Analysis of a bistable climate toy model with physics-based machine learning methods

Abstract

We propose a comprehensive framework able to address both the predictability of the first and of the second kind for high-dimensional chaotic models. For this purpose, we analyse the properties of a newly introduced multistable climate toy model constructed by coupling the Lorenz '96 model with a zero-dimensional energy balance model. First, the attractors of the system are identified with Monte Carlo Basin Bifurcation Analysis. Additionally, we are able to detect the Melancholia state separating the two attractors. Then, Neural Ordinary Differential Equations are applied in order to predict the future state of the system in both of the identified attractors.

arXiv comment17 pages, 7 figures. Eur. Phys. J. Spec. Top. (2021)

References in corpus (24)

• Tipping points in open systems: bifurcation, noise-induced and rate-dependent examples in the climate system
• The Physics of Climate Variability and Climate Change
• Stochastic Parameterization: Towards a new view of Weather and Climate Models
• Hybrid Forecasting of Chaotic Processes: Using Machine Learning in Conjunction with a Knowledge-Based Model
• A review of linear response theory for general differentiable dynamical systems
• Stochastic Climate Theory and Modelling
• Combining data assimilation and machine learning to emulate a dynamical model from sparse and noisy observations: a case study with the Lorenz 96 model
• Data-driven prediction of a multi-scale Lorenz 96 chaotic system using deep learning methods: Reservoir computing, ANN, and RNN-LSTM
• Edge States in the Climate System: Exploring Global Instabilities and Critical Transitions
• A Statistical Mechanical Approach for the Computation of the Climatic Response to General Forcings
• Transitions across Melancholia States in a Climate Model: Reconciling the Deterministic and Stochastic Points of View
• Thermodynamic Analysis of Snowball Earth Hysteresis Experiment: Efficiency, Entropy Production, and Irreversibility
• Equivalence of Non-Equilibrium Ensembles and Representation of Friction in Turbulent Flows: The Lorenz 96 Model
• A proof of concept for scale-adaptive parameterizations: the case of the Lorenz '96 model
• Stochastic perturbations to dynamical systems: a response theory approach
• Co-existing climate attractors in a coupled aquaplanet
• Nambu representation of an extended Lorenz model with viscous heating
• Travelling waves and their bifurcations in the Lorenz-96 model
• Response Formulae for $n$-point Correlations in Statistical Mechanical Systems and Application to a Problem of Coarse Graining
• Lyapunov analysis of multiscale dynamics: The slow bundle of the two-scale Lorenz 96 model
• Symmetries in the Lorenz-96 model
• Mechanics and Thermodynamics of a New Minimal Model of the Atmosphere
• Effects of Stochastic Parametrization on Extreme Value Statistics
• Rough basin boundaries in high dimension: Can we classify them experimentally?

Cited by in corpus (4)

• Framework for global stability analysis of dynamical systems
• Limits to predictability of the asymptotic state of the Atlantic Meridional Overturning Circulation in a conceptual climate model
• Heterogeneity of the Attractor of the Lorenz '96 Model: Lyapunov Analysis, Unstable Periodic Orbits, and Shadowing Properties
• Predictors and Predictands of Linear Response in Spatially Extended Systems