Fig. 2: Ensemble simulation results for prototypical R-tipping systems. | Nature Machine Intelligence

Fig. 2: Ensemble simulation results for prototypical R-tipping systems.

From: Deep learning for predicting rate-induced tipping

Fig. 2

a–c, Ensemble simulations are conducted on the saddle-node system (a), Bautin system (b) and compost-bomb system (c). For each simulated system, the same time-varying forcing parameter is used, but different noise perturbations (sampled from independently and identically distributed white noise) are applied to the simulations. Top panels: the simulated time series that do not tip (blue) and that exhibit R-tipping (red). Middle panels: the time-varying forcing parameter. Bottom panels: the probability density of the observed occurrence time of R-tipping. The blue-shaded area denotes the 99% confidence intervals of the ensemble realizations that do not manifest R-tipping.

Back to article page