2017 Contents 1 Batch and Recursive Estimation 2 Towards Bayesian Filtering 3 Kalman Filter and Bayesian Filtering and Smoothing

More information

Uncertainty Quantification for multiscale kinetic equations with high dimensional random inputs with sparse grids

Uncertainty Quantification for multiscale kinetic equations with high dimensional random inputs with sparse grids Uncertainty Quantification for multiscale kinetic equations with high dimensional random inputs with sparse grids Shi Jin University of Wisconsin-Madison, USA Kinetic equations Different Q Boltmann Landau

More information

Final Report: DE-FG02-95ER25239 Spectral Representations of Uncertainty: Algorithms and Applications

Final Report: DE-FG02-95ER25239 Spectral Representations of Uncertainty: Algorithms and Applications Final Report: DE-FG02-95ER25239 Spectral Representations of Uncertainty: Algorithms and Applications PI: George Em Karniadakis Division of Applied Mathematics, Brown University April 25, 2005 1 Objectives

More information

Organization. I MCMC discussion. I project talks. I Lecture.

Organization. I MCMC discussion. I project talks. I Lecture. Organization I MCMC discussion I project talks. I Lecture. Content I Uncertainty Propagation Overview I Forward-Backward with an Ensemble I Model Reduction (Intro) Uncertainty Propagation in Causal Systems

More information

Bayesian Inverse problem, Data assimilation and Localization

Bayesian Inverse problem, Data assimilation and Localization Bayesian Inverse problem, Data assimilation and Localization Xin T Tong National University of Singapore ICIP, Singapore 2018 X.Tong Localization 1 / 37 Content What is Bayesian inverse problem? What is

More information

Methods of Data Assimilation and Comparisons for Lagrangian Data

Methods of Data Assimilation and Comparisons for Lagrangian Data Methods of Data Assimilation and Comparisons for Lagrangian Data Chris Jones, Warwick and UNC-CH Kayo Ide, UCLA Andrew Stuart, Jochen Voss, Warwick Guillaume Vernieres, UNC-CH Amarjit Budiraja, UNC-CH

More information

Convergence of the Ensemble Kalman Filter in Hilbert Space

Convergence of the Ensemble Kalman Filter in Hilbert Space Convergence of the Ensemble Kalman Filter in Hilbert Space Jan Mandel Center for Computational Mathematics Department of Mathematical and Statistical Sciences University of Colorado Denver Parts based

More information

Uncertainty Quantification in MEMS

Uncertainty Quantification in MEMS Uncertainty Quantification in MEMS N. Agarwal and N. R. Aluru Department of Mechanical Science and Engineering for Advanced Science and Technology Introduction Capacitive RF MEMS switch Comb drive Various

More information

Adaptive ensemble Kalman filtering of nonlinear systems

Adaptive ensemble Kalman filtering of nonlinear systems Adaptive ensemble Kalman filtering of nonlinear systems Tyrus Berry George Mason University June 12, 213 : Problem Setup We consider a system of the form: x k+1 = f (x k ) + ω k+1 ω N (, Q) y k+1 = h(x

More information

Simulating with uncertainty : the rough surface scattering problem

Simulating with uncertainty : the rough surface scattering problem Simulating with uncertainty : the rough surface scattering problem Uday Khankhoje Assistant Professor, Electrical Engineering Indian Institute of Technology Madras Uday Khankhoje (EE, IITM) Simulating

More information

Dynamic System Identification using HDMR-Bayesian Technique

Dynamic System Identification using HDMR-Bayesian Technique Dynamic System Identification using HDMR-Bayesian Technique *Shereena O A 1) and Dr. B N Rao 2) 1), 2) Department of Civil Engineering, IIT Madras, Chennai 600036, Tamil Nadu, India 1) ce14d020@smail.iitm.ac.in

