Nearly ten years’ worth of Juno data analysed by a neural network reveal dust hidden in Jupiter’s immediate vicinity

A two-layer convolutional neural network sorted through the dust impacts and solitary waves detected by Juno’s Waves instrument, revealing a ring of dust that is much more extensive than can be seen through a telescope.

Jupiter's ring system is made of microscopic dust grains, whose movement is governed by the magnetic field of the planet, which rotates very quickly. Optical observations described the main ring and halo, but the very diffuse dust close to the planet largely escapes telescopes. A team of nine researchers took over nearly a decade of electrical field measurements made on site by the Waves instrument of the probe Juno, and entrusted them to a network of two-stage convolutive neurons, charged with distinguishing dust impacts, solitary waves and other plasma waves.

A tore-shaped halo

The detections show a structure that evolves with distance and extends well below the inner limit of the visible halo: near Jupiter, the dust forms a compact, symmetrical layer on both sides of the equator, which then thickens vertically into a wide halo before tightening towards the main ring. This direct measurement confirms the toroidal shape of the Jovian halo, with higher densities in the halo and transition with the main ring, and hemispheric asymmetries attributed to the geometry of the magnetic field.

The solitary waves are concentrated at high latitudes; projected along the field lines, they coincide closely with Jupiter's auroral structures, especially the main auroral oval.