Predicting Critical Transitions in Multiscale Dynamical Systems Using Reservoir Computing
This work addresses the challenge of forecasting critical transitions in multiscale systems, which is important for fields like climate science and engineering, though it appears incremental as it builds on existing reservoir computing techniques.
The authors tackled the problem of predicting rare critical transitions in slow-fast nonlinear dynamical systems by developing a data-driven reservoir computing method, which successfully predicted transitions several time steps in advance in numerical experiments across low to high-dimensional examples.
We study the problem of predicting rare critical transition events for a class of slow-fast nonlinear dynamical systems. The state of the system of interest is described by a slow process, whereas a faster process drives its evolution and induces critical transitions. By taking advantage of recent advances in reservoir computing, we present a data-driven method to predict the future evolution of the state. We show that our method is capable of predicting a critical transition event at least several numerical time steps in advance. We demonstrate the success as well as the limitations of our method using numerical experiments on three examples of systems, ranging from low dimensional to high dimensional. We discuss the mathematical and broader implications of our results.