Articles | Volume 26, issue 4
Nonlin. Processes Geophys., 26, 381–399, 2019
Nonlin. Processes Geophys., 26, 381–399, 2019

Research article 05 Nov 2019

Research article | 05 Nov 2019

Generalization properties of feed-forward neural networks trained on Lorenz systems

Sebastian Scher and Gabriele Messori

Related authors

A new view of heat wave dynamics and predictability over the eastern Mediterranean
Assaf Hochman, Sebastian Scher, Julian Quinting, Joaquim G. Pinto, and Gabriele Messori
Earth Syst. Dynam., 12, 133–149,,, 2021
Short summary
Technical Note: Temporal Disaggregation of Spatial Rainfall Fields with Generative Adversarial Networks
Sebastian Scher and Stefanie Peßenteiner
Hydrol. Earth Syst. Sci. Discuss.,,, 2020
Revised manuscript accepted for HESS
Short summary
Weather and climate forecasting with neural networks: using general circulation models (GCMs) with different complexity as a study ground
Sebastian Scher and Gabriele Messori
Geosci. Model Dev., 12, 2797–2809,,, 2019
Short summary

Related subject area

Subject: Predictability, probabilistic forecasts, data assimilation, inverse problems | Topic: Climate, atmosphere, ocean, hydrology, cryosphere, biosphere | Techniques: Simulation
Fast hybrid tempered ensemble transform filter formulation for Bayesian elliptical problems via Sinkhorn approximation
Sangeetika Ruchi, Svetlana Dubinkina, and Jana de Wiljes
Nonlin. Processes Geophys., 28, 23–41,,, 2021
Short summary
Simulating model uncertainty of subgrid-scale processes by sampling model errors at convective scales
Michiel Van Ginderachter, Daan Degrauwe, Stéphane Vannitsem, and Piet Termonia
Nonlin. Processes Geophys., 27, 187–207,,, 2020
Short summary
Data-driven versus self-similar parameterizations for stochastic advection by Lie transport and location uncertainty
Valentin Resseguier, Wei Pan, and Baylor Fox-Kemper
Nonlin. Processes Geophys., 27, 209–234,,, 2020
Short summary

Cited articles

Bakker, R., Schouten, J. C., Giles, C. L., Takens, F., and Bleek, C. M. v. d.: Learning Chaotic Attractors by Neural Networks, Neural Comput., 12, 2355–2383,, 2000. a, b
Bau, D., Zhu, J.-Y., Strobelt, H., Bolei, Z., Tenenbaum, J. B., Freeman, W. T., and Torralba, A.: GAN Dissection: Visualizing and Understanding Generative Adversarial Networks, in: Proceedings of the International Conference on Learning Representations (ICLR), 2019. a
Chattopadhyay, A., Hassanzadeh, P., Palem, K., and Subramanian, D.: Data-driven prediction of a multi-scale Lorenz 96 chaotic system using a hierarchy of deep learning methods: Reservoir computing, ANN, and RNN-LSTM, arXiv preprint arXiv:1906.08829, 2019. a, b
Dueben, P. D. and Bauer, P.: Challenges and design choices for global weather and climate models based on machine learning, Geosci. Model Dev., 11, 3999–4009,, 2018. a, b, c
Faranda, D., Messori, G., and Yiou, P.: Dynamical proxies of North Atlantic predictability and extremes, Sci. Rep., 7, 41278,, 2017. a
Short summary
Neural networks are a technique that is widely used to predict the time evolution of physical systems. For this the past evolution of the system is shown to the neural network – it is trained – and then can be used to predict the evolution in the future. We show some limitations in this approach for certain systems that are important to consider when using neural networks for climate- and weather-related applications.