Machine Learning Analysis of Jupiter's Far-Ultraviolet Auroral Morphology

Abstract

We present the first principal component analysis of Jupiter's far-ultraviolet auroras, in order to identify the most repeatable sources of variation in the auroral morphology. We show that the most recurrent source of variance is emission just poleward of the statistical oval on the dawnside. Further significant repeatable sources of variance are localized expansions of the main emission on the dawnside or duskside and poleward emission near noon and along the duskside. We go on to show using a density-based spatial clustering of applications with noise clustering analysis that the most significant auroral components form six repeatable auroral morphological classes. One class, exhibiting bright main and poleward dusk emissions, occurs solely during solar wind compressions. This presents an important new tool for diagnosing magnetospheric compressions at Jupiter.