[M-GI29] Data-driven geosciences

[MGI29-02] Jupiter's torsional oscillations, tropospheric variability, and dynamic mode decomposition

*Kumiko Hori¹, Chris A Jones², Steven M Tobias² (1.Graduate School of System Informatics, Kobe University, 2.Department of Applied Mathematics, University of Leeds)

Jupiter's weather layer exhibits long-term and quasi-periodic cycles of meteorological activity. There are cycles with several-year intervals, dependent on the latitude, which were detected in infrared radiation, which provides a window into the lower troposphere; however the origin of these cycles has been a mystery. We propose that magnetic torsional oscillations/waves arising from the dynamo region could trigger the variability of the tropospheric activity. These axisymmetric waves are magnetohydrodynamic waves influenced by the rapid rotation, which have been detected in Earth's fluid core. Using the Juno magnetic field model, we compute the expected speed of these waves. For the wave excited by variations in the zonal jet flows, their half period is estimated to be 3-5 years at mid latitudes, consistent with the intervals with the cycles of variability identified in the visible/infrared observations. The nature of these waves, including the wave speed and the wavelength, is revealed by dynamic mode decomposition, applied to the spatio-temporal data for the emission. Our results imply that exploration of these magnetohydrodynamic waves may provide a new window to the internal dynamics and dynamo of Jupiter.