Committor Functions for Climate Phenomena at the Predictability Margin: The Example of El Niño–Southern Oscillation in the Jin and Timmermann Model

Lucente, Dario; Herbert, Corentin; Bouchet, Freddy

doi:10.1175/jas-d-22-0038.1

Cited by 13 publications

(16 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The sampling of the phase space indeed becomes difficult when it displays complex structures. This result may be related to what Lucente et al (2022a) observed in the Jin-Timmermann model: there are certain zones of the phase space where the committor function displays a complex, fine-scale structure, which we cannot expect AMC to predict accurately due to its analog approximation. Even the testing phase that uses a search through a KD tree quickly becomes very computationally expensive with dimensionality and requires a lot of training data.…”

Section: Summary and Discussionmentioning

confidence: 75%

Data-driven methods to estimate the committor function in conceptual ocean models

Jacques-Dumas

Westen

Bouchet

et al. 2023

Nonlin. Processes Geophys.

Self Cite

View full text Add to dashboard Cite

Abstract. In recent years, several climate subsystems have been identified that may undergo a relatively rapid transition compared to the changes in their forcing. Such transitions are rare events in general, and simulating long-enough trajectories in order to gather sufficient data to determine transition statistics would be too expensive. Conversely, rare events algorithms like TAMS (trajectory-adaptive multilevel sampling) encourage the transition while keeping track of the model statistics. However, this algorithm relies on a score function whose choice is crucial to ensure its efficiency. The optimal score function, called the committor function, is in practice very difficult to compute. In this paper, we compare different data-based methods (analog Markov chains, neural networks, reservoir computing, dynamical Galerkin approximation) to estimate the committor from trajectory data. We apply these methods on two models of the Atlantic Ocean circulation featuring very different dynamical behavior. We compare these methods in terms of two measures, evaluating how close the estimate is from the true committor and in terms of the computational time. We find that all methods are able to extract information from the data in order to provide a good estimate of the committor. Analog Markov Chains provide a very reliable estimate of the true committor in simple models but prove not so robust when applied to systems with a more complex phase space. Neural network methods clearly stand out by their relatively low testing time, and their training time scales more favorably with the complexity of the model than the other methods. In particular, feedforward neural networks consistently achieve the best performance when trained with enough data, making this method promising for committor estimation in sophisticated climate models.

show abstract

Section: Summary and Discussionmentioning

confidence: 75%

Data-driven methods to estimate the committor function in conceptual ocean models

Jacques-Dumas

Westen

Bouchet

et al. 2023

Nonlin. Processes Geophys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The sampling of the phase space indeed becomes difficult when it displays complex structures. This result may be related to what Lucente et al (2022a) observed in the Jin-Timmermann model: there are certain zones of the phase space where the committor function displays a complex, fine scale structure, which we cannot expect AMC to predict accurately due to its analogues appproximation. Even the testing phase, that uses a search through a KD-tree quickly becomes very computationally expensive with dimensionality and requires a lot of training data.…”

Section: Summary and Discussionmentioning

confidence: 75%

Data-driven methods to estimate the committor function in conceptual ocean models

Jacques-Dumas

Westen

Bouchet

et al. 2022

Preprint

View full text Add to dashboard Cite

Abstract. In recent years, several climate subsystems have been identified that may undergo a relatively rapid transition compared to the changes in their forcing. Such transitions are rare events in general and simulating long-enough trajectories in order to gather sufficient data to determine transition statistics would be too expensive. Conversely, rare-events algorithms like TAMS (Trajectory-Adaptive Multilevel Sampling) encourage the transition while keeping track of the model statistics. However, this algorithm relies on a score function whose choice is crucial to ensure its efficiency. The optimal score function, called committor function, is in practice very difficult to compute. In this paper, we compare different data-based methods (Analogue Markov Chains, Neural Networks, Reservoir Computing, Dynamical Galerkin Approximation) to estimate the committor from trajectory data. We apply these methods on two models of the Atlantic Ocean circulation featuring very different dynamical behavior. We compare these methods in terms of two measures, evaluating how close the estimate is from the true committor, and in terms of the computational time. We find that all methods are able to extract information from the data in order to provide a good estimate of the committor. Analogue Markov Chains provide a very reliable estimate of the true committor in simple models but prove not so robust when applied to systems with a more complex phase space. Neural network methods clearly stand out by their relatively low testing time, and their training time scales more favorably with the complexity of the model than the other methods. In particular, feedforward neural networks consistently achieve the best performance when trained with enough data, making this method promising for committor estimation in sophisticated climate models.

show abstract

“…To include noise in the model, the approach described in [21] has been followed, where the noise components D x and D y are defined as:…”

Section: Models and Observationsmentioning

confidence: 99%

Detection of limit cycle signatures of El Niño in models and observations using reservoir computing

Guardamagna,

Wieners,

Fang

et al. 2024

J. Phys. Complex.

View full text Add to dashboard Cite

While the physics of the El Niño–Southern Oscillation (ENSO) phenomenon in the Tropical Pacific is quite well understood, there is still debate on several more fundamental aspects. The focus of this paper is on one of these issues that deals with whether ENSO variability, within the recharge-discharge oscillator theory arising from a stochastic Hopf bifurcation, is subcritical or supercritical. Using a Reservoir Computing method, we develop a criticality index as an indicator for the presence of a limit cycle in noisy time series. The utility of this index is shown in three members of a hierarchy of ENSO models: a conceptual box model, the classical Zebiak-Cane model and a state-of-the-art Global Climate Model. Finally, the criticality index is determined from observations, leading to the result that ENSO variability appears to be subcritical.

show abstract

Committor Functions for Climate Phenomena at the Predictability Margin: The Example of El Niño–Southern Oscillation in the Jin and Timmermann Model

Cited by 13 publications

References 50 publications

Data-driven methods to estimate the committor function in conceptual ocean models

Data-driven methods to estimate the committor function in conceptual ocean models

Data-driven methods to estimate the committor function in conceptual ocean models

Detection of limit cycle signatures of El Niño in models and observations using reservoir computing

Contact Info

Product

Resources

About