Abstract
Lagrangian techniques, such as the finite-time Lyapunov exponent (FTLE) and hyperbolic Lagrangian coherent structures, have become popular tools for analyzing unsteady fluid flows. These techniques identify regions where particles transported by a flow will converge to and diverge from over a finite-time interval, even in a divergence-free flow. Lagrangian analyses, however, are time consuming and computationally expensive, hence unsuitable for quickly assessing short-term material transport. A recently developed method called Objective Eulerian Coherent Structures (OECSs) (Serra and Haller in Chaos Interdiscip J Nonlinear Sci 26(5):053110, 2016) rigorously connected Eulerian quantities to short-term Lagrangian transport. This Eulerian method is faster and less expensive to compute than its Lagrangian counterparts, and needs only a single snapshot of a velocity field. Along the same line, here we define the instantaneous Lyapunov Exponent, the instantaneous counterpart of the FTLE, and connect the Taylor series expansion of the right Cauchy-Green deformation tensor to the infinitesimal integration time limit of the FTLE. We illustrate our results on geophysical fluid flows from numerical models as well as analytical flows, and demonstrate the efficacy of attracting and repelling instantaneous Lyapunov exponent structures in predicting short-term material transport.
Similar content being viewed by others
References
Shadden, S.C., Lekien, F., Marsden, J.E.: Definition and properties of Lagrangian coherent structures from finite-time Lyapunov exponents in two-dimensional aperiodic flows. Physica D 212(3), 271–304 (2005)
Lekien, F., Shadden, S.C., Marsden, J.E.: Lagrangian coherent structures in n-dimensional systems. J. Math. Phys. 48(6), 065404 (2007)
Lekien, F., Ross, S.D.: The computation of finite-time Lyapunov exponents on unstructured meshes and for non-Euclidean manifolds. Chaos Interdiscip. J. Nonlinear Sci. 20(1), 017505 (2010)
Brunton, S.L., Rowley, C.W.: Fast computation of FTLE fields for unsteady flows: a comparison of methods. Chaos Interdiscip. J. Nonlinear Sci. 20(1), 017503 (2010)
Lipinski, D., Mohseni, K.: A ridge tracking algorithm and error estimate for efficient computation of Lagrangian coherent structures. Chaos Interdiscip. J. Nonlinear Sci. 20(1), 017504 (2010)
Ross, S.D., Tanaka, M.L., Senatore, C.: Detecting dynamical boundaries from kinematic data in biomechanics. Chaos Interdiscip. J. Nonlinear Sci. 20(1), 017507 (2010)
Senatore, C., Ross, S.D.: Detection and characterization of transport barriers in complex flows via ridge extraction of the finite time Lyapunov exponent field. Int. J. Numer. Methods Eng. 86(9), 1163–1174 (2011)
Leung, S.: An Eulerian approach for computing the finite time Lyapunov exponent. J. Comput. Phys. 230(9), 3500–3524 (2011)
Haller, G.: A variational theory of hyperbolic Lagrangian coherent structures. Physica D 240(7), 574–598 (2011)
Tallapragada, P., Ross, S.D., Schmale III, D.G.: Lagrangian coherent structures are associated with fluctuations in airborne microbial populations. Chaos Interdiscip. J. Nonlinear Sci. 21, 033122 (2011)
Schindler, B., Peikert, R., Fuchs, R., Theisel, H.: Ridge concepts for the visualization of Lagrangian coherent structures. Topol. Methods Data Anal. Vis. II, 221–235 (2012)
Schindler, B., Fuchs, R., Barp, S., Waser, J., Pobitzer, A., Carnecky, R., Matković, K., Peikert, R.: Lagrangian coherent structures for design analysis of revolving doors. IEEE Trans. Visual Comput. Graph. 18(12), 2159–2168 (2012)
Peng, J., Peterson, R.: Attracting structures in volcanic ash transport. Atmos. Environ. 48, 230–239 (2012)
Tallapragada, P., Ross, S.D.: A set oriented definition of finite-time Lyapunov exponents and coherent sets. Commun. Nonlinear Sci. Numer. Simul. 18(5), 1106–1126 (2013)
BozorgMagham, A.E., Ross, S.D., Schmale III, D.G.: Real-time prediction of atmospheric Lagrangian coherent structures based on forecast data: an application and error analysis. Physica D 258, 47–60 (2013)
Michini, M., Hsieh, M.A., Forgoston, E., Schwartz, I.B.: Robotic tracking of coherent structures in flows. IEEE Trans. Robot. 30(3), 593–603 (2014)
Garaboa-Paz, D., Eiras-Barca, J., Huhn, F., Pérez-Muñuzuri, V.: Lagrangian coherent structures along atmospheric rivers. Chaos Interdiscip. J. Nonlinear Sci. 25(6), 063105 (2015)
