-
Analysis of tidal flows through the Strait of Gibraltar using Dynamic Mode Decomposition
Authors:
Sathsara Dias,
Sudam Surasinghe,
Kanaththa Priyankara,
Marko Budišić,
Larry Pratt,
José C. Sanchez-Garrido,
Erik M. Bollt
Abstract:
The Strait of Gibraltar is a region characterized by intricate oceanic sub-mesoscale features, influenced by topography, tidal forces, instabilities, and nonlinear hydraulic processes, all governed by the nonlinear equations of fluid motion. In this study, we aim to uncover the underlying physics of these phenomena within 3D MIT general circulation model simulations, including waves, eddies, and g…
▽ More
The Strait of Gibraltar is a region characterized by intricate oceanic sub-mesoscale features, influenced by topography, tidal forces, instabilities, and nonlinear hydraulic processes, all governed by the nonlinear equations of fluid motion. In this study, we aim to uncover the underlying physics of these phenomena within 3D MIT general circulation model simulations, including waves, eddies, and gyres. To achieve this, we employ Dynamic Mode Decomposition (DMD) to break down simulation snapshots into Koopman modes, with distinct exponential growth/decay rates and oscillation frequencies. Our objectives encompass evaluating DMD's efficacy in capturing known features, unveiling new elements, ranking modes, and exploring order reduction. We also introduce modifications to enhance DMD's robustness, numerical accuracy, and robustness of eigenvalues. DMD analysis yields a comprehensive understanding of flow patterns, internal wave formation, and the dynamics of the Strait of Gibraltar, its meandering behaviors, and the formation of a secondary gyre, notably the Western Alboran Gyre, as well as the propagation of Kelvin and coastal-trapped waves along the African coast. In doing so, it significantly advances our comprehension of intricate oceanographic phenomena and underscores the immense utility of DMD as an analytical tool for such complex datasets, suggesting that DMD could serve as a valuable addition to the toolkit of oceanographers.
△ Less
Submitted 19 January, 2025; v1 submitted 2 November, 2023;
originally announced November 2023.
-
Projected Data Assimilation using Sliding Window Proper Orthogonal Decomposition
Authors:
Aishah Albarakati,
Marko Budisic,
Erik Van Vleck
Abstract:
Prediction of the state evolution of complex high-dimensional nonlinear systems is challenging due to the nonlinear sensitivity of the evolution to small inaccuracies in the model. Data Assimilation (DA) techniques improve state estimates by combining model simulations with real-time data. Few DA techniques can simultaneously handle nonlinear evolution, non-Gaussian uncertainty, and the high dimen…
▽ More
Prediction of the state evolution of complex high-dimensional nonlinear systems is challenging due to the nonlinear sensitivity of the evolution to small inaccuracies in the model. Data Assimilation (DA) techniques improve state estimates by combining model simulations with real-time data. Few DA techniques can simultaneously handle nonlinear evolution, non-Gaussian uncertainty, and the high dimension of the state. We recently proposed addressing these challenges using a Proper Orthogonal Decomposition (POD) technique that projects the physical and data models into a reduced-dimensional subspace. POD is a tool to extract spatiotemporal patterns (modes) that dominate the observed data. We combined the POD-based projection operator, computed in an offline fashion, with a DA scheme that models non-Gaussian uncertainty in lower dimensional subspace. If the model parameters change significantly during time evolution, the offline computation of the projection operators ceases to be useful. We address this challenge using a sliding window POD (SWPOD), which recomputes the projection operator based on a sliding subset of snapshots from the entire evolution. The physical model projection is updated dynamically in terms of modes and number of modes, and the data model projection is also chosen to promote a sparse approximation. We test the efficacy of this technique on a modified Lorenz'96 model with a time-varying forcing and compare it with the time-invariant offline projected algorithm. In particular, dynamically determined physical and data model projections decrease the Root Mean Squared Error and the resampling rate.
△ Less
Submitted 6 July, 2023;
originally announced July 2023.