More information

EFFICIENT STOCHASTIC GALERKIN METHODS FOR RANDOM DIFFUSION EQUATIONS

EFFICIENT STOCHASTIC GALERKIN METHODS FOR RANDOM DIFFUSION EQUATIONS EFFICIENT STOCHASTIC GALERKIN METHODS FOR RANDOM DIFFUSION EQUATIONS DONGBIN XIU AND JIE SHEN Abstract. We discuss in this paper efficient solvers for stochastic diffusion equations in random media. We

More information

Introduction to Uncertainty Quantification in Computational Science Handout #3

Introduction to Uncertainty Quantification in Computational Science Handout #3 Introduction to Uncertainty Quantification in Computational Science Handout #3 Gianluca Iaccarino Department of Mechanical Engineering Stanford University June 29 - July 1, 2009 Scuola di Dottorato di

More information

Sequential Monte Carlo Samplers for Applications in High Dimensions

Sequential Monte Carlo Samplers for Applications in High Dimensions Sequential Monte Carlo Samplers for Applications in High Dimensions Alexandros Beskos National University of Singapore KAUST, 26th February 2014 Joint work with: Dan Crisan, Ajay Jasra, Nik Kantas, Alex

More information

A Polynomial Chaos Approach to Robust Multiobjective Optimization

A Polynomial Chaos Approach to Robust Multiobjective Optimization A Polynomial Chaos Approach to Robust Multiobjective Optimization Silvia Poles 1, Alberto Lovison 2 1 EnginSoft S.p.A., Optimization Consulting Via Giambellino, 7 35129 Padova, Italy s.poles@enginsoft.it

More information

On a Data Assimilation Method coupling Kalman Filtering, MCRE Concept and PGD Model Reduction for Real-Time Updating of Structural Mechanics Model

On a Data Assimilation Method coupling Kalman Filtering, MCRE Concept and PGD Model Reduction for Real-Time Updating of Structural Mechanics Model On a Data Assimilation Method coupling, MCRE Concept and PGD Model Reduction for Real-Time Updating of Structural Mechanics Model 2016 SIAM Conference on Uncertainty Quantification Basile Marchand 1, Ludovic

More information

Gaussian Process Approximations of Stochastic Differential Equations

Gaussian Process Approximations of Stochastic Differential Equations Gaussian Process Approximations of Stochastic Differential Equations Cédric Archambeau Dan Cawford Manfred Opper John Shawe-Taylor May, 2006 1 Introduction Some of the most complex models routinely run

More information

EnKF and Catastrophic filter divergence

EnKF and Catastrophic filter divergence EnKF and Catastrophic filter divergence David Kelly Kody Law Andrew Stuart Mathematics Department University of North Carolina Chapel Hill NC bkelly.com June 4, 2014 SIAM UQ 2014 Savannah. David Kelly

More information

Ensemble Kalman Filter

Ensemble Kalman Filter Ensemble Kalman Filter Geir Evensen and Laurent Bertino Hydro Research Centre, Bergen, Norway, Nansen Environmental and Remote Sensing Center, Bergen, Norway The Ensemble Kalman Filter (EnKF) Represents

More information

State and Parameter Estimation in Stochastic Dynamical Models

State and Parameter Estimation in Stochastic Dynamical Models State and Parameter Estimation in Stochastic Dynamical Models Timothy DelSole George Mason University, Fairfax, Va and Center for Ocean-Land-Atmosphere Studies, Calverton, MD June 21, 2011 1 1 collaboration

More information

Parametric Problems, Stochastics, and Identification

Parametric Problems, Stochastics, and Identification Parametric Problems, Stochastics, and Identification Hermann G. Matthies a B. Rosić ab, O. Pajonk ac, A. Litvinenko a a, b University of Kragujevac c SPT Group, Hamburg wire@tu-bs.de http://www.wire.tu-bs.de

More information

Analysis and Computation of Hyperbolic PDEs with Random Data

Analysis and Computation of Hyperbolic PDEs with Random Data Analysis and Computation of Hyperbolic PDEs with Random Data Mohammad Motamed 1, Fabio Nobile 2,3 and Raúl Tempone 1 1 King Abdullah University of Science and Technology, Thuwal, Saudi Arabia 2 EPFL Lausanne,

