Abstract:Real-world non-autonomous systems are open, out-of-equilibrium systems that evolve in and are driven by temporally varying environments. Such systems can show multiple timescale and transient dynamics together with transitions to very different and, at times, even disastrous dynamical regimes. Since such critical transitions disrupt the systems' intended or desired functionality, it is crucial to understand the underlying mechanisms, to identify precursors of such transitions and to reliably detect them in time series of suitable system observables to enable forecasts. This review critically assesses the various steps of investigation involved in time-series-analysis-based detection of critical transitions in real-world non-autonomous systems: from the data recording to evaluating the reliability of offline and online detections. It will highlight pros and cons to stimulate further developments, which would be necessary to advance understanding and forecasting nonlinear behavior such as critical transitions in complex systems.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to detect critical transitions through time - series analysis in non - autonomous systems in the real world. These systems are usually open, far from equilibrium, and affected by the environment that changes over time. These systems may exhibit multi - time - scale and transient dynamics, as well as transitions to very different or even catastrophic dynamic states. Such critical transitions will disrupt the expected or desired functions of the system. Therefore, it is crucial to understand the underlying mechanisms, identify early - warning indicators, and reliably detect these transitions in the time series observable in appropriate systems for prediction. The critical transitions mentioned in the paper refer to sudden changes in the state of complex dynamic systems, usually accompanied by extreme events that may lead to the loss of life or matter. Reliably detecting and predicting critical transitions can allow the development of adaptation and/or mitigation strategies, which is extremely important in many scientific fields. Although we still lack a comprehensive understanding of the mechanisms behind critical transitions, there is a large amount of observational data that may be suitable for transition detection based on time - series analysis. Successfully detecting and predicting critical transitions requires not only an understanding of various concepts and methods, but also an assessment of possible sources of error. To achieve this goal, the paper reviews recent conceptual and methodological developments aimed at studying critical transitions in real - world non - autonomous systems through time - series analysis. It critically evaluates each step in the investigation process - from experimental observation to dealing with non - stationarity, time - series analysis, off - line transition detection and its reliability assessment, and emphasizes the areas that are being actively researched and are expected to provide new insights. In addition, the paper also briefly discusses the possibility of on - line detection of critical transitions based on time - series analysis, which helps to further improve the reliability of prediction.

Time-series-analysis-based detection of critical transitions in real-world non-autonomous systems

Time-series-analysis-based detection of critical transitions in real-world non-autonomous systems

Real-time Detector for Time-variant Oscillation with Modified Intrinsic Time-scale Decomposition

Anticipating critical transitions in multidimensional systems driven by time- and state-dependent noise

Quantifying limits to detection of early warning for critical transitions

Bayesian on-line anticipation of critical transitions

Catching Critical Transition in Engineered Systems

A mathematical framework for critical transitions: bifurcations, fast-slow systems and stochastic dynamics

Early warning for critical transitions using machine-based predictability

Online data-driven changepoint detection for high-dimensional dynamical systems

Early Warning Signals for Critical Transitions: A Generalized Modeling Approach

Adaptable reservoir computing: A paradigm for model-free data-driven prediction of critical transitions in nonlinear dynamical systems

Learning from the past: predicting critical transitions with machine learning trained on surrogates of historical data

Early warning signs for saddle-escape transitions in complex networks

Early Warnings for Multistage Transitions in Dynamics on Networks

Towards a Critical Transition Theory under Different Temporal Scales and Noise Strengths.

Early warning signs of critical transitions -- The $\alpha$-stable case

Early warning indicators via latent stochastic dynamical systems

Visibility graphs of critical and off-critical time series for absorbing state phase transitions

Deep Learning for predicting rate-induced tipping

Predicting transitions across macroscopic states for railway systems