-
Modern Koopman Theory for Dynamical Systems
Authors:
Steven L. Brunton,
Marko Budišić,
Eurika Kaiser,
J. Nathan Kutz
Abstract:
The field of dynamical systems is being transformed by the mathematical tools and algorithms emerging from modern computing and data science. First-principles derivations and asymptotic reductions are giving way to data-driven approaches that formulate models in operator theoretic or probabilistic frameworks. Koopman spectral theory has emerged as a dominant perspective over the past decade, in wh…
▽ More
The field of dynamical systems is being transformed by the mathematical tools and algorithms emerging from modern computing and data science. First-principles derivations and asymptotic reductions are giving way to data-driven approaches that formulate models in operator theoretic or probabilistic frameworks. Koopman spectral theory has emerged as a dominant perspective over the past decade, in which nonlinear dynamics are represented in terms of an infinite-dimensional linear operator acting on the space of all possible measurement functions of the system. This linear representation of nonlinear dynamics has tremendous potential to enable the prediction, estimation, and control of nonlinear systems with standard textbook methods developed for linear systems. However, obtaining finite-dimensional coordinate systems and embeddings in which the dynamics appear approximately linear remains a central open challenge. The success of Koopman analysis is due primarily to three key factors: 1) there exists rigorous theory connecting it to classical geometric approaches for dynamical systems, 2) the approach is formulated in terms of measurements, making it ideal for leveraging big-data and machine learning techniques, and 3) simple, yet powerful numerical algorithms, such as the dynamic mode decomposition (DMD), have been developed and extended to reduce Koopman theory to practice in real-world applications. In this review, we provide an overview of modern Koopman operator theory, describing recent theoretical and algorithmic developments and highlighting these methods with a diverse range of applications. We also discuss key advances and challenges in the rapidly growing field of machine learning that are likely to drive future developments and significantly transform the theoretical landscape of dynamical systems.
△ Less
Submitted 29 October, 2021; v1 submitted 24 February, 2021;
originally announced February 2021.
-
Model and Data Reduction for Data Assimilation: Particle Filters Employing Projected Forecasts and Data with Application to a Shallow Water Model
Authors:
Aishah Albarakati,
Marko Budišić,
Rose Crocker,
Juniper Glass-Klaiber,
Sarah Iams,
John Maclean,
Noah Marshall,
Colin Roberts,
Erik S. Van Vleck
Abstract:
The understanding of nonlinear, high dimensional flows, e.g, atmospheric and ocean flows, is critical to address the impacts of global climate change. Data Assimilation techniques combine physical models and observational data, often in a Bayesian framework, to predict the future state of the model and the uncertainty in this prediction. Inherent in these systems are noise (Gaussian and non-Gaussi…
▽ More
The understanding of nonlinear, high dimensional flows, e.g, atmospheric and ocean flows, is critical to address the impacts of global climate change. Data Assimilation techniques combine physical models and observational data, often in a Bayesian framework, to predict the future state of the model and the uncertainty in this prediction. Inherent in these systems are noise (Gaussian and non-Gaussian), nonlinearity, and high dimensionality that pose challenges to making accurate predictions. To address these issues we investigate the use of both model and data dimension reduction based on techniques including Assimilation in Unstable Subspaces, Proper Orthogonal Decomposition, and Dynamic Mode Decomposition. Algorithms that take advantage of projected physical and data models may be combined with Data Analysis techniques such as Ensemble Kalman Filter and Particle Filter variants. The projected Data Assimilation techniques are developed for the optimal proposal particle filter and applied to the Lorenz'96 and Shallow Water Equations to test the efficacy of our techniques in high dimensional, nonlinear systems.
△ Less
Submitted 9 June, 2021; v1 submitted 22 January, 2021;
originally announced January 2021.
-
Finite-Time Braiding Exponents
Authors:
Marko Budisic,
Jean-Luc Thiffeault
Abstract:
Topological entropy is a common measure of the rate of mixing in a flow. It can be computed by partition methods, or by estimating the growth rate of material lines or other material elements. This requires detailed knowledge of the velocity field, which is not always available, such as when we only know a few particle trajectories (ocean float data, for example). We propose an alternative approxi…
▽ More
Topological entropy is a common measure of the rate of mixing in a flow. It can be computed by partition methods, or by estimating the growth rate of material lines or other material elements. This requires detailed knowledge of the velocity field, which is not always available, such as when we only know a few particle trajectories (ocean float data, for example). We propose an alternative approximation to topological entropy,applicable to two-dimensional flows, which uses only a finite number of trajectories as input data. To represent these sparse data sets, we use braids, algebraic objects that record how strands, i.e., trajectories, exchange positions with respect to a projection axis. Material curves advected by the flow are represented as simplified loop coordinates. The exponential rate at which a braid deforms loops over a finite time interval as the strands exchange places is the Finite-Time Braiding Exponent (FTBE) and serves as a proxy for topological entropy of the two-dimensional flow. We demonstrate that FTBEs are robust with respect to the value of numerical time step,details of braid representation, and choice of initial conditions inside the mixing region. We also explore how closely the FTBEs approximate topological entropy depending on the number and length of trajectories used.
