Articles | Volume 11, issue 3
https://doi.org/10.5194/esd-11-835-2020
https://doi.org/10.5194/esd-11-835-2020
Research article
 | 
18 Sep 2020
Research article |  | 18 Sep 2020

Reconstructing coupled time series in climate systems using three kinds of machine-learning methods

Yu Huang, Lichao Yang, and Zuntao Fu

Related authors

Conditional diffusion models for downscaling & bias correction of Earth system model precipitation
Michael Aich, Philipp Hess, Baoxiang Pan, Sebastian Bathiany, Yu Huang, and Niklas Boers
EGUsphere, https://doi.org/10.5194/egusphere-2025-2646,https://doi.org/10.5194/egusphere-2025-2646, 2025
Short summary

Cited articles

Badin, G. and Domeisen, D. I.: A search for chaotic behavior in stratospheric variability: comparison between the Northern and Southern Hemispheres, J. Atmos. Sci., 71, 4611–4620, 2014. 
Biancofiore, F., Busilacchio, M., Verdecchia, M., Tomassetti, B., Aruffo, E., Bianco, S., and Di Carlo, P.: Recursive neural network model for analysis and forecast of PM10 and PM2.5, Atmos. Pollut. Res., 8, 652–659, 2017. 
Brown, P. J.: Measurement, Regression, and Calibration, vol. 12 of Oxford Statistical Science Series, Oxford University Press, USA, 216 pp., 1994. 
Carroll, T. L.: Using reservoir computers to distinguish chaotic series, Phys. Rev. E., 98, 052209, https://doi.org/10.1103/PhysRevE.98.052209, 2018. 
Chattopadhyay, A., Hassanzadeh, P., and Subramanian, D.: Data-driven predictions of a multiscale Lorenz 96 chaotic system using machine-learning methods: reservoir computing, artificial neural network, and long short-term memory network, Nonlin. Processes Geophys., 27, 373–389, 2020. 
Download
Short summary
We investigate the applicability of machine learning (ML) on time series reconstruction and find that the dynamical coupling relation and nonlinear causality are crucial for the application of ML. Our results could provide insights into causality and ML approaches for paleoclimate reconstruction, parameterization schemes, and prediction in climate studies.
Share
Altmetrics
Final-revised paper
Preprint