BozorgMagham, A.E., Ross, S.D.: Atmospheric Lagrangian coherent structures considering unresolved turbulence and forecast uncertainty. Commun. Nonlinear Sci. Numer. Simul. 22(1), 964–979 (2015)
Schmale III, D.G., Ross, S.D.: Highways in the sky: scales of atmospheric transport of plant pathogens. Annu. Rev. Phytopathol. 53, 591–611 (2015)
BozorgMagham, A.E., Ross, S.D., Schmale III, D.G.: Local finite-time Lyapunov exponent, local sampling and probabilistic source and destination regions. Nonlinear Process. Geophys. 22(6), 663–677 (2015)
Haller, G.: Lagrangian coherent structures. Annu. Rev. Fluid Mech. 47(1), 137–162 (2015)
Mease, K., Topcu, U., Aykutluğ, E., Maggia, M.: Characterizing two-timescale nonlinear dynamics using finite-time Lyapunov exponents and subspaces. Commun. Nonlinear Sci. Numer. Simul. 36, 148–174 (2016)
Hadjighasem, A., Farazmand, M., Blazevski, D., Froyland, G., Haller, G.: A critical comparison of Lagrangian methods for coherent structure detection. Chaos Interdiscip. J. Nonlinear Sci. 27(5), 053104 (2017)
Onozaki, K., Yoshimura, H., Ross, S.D.: Tube dynamics and low energy Earth-Moon transfers in the 4-body system. Adv. Space Res. 60(10), 2117–2132 (2017)
Garaboa-Paz, D., Nieves, L., Pérez-Muñuzuri, V.: Influence of finite-time Lyapunov exponents on winter precipitation over the Iberian Peninsula. Nonlinear Process. Geophys. 24(2), 227 (2017)
Schmale III, D.G., Ross, S.D.: High-flying microbes: aerial drones and chaos theory help researchers explore the many ways that microorganisms spread havoc around the world. Sci. Am. 2, 32–37 (2017)
You, G., Wong, T., Leung, S.: Eulerian methods for visualizing continuous dynamical systems using Lyapunov exponents. SIAM J. Sci. Comput. 39(2), A415–A437 (2017)
Serra, M., Sathe, P., Beron-Vera, F., Haller, G.: Uncovering the edge of the polar vortex. J. Atmos. Sci. 74(11), 3871–3885 (2017)
Chang, H., Huntley, H.S., Kirwan Jr., A., Lipphardt Jr., B., Sulman, M.H.: Transport structures in a 3D periodic flow. Commun. Nonlinear Sci. Numer. Simul. 61, 84–103 (2018)
Liu, Y., Wilson, C., Green, M.A., Hughes, C.W.: Gulf stream transport and mixing processes via coherent structure dynamics. J. Geophys. Res. Oceans 123(4), 3014–3037 (2018)
Hsieh, M.A., Hajieghrary, H., Kularatne, D., Heckman, C.R., Forgoston, E., Schwartz, I.B., Yecko, P.A.: Small and adrift with self-control: using the environment to improve autonomy. In: Robotics Research, pp. 387–402. Springer (2018)
Balasuriya, S., Ouellette, N.T., Rypina, I.I.: Generalized Lagrangian coherent structures. Physica D 372, 31–51 (2018)
Shadden, S.C., Dabiri, J.O., Marsden, J.E.: Lagrangian analysis of fluid transport in empirical vortex ring flows. Phys. Fluids 18(4), 047105 (2006)
Olcay, A.B., Pottebaum, T.S., Krueger, P.S.: Sensitivity of Lagrangian coherent structure identification to flow field resolution and random errors. Chaos 20(1), 017506 (2010)
Raben, S.G., Ross, S.D., Vlachos, P.P.: Computation of finite-time Lyapunov exponents from time-resolved particle image velocimetry data. Exp. Fluids 55(1), 1638 (2014)
Raben, S.G., Ross, S.D., Vlachos, P.P.: Experimental determination of three-dimensional finite-time Lyapunov exponents in multi-component flows. Exp. Fluids 55(10), 1824 (2014)
Tanaka, M.L., Ross, S.D.: Separatrices and basins of stability from time series data: an application to biodynamics. Nonlinear Dyn. 58(1–2), 1–21 (2009)
Tanaka, M.L., Nussbaum, M.A., Ross, S.D.: Evaluation of the threshold of stability for the human spine. J. Biomech. 42(8), 1017–1022 (2009)
Tanaka, M.L., Ross, S.D., Nussbaum, M.A.: Mathematical modeling and simulation of seated stability. J. Biomech. 43(5), 906–912 (2010)
Shadden, S.C., Taylor, C.A.: Characterization of coherent structures in the cardiovascular system. Ann. Biomed. Eng. 36(7), 1152–1162 (2008)
Shadden, S.C., Astorino, M., Gerbeau, J.-F.: Computational analysis of an aortic valve jet with Lagrangian coherent structures. Chaos Interdiscip. J. Nonlinear Sci. 20(1), 017512 (2010)
Töger, J., Kanski, M., Carlsson, M., Kovács, S.J., Söderlind, G., Arheden, H., Heiberg, E.: Vortex ring formation in the left ventricle of the heart: analysis by 4D flow MRI and Lagrangian coherent structures. Ann. Biomed. Eng. 40(12), 2652–2662 (2012)
Shadden, S.C., Arzani, A.: Lagrangian postprocessing of computational hemodynamics. Ann. Biomed. Eng. 43(1), 41–58 (2015)
