Conditioning on PDE Parameters to Generalise Deep Learning Emulation of Stochastic and Chaotic Dynamics
This work addresses the challenge of efficiently exploring parameter spaces in complex dynamical systems for researchers in computational physics and climate science, though it is incremental as it builds on existing emulation methods with conditioning and attention mechanisms.
The paper tackles the problem of emulating stochastic and chaotic spatio-temporal systems by developing a deep learning model conditioned on PDE parameters, enabling generalization across parameter values and domain sizes, and demonstrates computational speed-ups over numerical integration with applications to chaotic and turbulent systems.
We present a deep learning emulator for stochastic and chaotic spatio-temporal systems, explicitly conditioned on the parameter values of the underlying partial differential equations (PDEs). Our approach involves pre-training the model on a single parameter domain, followed by fine-tuning on a smaller, yet diverse dataset, enabling generalisation across a broad range of parameter values. By incorporating local attention mechanisms, the network is capable of handling varying domain sizes and resolutions. This enables computationally efficient pre-training on smaller domains while requiring only a small additional dataset to learn how to generalise to larger domain sizes. We demonstrate the model's capabilities on the chaotic Kuramoto-Sivashinsky equation and stochastically-forced beta-plane turbulence, showcasing its ability to capture phenomena at interpolated parameter values. The emulator provides significant computational speed-ups over conventional numerical integration, facilitating efficient exploration of parameter space, while a probabilistic variant of the emulator provides uncertainty quantification, allowing for the statistical study of rare events.