More information

Local Ensemble Transform Kalman Filter

Local Ensemble Transform Kalman Filter Local Ensemble Transform Kalman Filter Brian Hunt 11 June 2013 Review of Notation Forecast model: a known function M on a vector space of model states. Truth: an unknown sequence {x n } of model states

More information

EnKF and filter divergence

EnKF and filter divergence EnKF and filter divergence David Kelly Andrew Stuart Kody Law Courant Institute New York University New York, NY dtbkelly.com December 12, 2014 Applied and computational mathematics seminar, NIST. David

More information

Overview. Bayesian assimilation of experimental data into simulation (for Goland wing flutter) Why not uncertainty quantification?

Overview. Bayesian assimilation of experimental data into simulation (for Goland wing flutter) Why not uncertainty quantification? Delft University of Technology Overview Bayesian assimilation of experimental data into simulation (for Goland wing flutter), Simao Marques 1. Why not uncertainty quantification? 2. Why uncertainty quantification?

More information

Uncertainty Quantification in Computational Models

Uncertainty Quantification in Computational Models Uncertainty Quantification in Computational Models Habib N. Najm Sandia National Laboratories, Livermore, CA, USA Workshop on Understanding Climate Change from Data (UCC11) University of Minnesota, Minneapolis,

More information

What do we know about EnKF?

What do we know about EnKF? What do we know about EnKF? David Kelly Kody Law Andrew Stuart Andrew Majda Xin Tong Courant Institute New York University New York, NY April 10, 2015 CAOS seminar, Courant. David Kelly (NYU) EnKF April

More information

Research Article A Pseudospectral Approach for Kirchhoff Plate Bending Problems with Uncertainties

Research Article A Pseudospectral Approach for Kirchhoff Plate Bending Problems with Uncertainties Mathematical Problems in Engineering Volume 22, Article ID 7565, 4 pages doi:.55/22/7565 Research Article A Pseudospectral Approach for Kirchhoff Plate Bending Problems with Uncertainties Ling Guo, 2 and

More information

Multilevel stochastic collocations with dimensionality reduction

Multilevel stochastic collocations with dimensionality reduction Multilevel stochastic collocations with dimensionality reduction Ionut Farcas TUM, Chair of Scientific Computing in Computer Science (I5) 27.01.2017 Outline 1 Motivation 2 Theoretical background Uncertainty

More information

Optimisation under Uncertainty with Stochastic PDEs for the History Matching Problem in Reservoir Engineering

Optimisation under Uncertainty with Stochastic PDEs for the History Matching Problem in Reservoir Engineering Optimisation under Uncertainty with Stochastic PDEs for the History Matching Problem in Reservoir Engineering Hermann G. Matthies Technische Universität Braunschweig wire@tu-bs.de http://www.wire.tu-bs.de

More information

A Note on the Particle Filter with Posterior Gaussian Resampling

A Note on the Particle Filter with Posterior Gaussian Resampling Tellus (6), 8A, 46 46 Copyright C Blackwell Munksgaard, 6 Printed in Singapore. All rights reserved TELLUS A Note on the Particle Filter with Posterior Gaussian Resampling By X. XIONG 1,I.M.NAVON 1,2 and

More information

Convergence of Square Root Ensemble Kalman Filters in the Large Ensemble Limit

Convergence of Square Root Ensemble Kalman Filters in the Large Ensemble Limit Convergence of Square Root Ensemble Kalman Filters in the Large Ensemble Limit Evan Kwiatkowski, Jan Mandel University of Colorado Denver December 11, 2014 OUTLINE 2 Data Assimilation Bayesian Estimation

More information

Gaussian Process Approximations of Stochastic Differential Equations

Gaussian Process Approximations of Stochastic Differential Equations Gaussian Process Approximations of Stochastic Differential Equations Cédric Archambeau Centre for Computational Statistics and Machine Learning University College London c.archambeau@cs.ucl.ac.uk CSML