Arzani, A., Gambaruto, A.M., Chen, G., Shadden, S.C.: Wall shear stress exposure time: a Lagrangian measure of near-wall stagnation and concentration in cardiovascular flows. Biomech. Model. Mechanobiol. 16(3), 787–803 (2017)
Shadden, S.C., Lekien, F., Paduan, J.D., Chavez, F.P., Marsden, J.E.: The correlation between surface drifters and coherent structures based on high-frequency radar data in Monterey Bay. Deep Sea Res. Part II Top. Stud. Oceanogr. 56(3–5), 161–172 (2009)
Rypina, I.I., Pratt, L.J., Lozier, M.S.: Near-surface transport pathways in the North Atlantic Ocean: looking for throughput from the subtropical to the subpolar gyre. J. Phys. Oceanogr. 41(5), 911–925 (2011)
Kirincich, A.: Remote sensing of the surface wind field over the coastal ocean via direct calibration of HF radar backscatter power. J. Atmos. Ocean. Technol. 33(7), 1377–1392 (2016)
D’Asaro, E.A., Shcherbina, A.Y., Klymak, J.M., Molemaker, J., Novelli, G., Guigand, C.M., Haza, A.C., Haus, B.K., Ryan, E.H., Jacobs, G.A., Huntley, H.S., Laxague, N.J.M., Chen, S., Judt, F., McWilliams, J.C., Barkan, R., Kirwan, A.D., Poje, A.C., Özgökmen, T.M.: Ocean convergence and the dispersion of flotsam. Proc. Nat. Acad. Sci. 115(6), 1162–1167 (2018)
Nolan, P.J., Pinto, J., González-Rocha, J., Jensen, A., Vezzi, C., Bailey, S., de Boer, G., Diehl, C., Laurence, R., Powers, C., Ross, S.D., Schmale III, D.G.: Coordinated unmanned aircraft system (UAS) and ground-based weather measurements to predict Lagrangian coherent structures (LCSs). Sensors 18(12), 4448 (2018)
González-Rocha, J., Woolsey, C.A., Sultan, C., De Wekker, S.F.J.: Sensing wind from quadrotor motion. J. Guidance Control Dyn. 42(4), 836–852 (2018). https://doi.org/10.2514/1.G003542
Van Dop, H., Addis, R., Fraser, G., Girardi, F., Graziani, G., Inoue, Y., Kelly, N., Klug, W., Kulmala, A., Nodop, K., Pretel, J.: ETEX: a European tracer experiment; observations, dispersion modelling and emergency response. Atmos. Environ. 32(24), 4089–4094 (1998)
De Dominicis, M., Falchetti, S., Trotta, F., Pinardi, N., Giacomelli, L., Napolitano, E., Fazioli, L., Sorgente, R., Haley Jr., P.J., Lermusiaux, P.F., Martins, F., Cocco, M.: A relocatable ocean model in support of environmental emergencies. Ocean Dyn. 64(5), 667–688 (2014)
Xie, X., Nolan, P.J., Ross, S.D., Iliescu, T.: Lagrangian data-driven reduced order modeling of finite time Lyapunov exponents (2019). arXiv preprint arXiv:1808.05635
Buesseler, K.O., Jayne, S.R., Fisher, N.S., Rypina, I.I., Baumann, H., Baumann, Z., Breier, C.F., Douglass, E.M., George, J., Macdonald, A.M., et al.: Fukushima-derived radionuclides in the ocean and biota off Japan. Proc. Natl. Acad. Sci. 109(16), 5984–5988 (2012)
Mezić, I., Loire, S., Fonoberov, V.A., Hogan, P.: A new mixing diagnostic and Gulf oil spill movement. Science 330(6003), 486–489 (2010)
Olascoaga, M.J., Haller, G.: Forecasting sudden changes in environmental pollution patterns. Proc. Natl. Acad. Sci. 109(13), 4738–4743 (2012)
García-Garrido, V., Ramos, A., Mancho, A., Coca, J., Wiggins, S.: A dynamical systems perspective for a real-time response to a marine oil spill. Mar. Pollut. Bull. 112(1–2), 201–210 (2016)
Allshouse, M.R., Ivey, G.N., Lowe, R.J., Jones, N.L., Beegle-Krause, C., Xu, J., Peacock, T.: Impact of windage on ocean surface Lagrangian coherent structures. Environ. Fluid Mech. 17(3), 473–483 (2017)
Breivik, Ø., Allen, A.A., Maisondieu, C., Roth, J.C.: Wind-induced drift of objects at sea: the leeway field method. Appl. Ocean Res. 33(2), 100–109 (2011)
Peacock, T., Haller, G.: Lagrangian coherent structures: the hidden skeleton of fluid flows. Phys. Today 66(2), 41 (2013)
Serra, M., Sathe, P., Rypina, I., Kirincich, A., Ross, S.D., Lermusiaux, P., Allen, A., Peacock, T., Haller, G.: Search and rescue at sea aided by hidden flow structures. Nat. Commun. 11, 2525 (2020)
Draxler, R.R., Hess, G.: An overview of the HYSPLIT\_4 modelling system for trajectories. Aust. Meteorol. Mag. 47(4), 295–308 (1998)
Rypina, I.I., Kirincich, A.R., Limeburner, R., Udovydchenkov, I.A.: Eulerian and Lagrangian correspondence of high-frequency radar and surface drifter data: effects of radar resolution and flow components. J. Atmos. Ocean. Technol. 31(4), 945–966 (2014)
