Kernel detrended fluctuation analysis: A nonlinear, multivariate method for detecting long-range persistence
Chaos: An Interdisciplinary Journal of Nonlinear Science · 01.03.2026 · 36 (3)
DOI: 10.1063/5.0316179 ↗Kurz gefasst
Fachartikel in der Zeitschrift Chaos: An Interdisciplinary Journal of Nonlinear Science, online erschienen am 18.03.2026. Verfasst von Tristan K E Williams und 6 weiteren Autor:innen. Beteiligte Einrichtung: Image Processing Laboratory, Universitat de València, València, Spain. Die Arbeit nennt 49 Literaturangaben. Gefördert von: Horizon 2020 Framework Programme, Microsoft AI, European Commission.
Abstract (Originalsprache)
We introduce Kernel Detrended Fluctuation Analysis (kDFA), a multivariate, nonlinear generalization of detrended fluctuation analysis for quantifying long-range persistence in complex systems. We show that kDFA generalizes the traditional variance-based fluctuation functional by replacing it with a kernel cross-covariance-based measure. This formulation connects the estimator to the Hilbert–Schmidt norm of the covariance operator in the reproducing kernel Hilbert space. This allows persistence to be inferred from linear to strongly nonlinear regimes via kernel learning. We showcase kDFA in synthetic and real experiments. On synthetic data, kDFA accurately retrieves Hurst exponents and generalizes the standard DFA to nonlinear cases. Comparisons of the Lorenz systems (both L63 and L96) against iterated amplitude-adjusted Fourier transform surrogates reveal genuine nonlinear persistence beyond linear autocorrelation. Applied to ecosystems, kDFA uncovers robust, long-term coupling between vegetation activity and its drivers across European vegetation sites, and detects patterns in this coupling relative to the long-term vegetation trend. kDFA, thus, provides a scalable, theory-grounded tool to uncover hidden multivariate memory in natural and engineered systems.
Autor:innen und Institutionen
- Image Processing Laboratory, Universitat de València, València, Spain.
- Institute for Earth System Science and Remote Sensing, Leipzig University, Leipzig, Germany.
- Max Planck Institute for the Physics of Complex Systems, Dresden, Germany.
ORCID-Profile: Tristan K E Williams, Homer Durand, Tobias Braun, Francesco Martinuzzi, David Montero, Miguel D Mahecha, Gustau Camps-Valls
Förderung
- Horizon 2020 Framework Programme
- Microsoft AI
- European Commission
- European Space Agency
Zitieren
APA
Williams, T. K. E., Durand, H., Braun, T., Martinuzzi, F., Montero, D., Mahecha, M. D., & Camps-Valls, G. (2026). Kernel detrended fluctuation analysis: A nonlinear, multivariate method for detecting long-range persistence. Chaos: An Interdisciplinary Journal of Nonlinear Science, 36(3). https://doi.org/10.1063/5.0316179
Vancouver
Williams TKE, Durand H, Braun T, Martinuzzi F, Montero D, Mahecha MD, et al. Kernel detrended fluctuation analysis: A nonlinear, multivariate method for detecting long-range persistence. Chaos: An Interdisciplinary Journal of Nonlinear Science. 2026;36(3). doi:10.1063/5.0316179
BibTeX
@article{Williams2026,
author = {Williams, Tristan K E and Durand, Homer and Braun, Tobias and Martinuzzi, Francesco and Montero, David and Mahecha, Miguel D and Camps-Valls, Gustau},
title = {Kernel detrended fluctuation analysis: A nonlinear, multivariate method for detecting long-range persistence},
journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science},
year = {2026},
volume = {36},
number = {3},
doi = {10.1063/5.0316179}
}
EUVIMED