Deep Learning Surrogates for Emulating Stochastic Climate Tipping Dynamics
For climate scientists, this provides a fast, differentiable surrogate to replace computationally expensive simulations of tipping dynamics, enabling large-scale uncertainty quantification and parameter studies.
This work develops a dynamics-informed Temporal Fusion Transformer as a data-driven surrogate for Earth system simulations, achieving 465x computational speedup while accurately forecasting stochastic climate tipping events (Atlantic and Pacific collapses) across thousands of time steps.
This work explores a dynamics-informed Temporal Fusion Transformer (TFT) as a data-driven surrogate for computationally intensive Earth system simulations. Focusing on multivariate time series describing global ocean transport, we demonstrate the surrogate's ability to forecast tip events across thousands of time steps. The data involve up to 21 non-stationary time series in addition to static covariates describing free parameters and initial conditions. Modifications to the architecture and objective function yield a surrogate that anticipates the timing of Atlantic and Pacific collapses to high fidelity and captures the stochastic uncertainty in transition timing across ensemble predictions. The learned surrogate achieves a 465x computational speedup over the numerical simulator while maintaining differentiability with respect to parameters and initial conditions.