The well-known banded structure of the gas giant Jupiter varies considerably on the multi-year time scale of 3 to 10 years. There are noticeable colour changes, brightening events, and stormy outbreaks visible at the cloud tops, capturing the imaginations of amateur and professional observers alike.  These bursts of activity are sometimes called ‘global upheavals’ and demonstrate quasi-predictable patterns. They have been witnessed for over 100 years by ground-based telescopes, but their underlying causes remain a long-standing mystery. 

Jupiter’s belts and zones are also seen in the infrared 5-µm radiation, which originates deeper down in its atmosphere, where the pressure is about 5 times that of the atmosphere at the Earth’s surface. Cloud-free bands appear bright at this wavelength, whereas thick clouds block the deep internal glow of Jupiter, rendering cloudy zones as dark.  This 5-µm radiation also varies on the intradecadal time scale (see figures 1a and b). Periodic signals have been detected in the 5-µm radiation, with periods of 4 to 8 years, depending on latitude. Since the signals are coherent at all longitudes, and seem to show a connection between events widely-separated latitudes north and south of the equator, it is likely that this periodicity is connected with something going on deep inside Jupiter.

We are proposing that torsional oscillations due to Jupiter’s magnetic field are responsible for the periodicity. The longitude-independent oscillations are a special type of magnetic waves, the Alfvén waves, that are provided by a tension of the magnetic field line: see figure 2 for an explanation of what these torsional oscillations are. The disturbances can propagate as waves, both towards or away from the rotation axis (see figures 2b-c), which means they move in latitude at the surface. Torsional oscillations have been detected in the Earth’s liquid iron core, with a 6-year oscillation period that is detected in the geomagnetic field record, and also through small variations in the length of day. They can also be seen in geodynamo models, which encouraged us to look for them in our models of Jupiter’s dynamo.  Our simulations did show torsional oscillations driven by the turbulent convection that transports heat outwards through the planet.

The Juno mission has measured Jupiter’s magnetic field with astonishing accuracy: see details of the findings, for example, in recent reports by Connerney et al. (J. Geophys. Res. Planets 2022). This has enabled us to work out the wave speed of torsional oscillations expected in Jupiter. Assuming the oscillations are excited by fluctuations in the zonal winds visible at the surface, we further worked out the oscillation periods at varying latitudes. Remarkably, these periods turned out to be very similar to those observed in the 5-µm radiation signal.

The Juno mission has now entered an extended phase, which allowed us to see how Jupiter’s field develops in time. The secular variation of the Earth’s field over time revealed the core torsional oscillations. Fortunately, there is a strong magnetic flux patch near Jupiter’s equator (Moore et al. Nature 2018) which makes it easier to detect changes in its field. Initially it looked as though the field was drifting steadily eastwards rather than oscillating, but the latest data shows the eastward drift is slowing down (Bloxham et al., J. Geophys. Res. Planets 2022), more compatible with a torsional oscillation, although further observations are needed to confirm this. Using dynamic mode decomposition (DMD) applied to the 5-µm signal, waves travelling in latitude at the wave speed expected from torsional oscillations can be seen (see figure 3), providing further support to our hypothesis.

A natural question is how does the torsional oscillation change the 5-µm radiation? The amplitude of the oscillation is small compared to the turbulent velocities in the weather layer where the 5-µm signal comes from, so it is more likely that the convective heat flux is disrupted hundreds of kilometers below, where conditions are expected to be much calmer. The shearing motion of the torsional oscillation could disrupt the heat transport there, and this modulation could propagate up to the weather layer and change environmental conditions to produce the observed signal, but exactly how this works needs more research.

The study of Jupiter’s interior has been largely disconnected from the study of its meteorology. This work suggests that they are really coupled systems. The torsional oscillations are driven mainly by fluctuations in the powerful jet streams in the atmosphere, but their motions extend throughout the planet. If there are stable stratified layers in the interior, as has recently been suggested, these will affect the structure of the torsional oscillations in ways which could be detectable. So these oscillations could be used to probe the interior structure of the planet, just as other types of waves have been used to gain understanding of planetary and stellar interiors.