Ameli, S., Shadden, S.C.: A transport method for restoring incomplete ocean current measurements. J. Geophys. Res. Oceans 124(1), 227–242 (2019)
Serra, M., Haller, G.: Objective Eulerian coherent structures. Chaos Interdiscip. J. Nonlinear Sci. 26(5), 053110 (2016)
Serra, M., Haller, G.: Forecasting long-lived Lagrangian vortices from their objective Eulerian footprints. J. Fluid Mech. 813, 436–457 (2017)
Serra, M., Haller, G.: Efficient computation of null geodesics with applications to coherent vortex detection. In: Proc. R. Soc. A, 2199, 20160807. The Royal Society (2017)
Nave Jr., G.K., Nolan, P.J., Ross, S.D.: Trajectory-free approximation of phase space structures using the trajectory divergence rate. Nonlinear Dyn. 96, 685–702 (2019)
Nolan, P.J., McClelland, H.G., Woolsey, C.A., Ross, S.D.: A method for detecting atmospheric Lagrangian coherent structures using a single fixed-wind unmanned aircraft system. Sensors 19(7), 1607 (2019)
Serra, M., Vétel, J., Haller, G.: Exact theory of material spike formation in flow separation. J. Fluid Mech. 845, 51–92 (2018)
Serra, M., Crouzat, S., Simon, G., Vétel, J., Haller, G.: Material spike formation in highly unsteady separated flows. J. Fluid Mech. (2019). arXiv:1903.10371
Klose, B.F., Jacobs, G.B., Serra, M.: Kinematics of Lagrangian flow separation in external aerodynamics. AIAA J. 58(5), 1926–1938 (2020)
Du Toit, P.C., Marsden, J.E.: Horseshoes in hurricanes. J. Fixed Point Theory Appl. 7(2), 351–384 (2010)
Ameli, S., Desai, Y., Shadden, S.C.: Development of an efficient and flexible pipeline for Lagrangian coherent structure computation. In: Topological Methods in Data Analysis and Visualization III, pp. 201–215. Springer (2014)
Rypina, I.I., Scott, S., Pratt, L.J., Brown, M.G.: Investigating the connection between complexity of isolated trajectories and Lagrangian coherent structures. Nonlinear Process. Geophys. 18, 977–987 (2011)
Pratt, L.J., Rypina, I.I., Özgökmen, T.M., Wang, P., Childs, H., Bebieva, Y.: Chaotic advection in a steady, three-dimensional, Ekman-driven eddy. J. Fluid Mech. 738, 143–183 (2014)
Truesdell, C., Noll, W.: Handbuch der Physik. Springer, Berlin (1965)
Weiss, J.: The dynamics of enstrophy transfer in two-dimensional hydrodynamics. Physica D 48, 273–294 (1991)
Dresselhaus, E., Tabor, M.: The kinematics of stretching and alignment of material elements in general flow fields. J. Fluid Mech. 236, 415–444 (1992)
Koh, T.-Y., Legras, B.: Hyperbolic lines and the stratospheric polar vortex. Chaos 12(2), 382–394 (2002)
Pérez-Munuzuri, V., Huhn, F.: Path-integrated Lagrangian measures from the velocity gradient tensor. Nonlinear Process. Geophys. 20(6), 987–991 (2013)
Stewart, H.: Kinematics and dynamics of fluid flow. In: Handbook of Meteorology, pp. 412–419. McGraw-Hill (1945)
Saucier, W.J.: Horizontal deformation in atmospheric motion. Trans. Am. Geophys. Union 34(5), 709–719 (1953)
Okubo, A.: Horizontal dispersion of floatable particles in the vicinity of velocity singularities such as convergences. Deep Sea Res. Oceanogr. Abstr. 17(3), 445–454 (1970)
Chelton, D.B., Schlax, M.G., Samelson, R.M.: Global observations of nonlinear mesoscale eddies. Prog. Oceanogr. 91(2), 167–216 (2011)
Vortmeyer-Kley, R., Gräwe, U., Feudel, U.: Detecting and tracking eddies in oceanic flow fields: a Lagrangian descriptor based on the modulus of vorticity. Nonlinear Process. Geophys. 23(4), 159–173 (2016)
Peikert, R., Schindler, B., Carnecky, R.: Ridge surface methods for the visualization of Lagrangian coherent structures. In: Proceedings of the Ninth International Conference on Flow Dynamics, Sendai, Japan, pp. 206–207 (2012)
Global Real-Time Ocean Forecast System. https://polar.ncep.noaa.gov/global/. Accessed 12 Aug 2019
Weather Research and Forecasting Model. https://www.mmm.ucar.edu/weather-research-and-forecasting-model. Accessed 2 Feb 2019
Schmale, D.G., Ross, S.D., Fetters, T.L., Tallapragada, P., Wood-Jones, A.K., Dingus, B.: Isolates of Fusarium graminearum collected 40–320 meters above ground level cause Fusarium head blight in wheat and produce trichothecene mycotoxins. Aerobiologia 28(1), 1–11 (2012)