More information

Steps in Uncertainty Quantification

Steps in Uncertainty Quantification Steps in Uncertainty Quantification Challenge: How do we do uncertainty quantification for computationally expensive models? Example: We have a computational budget of 5 model evaluations. Bayesian inference

More information

Sparse polynomial chaos expansions in engineering applications

Sparse polynomial chaos expansions in engineering applications DEPARTMENT OF CIVIL, ENVIRONMENTAL AND GEOMATIC ENGINEERING CHAIR OF RISK, SAFETY & UNCERTAINTY QUANTIFICATION Sparse polynomial chaos expansions in engineering applications B. Sudret G. Blatman (EDF R&D,

More information

Kalman Filter and Ensemble Kalman Filter

Kalman Filter and Ensemble Kalman Filter Kalman Filter and Ensemble Kalman Filter 1 Motivation Ensemble forecasting : Provides flow-dependent estimate of uncertainty of the forecast. Data assimilation : requires information about uncertainty

More information

New Fast Kalman filter method

New Fast Kalman filter method New Fast Kalman filter method Hojat Ghorbanidehno, Hee Sun Lee 1. Introduction Data assimilation methods combine dynamical models of a system with typically noisy observations to obtain estimates of the

More information

Fully Bayesian Deep Gaussian Processes for Uncertainty Quantification

Fully Bayesian Deep Gaussian Processes for Uncertainty Quantification Fully Bayesian Deep Gaussian Processes for Uncertainty Quantification N. Zabaras 1 S. Atkinson 1 Center for Informatics and Computational Science Department of Aerospace and Mechanical Engineering University

More information

Estimating functional uncertainty using polynomial chaos and adjoint equations

Estimating functional uncertainty using polynomial chaos and adjoint equations 0. Estimating functional uncertainty using polynomial chaos and adjoint equations February 24, 2011 1 Florida State University, Tallahassee, Florida, Usa 2 Moscow Institute of Physics and Technology, Moscow,

More information

Data assimilation with and without a model

Data assimilation with and without a model Data assimilation with and without a model Tyrus Berry George Mason University NJIT Feb. 28, 2017 Postdoc supported by NSF This work is in collaboration with: Tim Sauer, GMU Franz Hamilton, Postdoc, NCSU

More information

Uncertainty Quantification for multiscale kinetic equations with random inputs. Shi Jin. University of Wisconsin-Madison, USA

Uncertainty Quantification for multiscale kinetic equations with random inputs. Shi Jin. University of Wisconsin-Madison, USA Uncertainty Quantification for multiscale kinetic equations with random inputs Shi Jin University of Wisconsin-Madison, USA Where do kinetic equations sit in physics Kinetic equations with applications

More information

Adaptive Collocation with Kernel Density Estimation

Adaptive Collocation with Kernel Density Estimation Examples of with Kernel Density Estimation Howard C. Elman Department of Computer Science University of Maryland at College Park Christopher W. Miller Applied Mathematics and Scientific Computing Program

More information

Quadrature for Uncertainty Analysis Stochastic Collocation. What does quadrature have to do with uncertainty?

Quadrature for Uncertainty Analysis Stochastic Collocation. What does quadrature have to do with uncertainty? Quadrature for Uncertainty Analysis Stochastic Collocation What does quadrature have to do with uncertainty? Quadrature for Uncertainty Analysis Stochastic Collocation What does quadrature have to do with

More information

Covariance function estimation in Gaussian process regression

Covariance function estimation in Gaussian process regression Covariance function estimation in Gaussian process regression François Bachoc Department of Statistics and Operations Research, University of Vienna WU Research Seminar - May 2015 François Bachoc Gaussian

More information

The Ensemble Kalman Filter:

The Ensemble Kalman Filter: p.1 The Ensemble Kalman Filter: Theoretical formulation and practical implementation Geir Evensen Norsk Hydro Research Centre, Bergen, Norway Based on Evensen 23, Ocean Dynamics, Vol 53, No 4 p.2 The Ensemble

More information

Stochastic structural dynamic analysis with random damping parameters

Stochastic structural dynamic analysis with random damping parameters Stochastic structural dynamic analysis with random damping parameters K. Sepahvand 1, F. Saati Khosroshahi, C. A. Geweth and S. Marburg Chair of Vibroacoustics of Vehicles and Machines Department of Mechanical

More information

Uncertainty Quantification and hypocoercivity based sensitivity analysis for multiscale kinetic equations with random inputs.