△ Less
Submitted 7 February, 2015;
originally announced February 2015.
-
Conditioning moments of singular measures for entropy maximization II: Numerical examples
Authors:
Marko Budišić,
Mihai Putinar
Abstract:
If moments of singular measures are passed as inputs to the entropy maximization procedure, the optimization algorithm might not terminate. The framework developed in our previous paper demonstrated how input moments of measures, on a broad range of domains, can be conditioned to ensure convergence of the entropy maximization. Here we numerically illustrate the developed framework on simplest poss…
▽ More
If moments of singular measures are passed as inputs to the entropy maximization procedure, the optimization algorithm might not terminate. The framework developed in our previous paper demonstrated how input moments of measures, on a broad range of domains, can be conditioned to ensure convergence of the entropy maximization. Here we numerically illustrate the developed framework on simplest possible examples: measures with one-dimensional, bounded supports. Three examples of measures are used to numerically compare approximations obtained through entropy maximization with and without the conditioning step.
△ Less
Submitted 26 March, 2015; v1 submitted 31 October, 2014;
originally announced November 2014.
-
Braidlab: A Software Package for Braids and Loops
Authors:
Jean-Luc Thiffeault,
Marko Budisic
Abstract:
Braidlab is a Matlab package for analyzing data using braids. It was designed to be fast, so it can be used on relatively large problems. It uses the object-oriented features of Matlab to provide a class for braids on punctured disks and a class for equivalence classes of simple closed loops. The growth of loops under iterated action by braids is used to compute the topological entropy of braids,…
▽ More
Braidlab is a Matlab package for analyzing data using braids. It was designed to be fast, so it can be used on relatively large problems. It uses the object-oriented features of Matlab to provide a class for braids on punctured disks and a class for equivalence classes of simple closed loops. The growth of loops under iterated action by braids is used to compute the topological entropy of braids, as well as for determining the equality of braids. This guide is a survey of the main capabilities of braidlab, with many examples; the help messages of the various commands provide more details. Some of the examples contain novel observations, such as the existence of cycles of the linear effective action for arbitrary braids.
△ Less
Submitted 15 November, 2019; v1 submitted 2 October, 2014;
originally announced October 2014.
-
Frontiers of chaotic advection
Authors:
Hassan Aref,
John R. Blake,
Marko Budišić,
Silvana S. S. Cardoso,
Julyan H. E. Cartwright,
Herman J. H. Clercx,
Kamal El Omari,
Ulrike Feudel,
Ramin Golestanian,
Emmanuelle Gouillart,
GertJan F. van Heijst,
Tatyana S. Krasnopolskaya,
Yves Le Guer,
Robert S. MacKay,
Vyacheslav V. Meleshko,
Guy Metcalfe,
Igor Mezić,
Alessandro P. S. de Moura,
Oreste Piro,
Michel F. M. Speetjens,
Rob Sturman,
Jean-Luc Thiffeault,
Idan Tuval
Abstract:
This work reviews the present position of and surveys future perspectives in the physics of chaotic advection: the field that emerged three decades ago at the intersection of fluid mechanics and nonlinear dynamics, which encompasses a range of applications with length scales ranging from micrometers to hundreds of kilometers, including systems as diverse as mixing and thermal processing of viscous…
▽ More
This work reviews the present position of and surveys future perspectives in the physics of chaotic advection: the field that emerged three decades ago at the intersection of fluid mechanics and nonlinear dynamics, which encompasses a range of applications with length scales ranging from micrometers to hundreds of kilometers, including systems as diverse as mixing and thermal processing of viscous fluids, microfluidics, biological flows, and oceanographic and atmospheric flows.
△ Less
Submitted 14 June, 2017; v1 submitted 12 March, 2014;
originally announced March 2014.
-
Applied Koopmanism
Authors:
Marko Budišić,
Ryan M. Mohr,
Igor Mezić
Abstract:
A majority of methods from dynamical systems analysis, especially those in applied settings, rely on Poincaré's geometric picture that focuses on "dynamics of states". While this picture has fueled our field for a century, it has shown difficulties in handling high-dimensional, ill-described, and uncertain systems, which are more and more common in engineered systems design and analysis of "big da…
▽ More
A majority of methods from dynamical systems analysis, especially those in applied settings, rely on Poincaré's geometric picture that focuses on "dynamics of states". While this picture has fueled our field for a century, it has shown difficulties in handling high-dimensional, ill-described, and uncertain systems, which are more and more common in engineered systems design and analysis of "big data" measurements.