Barbieri, L.K., Kral, S.T., Bailey, S.C.C., Frazier, A.E., Jacob, J.D., Brus, D., Chilson, P.B., Crick, C., Elston, J., Foroutan, H., González-Rocha, J., Greene, B.R., Guzman, M.I., Houston, A.L., Islam, A., Kemppinen, O., Pillar-Little, E.A., Reuder, J., Ross, S.D., Sama, M., Schmale III, D.G., Schuyler, T.J., Smith, S., Waugh, S., Doddi, A., Lawrence, D., Dixon, C., Borenstein, S., de Boer, G.: Intercomparison of small unmanned aircraft system (sUAS) measurements for atmospheric science during the LAPSE-RATE campaign. Sensors 19, 2179 (2019)
Arnold, V.I.: Sur la topologie des écoulements stationnaires des fluides parfaits, pp. 15–18. Springer, Berlin (2014)
Dombre, T., Frisch, U., Greene, J.M., Hénon, M., Mehr, A., Soward, A.M.: Chaotic streamlines in the ABC flows. J. Fluid Mech. 167, 353–391 (1986)
Henon, M.: Sur la topologie des lignes de courant dans un cas particulier. Comptes Rendus Acad. Sci. Paris A 262, 312–314 (1966)
Gómez, G., Koon, W.S., Lo, M.W., Marsden, J.E., Masdemont, J., Ross, S.D.: Connecting orbits and invariant manifolds in the spatial restricted three-body problem. Nonlinearity 17(5), 1571 (2004)
Farrés, A., Jorba, A.: On the high order approximation of the centre manifold for ODEs. Discrete Contin. Dyn. Syst. Ser. B 14(3), 977–1000 (2010)
Koon, W.S., Lo, M.W., Marsden, J.E., Ross, S.D. Dynamical Systems, the Three-Body Problem and Space Mission Design. Marsden Books. ISBN 978-0-615-24095-4 (2011)
Nave Jr., G.K., Ross, S.D.: Global phase space structures in a model of passive descent. Commun. Nonlinear Sci. Numer. Simul. 77, 54–80 (2019)
Doedel, E.J., Krauskopf, B., Osinga, H.M.: Global organization of phase space in the transition to chaos in the Lorenz system. Nonlinearity 28(11), R113–R139 (2015)
Gabern, F., Koon, W.S., Marsden, J.E., Ross, S.D.: Theory and computation of non-RRKM lifetime distributions and rates in chemical systems with three or more degrees of freedom. Physica D 211(3–4), 391–406 (2005)
Surana, A., Jacobs, G.B., Grunberg, O., Haller, G.: An exact theory of three-dimensional fixed separation in unsteady flows. Phys. Fluids 20(10), 107101 (2008)
Ross, S.D., BozorgMagham, A.E., Naik, S., Virgin, L.N.: Experimental validation of phase space conduits of transition between potential wells. Phys. Rev. E 98(5), 052214 (2018)
Truesdell, C., Noll, W.: The Non-linear Field Theories of Mechanics. Springer, Berlin (2004)
Acknowledgements
We thank Hosein Foroutan for providing us with WRF model data, Nicole Abaid for assistance with the perturbation analysis, Gary Nave for fruitful conversations related to some of the ideas covered in this work during their nascent stage, and Siavash Ameli for developing and maintaining the TRACE server used for the geophysical particle integration calculations. M.S. acknowledges support from the Schmidt Science Fellowship and the Postdoc Mobility Fellowship from the Swiss National Science Foundation.
Funding
This research was supported in part by grants from the National Science Foundation (NSF) under grant numbers AGS 1520825 (Hazards SEES: Advanced Lagrangian Methods for Prediction, Mitigation and Response to Environmental Flow Hazards) and DMS 1821145 (Data-Driven Computation of Lagrangian Transport Structure in Realistic Flows). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsors.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendices
Expansion of the right Cauchy-Green tensor in the integration time
For tensor fields in what follows, the dependence on \({\mathbf {x}}_0\) and \(t_0\) will be notationally dropped for clarity, as it will be understood. For small integration time \(T=t-t_0\), the right Cauchy-Green tensor, \({\mathbf {C}}\), may be expanded, as in [65, 68, 77], in terms of the integration time T,
where the dependence on the initial position and time is understood. Because all derivatives are evaluated at \(T=0\), \(\left. \tfrac{\mathrm{d}}{\mathrm{d}t}\right| _{t=t_0}=\left. \tfrac{\mathrm{d}}{\mathrm{d}T}\right| _{T=0}\). The first term on the right denotes the situation of no deformation, therefore, \({\mathbf {C}}|_{T=0}=\mathbb {1}\). The derivatives of the right Cauchy-Green tensor are given to any order by the Rivlin-Ericksen tensors [68, 77, 103],
For small \(|T|\ll 1\), the leading order behavior is given by the first Rivlin-Ericksen tensor \((\nabla {\mathbf {v}}+(\nabla {\mathbf {v}})^{\top })\), which is twice \({\mathbf {S}}\), from (7). The second-order term is,
where \({\mathbf {B}}\) is the same as given in (11).
The third-order term is,
where \({\mathbf {Q}}\) is the same as given in (13).