Uncertainty Quantification and hypocoercivity based sensitivity analysis for multiscale kinetic equations with random inputs. Uncertainty Quantification and hypocoercivity based sensitivity analysis for multiscale kinetic equations with random inputs Shi Jin University of Wisconsin-Madison, USA Shanghai Jiao Tong University,

More information

Schwarz Preconditioner for the Stochastic Finite Element Method

Schwarz Preconditioner for the Stochastic Finite Element Method Schwarz Preconditioner for the Stochastic Finite Element Method Waad Subber 1 and Sébastien Loisel 2 Preprint submitted to DD22 conference 1 Introduction The intrusive polynomial chaos approach for uncertainty

More information

NON-LINEAR APPROXIMATION OF BAYESIAN UPDATE

NON-LINEAR APPROXIMATION OF BAYESIAN UPDATE tifica NON-LINEAR APPROXIMATION OF BAYESIAN UPDATE Alexander Litvinenko 1, Hermann G. Matthies 2, Elmar Zander 2 http://sri-uq.kaust.edu.sa/ 1 Extreme Computing Research Center, KAUST, 2 Institute of Scientific

More information

Solving the Stochastic Steady-State Diffusion Problem Using Multigrid

Solving the Stochastic Steady-State Diffusion Problem Using Multigrid Solving the Stochastic Steady-State Diffusion Problem Using Multigrid Tengfei Su Applied Mathematics and Scientific Computing Advisor: Howard Elman Department of Computer Science Sept. 29, 2015 Tengfei

More information

arxiv: v2 [math.na] 6 Aug 2018

arxiv: v2 [math.na] 6 Aug 2018 Bayesian model calibration with interpolating polynomials based on adaptively weighted Leja nodes L.M.M. van den Bos,2, B. Sanderse, W.A.A.M. Bierbooms 2, and G.J.W. van Bussel 2 arxiv:82.235v2 [math.na]

More information

Nonlinear stochastic Galerkin and collocation methods: application to a ferromagnetic cylinder rotating at high speed

Nonlinear stochastic Galerkin and collocation methods: application to a ferromagnetic cylinder rotating at high speed Nonlinear stochastic Galerkin and collocation methods: application to a ferromagnetic cylinder rotating at high speed Eveline Rosseel Herbert De Gersem Stefan Vandewalle Report TW 541, July 29 Katholieke

More information

Implementation of Sparse Wavelet-Galerkin FEM for Stochastic PDEs

Implementation of Sparse Wavelet-Galerkin FEM for Stochastic PDEs Implementation of Sparse Wavelet-Galerkin FEM for Stochastic PDEs Roman Andreev ETH ZÜRICH / 29 JAN 29 TOC of the Talk Motivation & Set-Up Model Problem Stochastic Galerkin FEM Conclusions & Outlook Motivation

More information

Uncertainty Quantification in Computational Science

Uncertainty Quantification in Computational Science DTU 2010 - Lecture I Uncertainty Quantification in Computational Science Jan S Hesthaven Brown University Jan.Hesthaven@Brown.edu Objective of lectures The main objective of these lectures are To offer

More information

Surrogate and reduced-order modeling: a comparison of approaches for large-scale statistical inverse problems

Surrogate and reduced-order modeling: a comparison of approaches for large-scale statistical inverse problems Surrogate and reduced-order modeling: a comparison of approaches for large-scale statistical inverse problems M. Frangos, Y. Marzouk, K. Willcox, Massachusetts Institute of Technology; B. van Bloemen Waanders,

More information

The Ensemble Kalman Filter:

The Ensemble Kalman Filter: p.1 The Ensemble Kalman Filter: Theoretical formulation and practical implementation Geir Evensen Norsk Hydro Research Centre, Bergen, Norway Based on Evensen, Ocean Dynamics, Vol 5, No p. The Ensemble

More information

Lecture 6: Bayesian Inference in SDE Models

Lecture 6: Bayesian Inference in SDE Models Lecture 6: Bayesian Inference in SDE Models Bayesian Filtering and Smoothing Point of View Simo Särkkä Aalto University Simo Särkkä (Aalto) Lecture 6: Bayesian Inference in SDEs 1 / 45 Contents 1 SDEs

More information

Superparameterization and Dynamic Stochastic Superresolution (DSS) for Filtering Sparse Geophysical Flows

Superparameterization and Dynamic Stochastic Superresolution (DSS) for Filtering Sparse Geophysical Flows SP and DSS for Filtering Sparse Geophysical Flows Superparameterization and Dynamic Stochastic Superresolution (DSS) for Filtering Sparse Geophysical Flows Presented by Nan Chen, Michal Branicki and Chenyue

More information

STOCHASTIC SAMPLING METHODS

STOCHASTIC SAMPLING METHODS STOCHASTIC SAMPLING METHODS APPROXIMATING QUANTITIES OF INTEREST USING SAMPLING METHODS Recall that quantities of interest often require the evaluation of stochastic integrals of functions of the solutions

More information

Gaussian Filtering Strategies for Nonlinear Systems

Gaussian Filtering Strategies for Nonlinear Systems Gaussian Filtering Strategies for Nonlinear Systems Canonical Nonlinear Filtering Problem ~u m+1 = ~ f (~u m )+~ m+1 ~v m+1 = ~g(~u m+1 )+~ o m+1 I ~ f and ~g are nonlinear & deterministic I Noise/Errors

More information

An Empirical Chaos Expansion Method for Uncertainty Quantification

An Empirical Chaos Expansion Method for Uncertainty Quantification An Empirical Chaos Expansion Method for Uncertainty Quantification Melvin Leok and Gautam Wilkins Abstract. Uncertainty quantification seeks to provide a quantitative means to understand complex systems

More information

Smoothers: Types and Benchmarks

Smoothers: Types and Benchmarks Smoothers: Types and Benchmarks Patrick N. Raanes Oxford University, NERSC 8th International EnKF Workshop May 27, 2013 Chris Farmer, Irene Moroz Laurent Bertino NERSC Geir Evensen Abstract Talk builds

More information

Winter 2019 Math 106 Topics in Applied Mathematics. Lecture 1: Introduction

Winter 2019 Math 106 Topics in Applied Mathematics. Lecture 1: Introduction Winter 2019 Math 106 Topics in Applied Mathematics Data-driven Uncertainty Quantification Yoonsang Lee (yoonsang.lee@dartmouth.edu) Lecture 1: Introduction 19 Winter M106 Class: MWF 12:50-1:55 pm @ 200

More information

Solving the steady state diffusion equation with uncertainty Final Presentation

Solving the steady state diffusion equation with uncertainty Final Presentation Solving the steady state diffusion equation with uncertainty Final Presentation Virginia Forstall vhfors@gmail.com Advisor: Howard Elman elman@cs.umd.edu Department of Computer Science May 6, 2012 Problem

More information

Bayesian parameter estimation in predictive engineering

Bayesian parameter estimation in predictive engineering Bayesian parameter estimation in predictive engineering Damon McDougall Institute for Computational Engineering and Sciences, UT Austin 14th August 2014 1/27 Motivation Understand physical phenomena Observations

More information

A new Hierarchical Bayes approach to ensemble-variational data assimilation

A new Hierarchical Bayes approach to ensemble-variational data assimilation A new Hierarchical Bayes approach to ensemble-variational data assimilation Michael Tsyrulnikov and Alexander Rakitko HydroMetCenter of Russia College Park, 20 Oct 2014 Michael Tsyrulnikov and Alexander

More information

Data assimilation with and without a model

Data assimilation with and without a model Data assimilation with and without a model Tim Sauer George Mason University Parameter estimation and UQ U. Pittsburgh Mar. 5, 2017 Partially supported by NSF Most of this work is due to: Tyrus Berry,

More information

Nonlinear State Estimation! Particle, Sigma-Points Filters!

