A data-driven, physics-informed framework for forecasting the spatiotemporal evolution of chaotic dynamics with nonlinearities modeled as exogenous forcings

M. A. Khodkar, Pedram Hassanzadeh

Abstract: Highlights•A novel data-driven linear method for spatiotemporal forecasting of chaotic and turbulent systems is introduced.•The method is linear but accounts for nonlinearities as exogenous forcings.•The method captures a key aspect of the nonlinear dynamics through the exogenous forcing component.•The method is shown to successfully forecast the evolution of several chaotic systems such as Lorenz 96.•The method also shows success when applied to 2D cavity flows at high Reynolds numbers.