[A-AS01] From Weather Predictability to Controllability

[AAS01-P07] Bottom-up approach for mitigating extreme events under limited intervention options: a case study with Lorenz 96

*Takahito Mitsui¹, Shunji Kotsuki², Naoya Fujiwara³, Atsushi Okazaki², Tokuda Keita¹ (1.Juntendo University, 2.Chiba University, 3.Tohoku University)

Keywords:Extreme weather, Control, Intervention, Kalman filter, Lorenz

Prediction and mitigation of extreme weather events are important challenges in science and society. Recently, Miyoshi et al. proposed the control simulation framework to assess the controllability of chaotic systems under observational uncertainty. They developed a method for reducing extreme events in the Lorenz 96 model by exploiting its sensitivity to initial conditions and guiding trajectories toward desired outcomes with small control inputs (Sun et al., Nonlin. Processes Geophys., 30, 117-128, 2023). However, their method, which applies control inputs to all grid variables, reduces the success rate to about 60% when applied to only one site, at least in a specific setting. In this study, we present a bottom-up approach to mitigating extreme events by effectively utilizing limited intervention options. This contrasts with control-theoretic approaches, which take a top-down strategy, determining inputs by optimizing cost functions. We mitigate extreme events through local interventions based on multi-scenario ensemble forecasts. Our method achieves a success rate of 94%, even when interventions are applied to only one site per step. This represents a significant improvement over the ~60% success rate of the previous study, albeit at a moderately higher intervention cost. Furthermore, the success rate increases to 99.4% when interventions are applied to two sites.