Nonlinear State Estimation! Particle, Sigma-Points Filters! Nonlinear State Estimation! Particle, Sigma-Points Filters! Robert Stengel! Optimal Control and Estimation, MAE 546! Princeton University, 2017!! Particle filter!! Sigma-Points Unscented Kalman ) filter!!

More information

LARGE-SCALE TRAFFIC STATE ESTIMATION

LARGE-SCALE TRAFFIC STATE ESTIMATION Hans van Lint, Yufei Yuan & Friso Scholten A localized deterministic Ensemble Kalman Filter LARGE-SCALE TRAFFIC STATE ESTIMATION CONTENTS Intro: need for large-scale traffic state estimation Some Kalman

More information

c 2004 Society for Industrial and Applied Mathematics

c 2004 Society for Industrial and Applied Mathematics SIAM J. SCI. COMPUT. Vol. 6, No., pp. 578 59 c Society for Industrial and Applied Mathematics STOCHASTIC SOLUTIONS FOR THE TWO-DIMENSIONAL ADVECTION-DIFFUSION EQUATION XIAOLIANG WAN, DONGBIN XIU, AND GEORGE

More information

Fundamentals of Data Assimila1on

Fundamentals of Data Assimila1on 2015 GSI Community Tutorial NCAR Foothills Campus, Boulder, CO August 11-14, 2015 Fundamentals of Data Assimila1on Milija Zupanski Cooperative Institute for Research in the Atmosphere Colorado State University

More information

Bayesian Statistics and Data Assimilation. Jonathan Stroud. Department of Statistics The George Washington University

Bayesian Statistics and Data Assimilation. Jonathan Stroud. Department of Statistics The George Washington University Bayesian Statistics and Data Assimilation Jonathan Stroud Department of Statistics The George Washington University 1 Outline Motivation Bayesian Statistics Parameter Estimation in Data Assimilation Combined

More information

Polynomial chaos expansions for sensitivity analysis

Polynomial chaos expansions for sensitivity analysis c DEPARTMENT OF CIVIL, ENVIRONMENTAL AND GEOMATIC ENGINEERING CHAIR OF RISK, SAFETY & UNCERTAINTY QUANTIFICATION Polynomial chaos expansions for sensitivity analysis B. Sudret Chair of Risk, Safety & Uncertainty

More information

DATA ASSIMILATION FOR FLOOD FORECASTING

DATA ASSIMILATION FOR FLOOD FORECASTING DATA ASSIMILATION FOR FLOOD FORECASTING Arnold Heemin Delft University of Technology 09/16/14 1 Data assimilation is the incorporation of measurement into a numerical model to improve the model results

More information

EnKF-based particle filters

EnKF-based particle filters EnKF-based particle filters Jana de Wiljes, Sebastian Reich, Wilhelm Stannat, Walter Acevedo June 20, 2017 Filtering Problem Signal dx t = f (X t )dt + 2CdW t Observations dy t = h(x t )dt + R 1/2 dv t.

More information

Lagrangian Data Assimilation and Its Application to Geophysical Fluid Flows

Lagrangian Data Assimilation and Its Application to Geophysical Fluid Flows Lagrangian Data Assimilation and Its Application to Geophysical Fluid Flows Laura Slivinski June, 3 Laura Slivinski (Brown University) Lagrangian Data Assimilation June, 3 / 3 Data Assimilation Setup:

More information

Spectral Representation of Random Processes

Spectral Representation of Random Processes Spectral Representation of Random Processes Example: Represent u(t,x,q) by! u K (t, x, Q) = u k (t, x) k(q) where k(q) are orthogonal polynomials. Single Random Variable:! Let k (Q) be orthogonal with