The expansion of the right Cauchy-Green tensor (46) can be written as,
which is a form convenient for matrix perturbation analysis, as in Appendix “Eigenvalues of the Taylor-expanded right Cauchy-Green tensor”.
Details of approximating the FTLE to second-order in integration time
Note the following general result for the eigenvalues,
of \(n \times n\) real symmetric matrices \({\mathbf {A}}\). Here, we use \(\lambda _{-}({\mathbf {A}})\) and \(\lambda _{+}({\mathbf {A}})\) as shorthand for \(\lambda _{\mathrm{min}}({\mathbf {A}})\) and \(\lambda _{\mathrm{max}}({\mathbf {A}})\), the minimum and maximum eigenvalues of \({\mathbf {A}}\), respectively. For scalar \(c \ne 0\),
where,
See Appendix “Proof of equation (52)” for the proof.
In (6), \(\lambda _{n} = \lambda _{+}({\mathbf {C}}_{t_{0}}^{t}({\mathbf {x}}))\). For small \(T>0\), where the \({\mathcal {O}}(T^2)\) and higher terms can be neglected,
Thus,
in the limit of small T using the Taylor expansion, \(\log (1+\delta ) = \delta + {\mathcal {O}}(\delta ^2)\) for small \(|\delta |\).
From (6), and noting that \(|T|=T\) for \(T>0\),
Therefore, the maximum eigenvalue of \({\mathbf {S}}(\mathbf{x} ,t_0)\) is the limit of the FTLE value for forward time as \(T \rightarrow 0^{+}\).
For \(T<0\) with small T,
Thus,
in the limit of small T.
From (6), and noting that \(|T|=-T\) for \(T<0\),
Therefore, the negative of the minimum eigenvalue of \({\mathbf {S}}(\mathbf{x} ,t_0)\) is the limit of the FTLE value for backward time as \(T \rightarrow 0^{-}\).
Consider now the third term, the order \(T^2\) term, in the expansion (10) of the right Cauchy-Green tensor. Then, (54) becomes,
Note that \({\mathbf {B}}(\mathbf{x} ,t_0)\), like \({\mathbf {S}}(\mathbf{x} ,t_0)\), is symmetric.
Below, we adopt the notation of \(s_-\) and \(s_+\) for \(s_1\) and \(s_n\), respectively, as in the main text.
It can be shown via matrix perturbation techniques (see Appendix “Eigenvalues of the Taylor-expanded right Cauchy-Green tensor”) that,
Using the Taylor expansion \(\log (1+\delta ) = \delta -\tfrac{1}{2}\delta ^2 + \tfrac{1}{3}\delta ^3 +{\mathcal {O}}(\delta ^4)\) for small \(|\delta |\), by a similar argument as before, for small T,
Therefore, for \(T>0\) with small |T|,
And similarly, for \(T<0\) with small |T|,
If we continue this procedure to obtain the approximate forward and backward FTLE through second order in T, we get (14).
Proof of equation (52)
Let \({\mathbf {A}}\) be an \(n \times n\) matrix, \(\lambda \) an eigenvalue of \({\mathbf {A}}\), \(\varvec{\xi }\) the corresponding eigenvector of \({\mathbf {A}}\), \(\mathbb {1}\) the \(n \times n\) identity matrix and \(\zeta \in {\mathbb {C}}\). By the definition of an eigenvalue \({\mathbf {A}}\varvec{\xi }\) = \(\lambda \varvec{\xi }\), we have,
Therefore, if \(\lambda \) is an eigenvalue of \({\mathbf {A}}\) with eigenvector \(\varvec{\xi }\), then \((\zeta +\lambda )\) is an eigenvalue of \(\zeta \,\mathbb {1}+{\mathbf {A}}\) with the same eigenvector \(\varvec{\xi }\). In particular, this holds when \(\zeta =1\), as in (52).
Equality of the eigenvectors of S and C as integration time goes to zero
Let \(T>0\) be small enough that the relationships in (10) and (54) hold and \({\mathcal {O}}(T^2)\) terms are negligible. As before, let \(\varvec{e}_{i}\) be the eigenvector of \({\mathbf {S}}\) associated with \(s_{i}\), then,
where the dependence on \({\mathbf {x}}\) and \(t_0\) is understood and we used the order-T approximation for \({\mathbf {C}}\). But from (4) and (5),
thus,
that is, if \(\varvec{e}_{i}\) is an eigenvector of \({\mathbf {S}}\), then it is also an eigenvector of \({\mathbf {C}}\) in the limit as \(T\rightarrow 0\). Now, assuming that \(\varvec{\xi }_{\lambda _i}\) is the eigenvector of \({\mathbf {C}}\) associated with \(\lambda _{i}\), and working through (66-69) in reverse proves that if \(\varvec{\xi }_{\lambda _i}\) is an eigenvector of \({\mathbf {C}}\), then it is also an eigenvector of \({\mathbf {S}}\) in the limit as T goes to 0. For \(T<0\) an analogous argument holds using (57) in place of (54) and with the ordering of the eigenvalues opposed, i.e., \(\lambda _i \sim s_{n-i+1},\,i\in \{1,\ldots ,n\}\).
