[A-AS01] From Weather Predictability to Controllability

[AAS01-07] Ensemble Kalman Control: Mathematical Platform to Explore Tropical Cyclone’s Controllability

*Yohei Sawada¹, Duc Le¹, Yuyue Yan², Kazumune Hashimoto², Masashi Minamide¹ (1.The University of Tokyo, 2.Osaka University)

Keywords:Controllability, Ensemble Kalman Control, data assimilation, tropical cyclone

It is a grand challenge to find a feasible weather modification method to mitigate the impact of extreme weather events such as tropical cyclones. Previous works have proposed potentially effective actuators and assessed their capabilities to achieve weather modification objectives through numerical simulations. However, few studies have explored efficient mathematical and computational methods to inversely determine optimal actuators from specific modification goals. Here we demonstrate the utility of the ensemble Kalman filter (EnKF)-based control method, referred to as ensemble Kalman control (EnKC). In EnKC, the reference vector, which serves as the control target, is assimilated into the state space as a pseudo-observation by ensemble Kalman smoother to obtain the appropriate perturbation to be added to a system. We demonstrated the efficiency of EnKC for controlling extremely high-dimensional spatio-temporally chaotic systems. The series of numerical experiments of idealized tropical cyclones indicate that EnKC efficiently identifies local, small, and intermittent control perturbations that can mitigate the intensity of tropical cyclones. The existing techniques of EnKF, such as background error covariance localization, can improve the sparsity and efficiency of the control. This work paves the way toward the real-world applications of EnKC to explore the controllability of extreme atmospheric events.