Data-driven low-dimensional dynamic model of Kolmogorov flow

Minimal dimensional models are desirable for reduced computational costs in simulations as well applications such model-based control. Long-time dynamics of flows often evolve on a low-dimensional manifold M the full state space. We use neural networks to estimate and it two-dimensional Kolmogorov flow chaotic bursting regime. Outcomes include: minimal dimension estimate, good short-time tracking long-time statistics, accurate predictions events.