This overview article presents an alternative framework for dynamical systems, based on the "dynamics of observables" picture. The central object is the Koopman operator: an infinite-dimensional, linear operator that is nonetheless capable of capturing the full nonlinear dynamics. The first goal of this paper is to make it clear how methods that appeared in different papers and contexts all relate to each other through spectral properties of the Koopman operator. The second goal is to present these methods in a concise manner in an effort to make the framework accessible to researchers who would like to apply them, but also, expand and improve them. Finally, we aim to provide a road map through the literature where each of the topics was described in detail.
We describe three main concepts: Koopman mode analysis, Koopman eigenquotients, and continuous indicators of ergodicity. For each concept we provide a summary of theoretical concepts required to define and study them, numerical methods that have been developed for their analysis, and, when possible, applications that made use of them.
The Koopman framework is showing potential for crossing over from academic and theoretical use to industrial practice. Therefore, the paper highlights its strengths, in applied and numerical contexts. Additionally, we point out areas where an additional research push is needed before the approach is adopted as an off-the-shelf framework for analysis and design.
△ Less
Submitted 30 November, 2012; v1 submitted 14 June, 2012;
originally announced June 2012.
-
Geometry of the ergodic quotient reveals coherent structures in flows
Authors:
Marko Budišić,
Igor Mezić
Abstract:
Dynamical systems that exhibit diverse behaviors can rarely be completely understood using a single approach. However, by identifying coherent structures in their state spaces, i.e., regions of uniform and simpler behavior, we could hope to study each of the structures separately and then form the understanding of the system as a whole. The method we present in this paper uses trajectory averages…
▽ More
Dynamical systems that exhibit diverse behaviors can rarely be completely understood using a single approach. However, by identifying coherent structures in their state spaces, i.e., regions of uniform and simpler behavior, we could hope to study each of the structures separately and then form the understanding of the system as a whole. The method we present in this paper uses trajectory averages of scalar functions on the state space to: (a) identify invariant sets in the state space, (b) form coherent structures by aggregating invariant sets that are similar across multiple spatial scales. First, we construct the ergodic quotient, the object obtained by mapping trajectories to the space of trajectory averages of a function basis on the state space. Second, we endow the ergodic quotient with a metric structure that successfully captures how similar the invariant sets are in the state space. Finally, we parametrize the ergodic quotient using intrinsic diffusion modes on it. By segmenting the ergodic quotient based on the diffusion modes, we extract coherent features in the state space of the dynamical system. The algorithm is validated by analyzing the Arnold-Beltrami-Childress flow, which was the test-bed for alternative approaches: the Ulam's approximation of the transfer operator and the computation of Lagrangian Coherent Structures. Furthermore, we explain how the method extends the Poincaré map analysis for periodic flows. As a demonstration, we apply the method to a periodically-driven three-dimensional Hill's vortex flow, discovering unknown coherent structures in its state space. In the end, we discuss differences between the ergodic quotient and alternatives, propose a generalization to analysis of (quasi-)periodic structures, and lay out future research directions.
△ Less
Submitted 10 April, 2012;
originally announced April 2012.
-
Conditioning moments of singular measures for entropy optimization. I
Authors:
Marko Budišić,
Mihai Putinar
Abstract:
In order to process a potential moment sequence by the entropy optimization method one has to be assured that the original measure is absolutely continuous with respect to Lebesgue measure. We propose a non-linear exponential transform of the moment sequence of any measure, including singular ones, so that the entropy optimization method can still be used in the reconstruction or approximation of…
▽ More
In order to process a potential moment sequence by the entropy optimization method one has to be assured that the original measure is absolutely continuous with respect to Lebesgue measure. We propose a non-linear exponential transform of the moment sequence of any measure, including singular ones, so that the entropy optimization method can still be used in the reconstruction or approximation of the original. The Cauchy transform in one variable, used for this very purpose in a classical context by A.\ A.\ Markov and followers, is replaced in higher dimensions by the Fantappiè transform. Several algorithms for reconstruction from moments are sketched, while we intend to provide the numerical experiments and computational aspects in a subsequent article. The essentials of complex analysis, harmonic analysis, and entropy optimization are recalled in some detail, with the goal of making the main results more accessible to non-expert readers.
Keywords: Fantappiè transform; entropy optimization; moment problem; tube domain; exponential transform
△ Less
Submitted 18 May, 2012; v1 submitted 14 March, 2012;
originally announced March 2012.