More information

AN EFFICIENT COMPUTATIONAL FRAMEWORK FOR UNCERTAINTY QUANTIFICATION IN MULTISCALE SYSTEMS

AN EFFICIENT COMPUTATIONAL FRAMEWORK FOR UNCERTAINTY QUANTIFICATION IN MULTISCALE SYSTEMS AN EFFICIENT COMPUTATIONAL FRAMEWORK FOR UNCERTAINTY QUANTIFICATION IN MULTISCALE SYSTEMS A Dissertation Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfillment of

More information

Lecture 2: From Linear Regression to Kalman Filter and Beyond

Lecture 2: From Linear Regression to Kalman Filter and Beyond Lecture 2: From Linear Regression to Kalman Filter and Beyond Department of Biomedical Engineering and Computational Science Aalto University January 26, 2012 Contents 1 Batch and Recursive Estimation

More information

Sparse Quadrature Algorithms for Bayesian Inverse Problems

Sparse Quadrature Algorithms for Bayesian Inverse Problems Sparse Quadrature Algorithms for Bayesian Inverse Problems Claudia Schillings, Christoph Schwab Pro*Doc Retreat Disentis 2013 Numerical Analysis and Scientific Computing Disentis - 15 August, 2013 research

More information

Ensemble Data Assimilation and Uncertainty Quantification

Ensemble Data Assimilation and Uncertainty Quantification Ensemble Data Assimilation and Uncertainty Quantification Jeff Anderson National Center for Atmospheric Research pg 1 What is Data Assimilation? Observations combined with a Model forecast + to produce

More information

Estimation of State Noise for the Ensemble Kalman filter algorithm for 2D shallow water equations.

Estimation of State Noise for the Ensemble Kalman filter algorithm for 2D shallow water equations. Estimation of State Noise for the Ensemble Kalman filter algorithm for 2D shallow water equations. May 6, 2009 Motivation Constitutive Equations EnKF algorithm Some results Method Navier Stokes equations

More information

PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 13: SEQUENTIAL DATA

PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 13: SEQUENTIAL DATA PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 13: SEQUENTIAL DATA Contents in latter part Linear Dynamical Systems What is different from HMM? Kalman filter Its strength and limitation Particle Filter

More information

Filtering Sparse Regular Observed Linear and Nonlinear Turbulent System

Filtering Sparse Regular Observed Linear and Nonlinear Turbulent System Filtering Sparse Regular Observed Linear and Nonlinear Turbulent System John Harlim, Andrew J. Majda Department of Mathematics and Center for Atmosphere Ocean Science Courant Institute of Mathematical

More information

Numerical Solutions of ODEs by Gaussian (Kalman) Filtering

Numerical Solutions of ODEs by Gaussian (Kalman) Filtering Numerical Solutions of ODEs by Gaussian (Kalman) Filtering Hans Kersting joint work with Michael Schober, Philipp Hennig, Tim Sullivan and Han C. Lie SIAM CSE, Atlanta March 1, 2017 Emmy Noether Group

More information

NEW ALGORITHMS FOR UNCERTAINTY QUANTIFICATION AND NONLINEAR ESTIMATION OF STOCHASTIC DYNAMICAL SYSTEMS. A Dissertation PARIKSHIT DUTTA

NEW ALGORITHMS FOR UNCERTAINTY QUANTIFICATION AND NONLINEAR ESTIMATION OF STOCHASTIC DYNAMICAL SYSTEMS. A Dissertation PARIKSHIT DUTTA NEW ALGORITHMS FOR UNCERTAINTY QUANTIFICATION AND NONLINEAR ESTIMATION OF STOCHASTIC DYNAMICAL SYSTEMS A Dissertation by PARIKSHIT DUTTA Submitted to the Office of Graduate Studies of Texas A&M University

More information

Robust exponential convergence of hp-fem for singularly perturbed systems of reaction-diffusion equations

Robust exponential convergence of hp-fem for singularly perturbed systems of reaction-diffusion equations Robust exponential convergence of hp-fem for singularly perturbed systems of reaction-diffusion equations Christos Xenophontos Department of Mathematics and Statistics University of Cyprus joint work with

More information