Controlling eddies in the non-autonomous Lorenz-84 Model

Extreme weather events emerge from the chaotic dynamics of the atmosphere. Adaptive chaos control has been applied to Lorenz models in this context. Weather Jiu-Jitsu is a control paradigm that seeks to steer trajectories away from dangerous weather regimes using small, well-timed perturbations. The seasonally forced, non-autonomous Lorenz model has a much more complex attractor than the "toy models" in the existing literature. Noise or stochastic terms can also significantly increase the complexity of control via small perturbations. We present the first example of finite time adaptive chaos control for a seasonally forced and noise perturbed Lorenz–84 model. We demonstrate two strategies for triggering control: (1) local Lyapunov exponents (LLE), and (2) transition probabilities for the latent states of a non-homogeneous Hidden Markov Model (NHMM). The second approach is new. It is motivated by thinking of future applications to a latent embedding space of planetary atmospheric circulation that would get us closer to real world analyses. The NHMM "triggers" coincide with strongly positive LLE regimes, confirming their dynamical interpretability. Thus, latent-state triggers complement instability diagnostics and provide a conceptual bridge to weather foundation models where hidden states are already identified and could be used for triggering control.