Therefore, in the limit as |T| goes to 0, the eigenvectors of \({\mathbf {S}}\) and \({\mathbf {C}}\) are equal. For small |T|, we can also use the perturbation expansion of \({\mathbf {C}}\), to get the estimated eigenvectors of \({\mathbf {C}}\) from (73) in Appendix “Eigenvalues of the Taylor-expanded right Cauchy-Green tensor”, which provides the eigenvectors through order \(T^2\) using only the velocity field \({\mathbf {v}}\) from (1) evaluated at \({\mathbf {x}}\) and time \(t_0\) as well as appropriate derivatives.
Eigenvalues of the Taylor-expanded right Cauchy-Green tensor
Let \({\mathbf {S}}\) be a real, symmetric \(n \times n\) matrix with n distinct eigenvalues, and let \({\mathbf {B}}\) and \({\mathbf {Q}}\) also be real, symmetric \(n \times n\) matrices. We seek the eigenvalues of,
a perturbation of \({\mathbf {S}}\), where \(|\varepsilon |\) is a small scalar. In our case of interest, from (50), the small parameter is \(\varepsilon = \tfrac{1}{2}T\).
Consider the eigenvalue \(\mu _0\) of \({\mathbf {S}}\) with corresponding normalized eigenvector \(\varvec{\xi }_0\). Let’s refer to the perturbed eigenvalue and corresponding perturbed eigenvector of \({\mathbf {S}}_{\varepsilon }\) as \(\mu _{\varepsilon }\) and \(\varvec{\xi }_{\varepsilon }\). One can expand \(\varvec{\xi }_{\varepsilon }\) and \(\mu _{\varepsilon }\) in powers of \(\varepsilon \) as
The eigenvector equation, \({\mathbf {S}}_{\varepsilon }\varvec{\xi }_{\varepsilon } = \mu _{\varepsilon } \varvec{\xi }_{\varepsilon }\), can be approximated as
which leads to the following three expressions, corresponding to the order one terms, order \(\varepsilon \), and order \(\varepsilon ^2\) terms, respectively,
Multiply (77) by \(\varvec{\xi }_{0}^{\top }\) to get,
Since \(\varvec{\xi }_{0}\) is normalized, \(\varvec{\xi }_{0}^{\top }\varvec{\xi }_{0}=1\). Also, since \({\mathbf {S}}\) is symmetric,
where (76) was used. Now (79) is,
Thus,
which, since \({\mathbf {B}}\) is symmetric, represents a quadratic form.
A bound can be put on the term \(\varvec{\xi }_{0}^{\top }{\mathbf {B}} \varvec{\xi }_{0}\), noting that \(\varvec{\xi }_{0}\) is a unit vector. If \(b_n\) is the maximum eigenvalue of \({\mathbf {B}}\), then,
Similarly, if \(b_1\) is the minimum eigenvalue of \({\mathbf {B}}\), then,
So,
So (74) becomes,
where \(\mu _1\) is from (82).
With \(\mu _1\) in hand, \(\varvec{\xi }_{1}\) can also be determined as the solution of the following re-arranged version of (77),
Note that \(({\mathbf {S}} - \mu _0 \mathbb {1})\) is not invertible as it has zero determinant, since \(\mu _0\) is an eigenvalue of \({\mathbf {S}}\). The null space of \(({\mathbf {S}} - \mu _0 \mathbb {1})\) is span\(\{\varvec{\xi }_{0}\}\). Note that (87) is of the form \({\mathbf {A}}{\mathbf {x}}={\mathbf {b}}\) with a square matrix \({\mathbf {A}}\) of nullity 1 and a vector \({\mathbf {b}}\) which is in the image of \({\mathbf {A}}\), as shown below.
Note that, as a consequence of (82), the vector \({\mathbf {B}}\varvec{\xi }_{0}\) can be written as,
where \(d \in \mathbb {R}\) and \(\varvec{\xi }_{0}^{\prime \perp }\) is, in general, a vector in \(\mathrm{im}({\mathbf {S}} - \mu _0 \mathbb {1})\). This equation can be re-arranged to yield,
The left-hand side of (89) is the same as the right-hand side of (87), which means the right-hand side of (87) is a vector \({\mathbf {b}}\) which is in \(\mathrm{im}({\mathbf {S}} - \mu _0 \mathbb {1})\), which we will use below.
One can determine \(\mu _2\) by multiplying (78) by \(\varvec{\xi }_{0}^{\top }\) to get, by a similar procedure as before,
Canceling the identical terms on both sides, we get,
But take the transpose and,
where (87) was used. One can write \(\varvec{\xi }_{1}\) as,
where \(a,b \in \mathbb {R}\) and \(\varvec{\xi }_{0}^{\perp } \in \mathrm{im}({\mathbf {S}} - \mu _0 \mathbb {1})\), which is, in general, not equal to \(\varvec{\xi }_{0}^{\prime \perp }\) from (89). Hence,
Therefore, the only part of \(\varvec{\xi }_{1}\) which contributes to \(\mu _2\) is the part which is in \(\mathrm{im}({\mathbf {S}} - \mu _0 \mathbb {1})\).
