The Optical Aurora of Jupiter's Galilean Satellites; or, What They Do in the Shadows

Author: Milby, Zachariah

Year: 2026

Degree: Dissertation (Ph.D.)

Advisor: de Kleer, Katherine R.

Committee Members: Knutson, Heather A.; Brown, Michael E.; de Kleer, Katherine R.; Blake, Geoffrey A.

Option: Planetary Sciences

Abstract

When electrons from Jupiter's rapidly rotating magnetosphere collide with the thin atmospheres of the Galilean satellite—Io, Europa, Ganymede and Callisto—they create faint auroras like Earth's northern and southern lights. However, reflected sunlight from their icy surfaces overwhelms these faint auroral emissions at visible wavelengths. Fortunately, the satellites transit Jupiter's large shadow once per synodic orbit. When Jupiter is near elongation, we can view these transits from Earth and observe the satellites' faint auroral emissions in the absence of reflected sunlight. Over the past five years, we have used the High Resolution Echelle Spectrometer (HIRES) on the Keck I telescope at the summit of Maunakea to observe the optical wavelength auroras of these four satellites during the eclipse phase of their orbits. My analysis of the auroras and the modulation of their brightnesses over five- to ten-minute timescales has provided new insight into the composition and density of their atmospheres and how they interact with the local plasma environment.

At Io, I found 13 auroral emissions not previously detected. I compared these detections to optical wavelength images taken through a variety of narrowband filters during the Cassini flyby to understand the morphology of the different auroral emissions. Using an auroral emission model, I showed that the oxygen emissions were consistent with electron impact on an atmosphere composed of O, SO and SO₂, disagreeing with prior studies which posited the oxygen emissions came from electron impact on its extended atomic oxygen corona. The brightest oxygen emissions did not vary with ambient upstream electron density, suggesting the electron flux into Io's atmosphere is more complex.

At Europa, the oxygen emissions varied systematically with the upstream electron density, indicating a simpler electron precipitation process than that at Io. This allowed me to use the variability in the auroral emissions as a proxy to understand variability in Europa's plasma environment. I evaluated how the brightnesses changed on short timescales during single eclipses as well as systematically on longer timescales over the four years we made the observations. I validated the results by comparing to in situ measurements taken during the Galileo mission. I also considered the timescale under which a sublimated water atmosphere would need to freeze back onto the surface as Europa passed into Jupiter's shadow.

At Ganymede, I showed the brightness had a hemispheric asymmetry that correlated with the geometry between the satellite and the centrifugal equator of Jupiter's magnetosphere. The brightness ratio between the northern and southern hemispheres suggested that its auroras are triggered by the bounce motion of electrons along Jovian magnetic field lines rather than by ram flux from the upstream plasma population. I compared the results to model simulations of Ganymede's O₂ and H₂O atmospheres at different times during the eclipse to understand the detectability of a sublimated water atmosphere which, like with Europa, would need to collapse rapidly at the onset of the eclipse.

Finally, at Callisto, detections of auroral emission have been challenging because of the very low magnetospheric electron densities at its orbital distance from Jupiter. However, through careful data processing I was able to retrieve brightnesses of the 630.0 nm [O I] emission from individual spectra, allowing for the first time-series analysis of Callisto's aurora. Like Europa, I found the brightness modulated with distance from the centrifugal equator as expected for excitation by the upstream ram electron flux. This indicated the absence of a substantial ionosphere and supported the paradigm that the ionosphere exists only over the trailing hemisphere when it is illuminated by sunlight. I did not detect enough emission lines to perform the atmospheric compositional analysis I did with the other satellites, but I forward-modeled the expected brightness of the O, O₂, H₂O and CO₂ atmospheres inferred or measured by other studies and found agreement with the aurora brightnesses and upper limits I observed with HIRES.