When dealing with a two-dimensional flow field, \(\mathrm{im}({\mathbf {S}} - \mu _0 \mathbb {1})\) is just a 1-dimensional subspace of \(\mathbb {R}^2\), and thus \(\varvec{\xi }_{0}^{\prime \perp }\) in (89) is parallel to \(\varvec{\xi }_{0}^{\perp }\) in (94). Without loss of generality, they can be taken to be equal unit vectors, \(\varvec{\xi }_{0}^{\perp }=\varvec{\xi }_{0}^{\prime \perp }\). Thus, (87) becomes,
or, assuming \(b\ne 0\),
which is an eigenvector equation for the matrix \(({\mathbf {S}} - \mu _0 \mathbb {1})\) with the eigenvector \(\varvec{\xi }_{0}^{\perp }\) and corresponding eigenvalue \({\bar{\mu }}=- \tfrac{d}{b}\). Note that if \(b=0\), then \(d=0\) also, from (96).
For two-dimensional flows, from \(\varvec{\xi }_{0}\), one can easily obtain \(\varvec{\xi }_{0}^{\perp }\) from a \(90^{\circ }\) counterclockwise rotation,
where,
Now, \(\varvec{\xi }_{0}^{\perp }\) can be used to obtain \({\bar{\mu }}\) from (97) for the case \(d\ne 0\). With (98) in (97), (97) becomes the following eigenvector equation for \({\mathbf {R}}^{\top }({\mathbf {S}} - \mu _0 \mathbb {1}){\mathbf {R}}\) with eigenvector \(\varvec{\xi }_{0}\),
Therefore \({\bar{\mu }}\) is obtained by taking the dot product with \(\varvec{\xi }_{0}\),
and d is obtained from (89), noting that \(\varvec{\xi }_{0}^{\perp {\top }}\varvec{\xi }_{0}=0\),
Thus, (95), for two-dimensional systems, simplifies to,
where \({\bar{\mu }}\) and d are from (101) and (102), respectively.
Details for the examples
1.1 Details for nonlinear saddle example
Writing the \(\log \) term of (35) as follows, using Taylor series approximations for small |T|, we have,
So the backward FTLE is expanded in T as follows, obtained by dividing by \(-2T\),
which is the same as (36).
The FTLE can be approximated by the first, second, and third terms (the zeroth-order, first-order, and second-order in T, respectively) using the procedure outlined in Sect. 3. The gradient of the velocity is,
which is also \({\mathbf {S}}({\mathbf {x}}_0)\), since the gradient is diagonal. This has a minimum eigenvalue \(s_{-} = - 1 + 3y_0^2\), the negative of which matches the first term of (105), as prescribed by (14). To calculate the second term of (105), the term first-order in T, the acceleration field needs to be calculated and then \({\mathbf {B}}({\mathbf {x}}_0)\). The acceleration field is, following (12),
Therefore (11) gives,
The normalized eigenvector of \({\mathbf {S}}({\mathbf {x}}_0)\) corresponding to \(s_-\) is simply \(\varvec{e}_{-}=[ 0, 1]^{\top }\), which, via (16), yields,
hence,
the negative of which matches the T coefficient of the second term of (105), as prescribed by (14).
For the term second-order in T, note that, as prescribed by (13),
and since (87) implies that \(\varvec{\xi }_{1-}\) is parallel to \(\varvec{e}_-\), (16) yields,
According to (14), the second-order term is,
which matches the \(T^2\) term of the true FTLE field (105).
1.2 Details for the time-varying double-gyre example
The gradient tensor for the double-gyre velocity field (40) is,
The acceleration field, \({\mathbf {a}}=\tfrac{d}{dt}{\mathbf {v}}=(a_x,a_y)\), for the double-gyre, (40), is given by,
where the dependence of the function f, from (39), is understood.
The components of the symmetric \({\mathbf {B}}\) matrix are,
The eigenvalue \(s_-({\mathbf {x}}_0,t_0)\) of \({\mathbf {S}}({\mathbf {x}}_0,t_0)\) is,
The normalized eigenvector of \({\mathbf {S}}({\mathbf {x}}_0,t_0)\) corresponding to the eigenvalue \(s_-({\mathbf {x}}_0,t_0)\) is given by,
where,
The coefficient of T in the approximation of the backward-time FTLE for the double-gyre is thus given by \(s_-^2 - \tfrac{1}{2} \varvec{e}_{-}^{\top } {\mathbf {B}} \varvec{e}_{-}\) which can be expressed in terms of,
using the above formulas. This yields a backward-time FTLE approximation for small backward times \(T<0\) of,
Note that the first and second terms have explicit dependence on both initial position and initial time.
1.3 Details for the ABC flow example
For the ABC velocity field (45), the characteristic polynomial for the rate-of-strain tensor \({\mathbf {S}}\) for this system is,
where
The repulsion and attraction rate fields, \(s_+\) and \(s_-\), are given by
where the dependence on initial position \({\mathbf {x}}\) is understood and \(\rho \) and \(\theta \) are given by,
where,
Rights and permissions
About this article
Cite this article
Nolan, P.J., Serra, M. & Ross, S.D. Finite-time Lyapunov exponents in the instantaneous limit and material transport. Nonlinear Dyn 100, 3825–3852 (2020). https://doi.org/10.1007/s11071-020-05713-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-020-05713-4