Jupiter's ultraviolet auroral bridge: the influence of the solar wind on polar auroral morphology

Introduction. Jupiter's ultraviolet aurora frequently shows a number of arcs between the dusk-side polar region and the main emission, which are denoted as “bridges” (Figure 1, left).  Thus far, the only dedicated study of the auroral bridge was part of a Master’s thesis which analysed three images of the aurora taken by the Space Telescope Imaging Spectrograph (STIS) instrument aboard the Hubble Space Telescope (HST) that contained bridges. The bridges were found to map to the dusk-side magnetosphere between 10 and 22 magnetic local time (MLT) and be largely confined to radial distances greater than 60 RJ. Other works have identified the appearance of dusk-side polar arcs with compression by the solar wind, though these conclusions were based a limited number of observations, and the origins of auroral bridges remain poorly understood.

Methods. We present a largely automated detection and statistical analysis of bridges over 248 HST-STIS observations, alongside a multi-instrument study of crossings of magnetic field lines connected to bridges by the Juno spacecraft during its first 30 perijoves. The detection of bridges in the large number of HST images is performed automatically using a bespoke arc-tracing algorithm (Figure 1, right), which is combined with manual arc designations and a random-forest filter to provide a model test accuracy of 82%. In the Juno-UltraViolet-Spectrograph (Juno-UVS) images, bridge locations (and hence bridge-crossing timestamps) were identified manually to avoid the introduction of artefacts of the automated method.

Results. Bridges are observed to arise on timescales of ~2 hours, can persist over a full Jupiter rotation, and are conjugate between hemispheres. The appearance of bridges is strongly associated with compression of the magnetosphere by the solar wind (Figure 2, left); there is a clear distinction between green points (cases where the magnetosphere was compressed) and red points (cases where the magnetosphere was uncompressed) in the total detected bridge length. Low-altitude bridge crossings are associated with upward-dominated, broadband electron distributions (Figure 2, left), consistent with Zone-II (Mauk et al. 2020; doi:10.1029/2019JA027699) aurorae, notable since similar regions in Earth and Saturn’s aurorae show no appreciable emission. Bridge crossings are also associated with plasma-wave bursts observed by Juno-Waves, in agreement with existing theoretical models for the generation of polar-region aurorae. Electron populations associated with crossings of field lines threading the main emission by Juno also become more downward-dominated when separate bridges are present in the nearby aurora. Figure 2 (right) shows that main emission crossings where bridges are present in the aurora (green) did not show the same bridge-like/Zone-II upward-travelling population of electrons as cases where no separate bridge was observed in the aurora (orange), which indicates that the “bridge” aurora was merely spatially indistinguishable from the main emission in these latter cases.

Conclusions. Combining these two sets of results, bridges are identified as Zone-II aurorae that have become spatially separated from the Zone-I aurorae under the influence of the solar wind. In the absence of separate bridges in the polar region, the main emission takes on a decidedly mixed Zone-I-II character (bidirectional electrons, plasma-wave bursts) which become uniquely Zone-I-like when bridges are present. This is consistent with the previously observed adjacency of the auroral Zone-I and Zone-II; the gap between the two is thus observed to increase during solar wind compression. The solar wind is thus implied to be able to exert a considerable influence on Jupiter’s internally driven magnetosphere that is reflected in the morphology of its aurora.

Figure 1: (left) An image of the northern jovian UV aurora captured by HST during the GO-15638 campaign (exposure ID: odxc01okq). A 15°-by-15° grid in System-III longitude and planetocentric latitude is included; the System-III longitude of certain meridians are given in white, and certain planetocentric latitudes in magenta. The average subsolar longitude during this exposure (170°) is denoted by a solid yellow line, and the positions of the dawn and dusk hemispheres are included to guide the reader. The approximate location of the polar collar is enclosed by red dashed lines, and that of the noon active region by a green dashed ellipse. Bridges are highlighted with yellow arrows. (right) Results of the bridge-detection algorithm after filtering. Red lines denote arcs accepted by the filter, and grey the arcs that have been discarded. The seed point of each arc is given in magenta. Manually designated arcs are given in yellow. White contours give the region of validity of the JRM33 flux-equivalence mapping along closed field lines.

Figure 2: (left) Detected average expansion of the main emission vs. the total detected magnetosphere-mapped bridge length for each northern HST-STIS series considered in this work. Negative expansions imply a contracted main emission. Bridge-length error bars are determined as described in Appendix A. Green crosses denote those cases where the magnetosphere was compressed, and red pluses those cases where the magnetosphere was uncompressed; grey points denote cases where the compression state of the magnetosphere is unknown. The least-squares best fit is given by the solid blue line. (right) Median-average properties of dusk-side (12<MLT<18), low-altitude (<3 RJ) bridge crossings (red, solid) and main-emission crossings, both with (green, dotted) and without (orange, dashed) local bridges, observed by Juno. From top to bottom: JEDI electron flux vs. pitch angle profiles; Waves-E LFR-Lo spectral intensity. The shaded regions denote the 25-to-75th percentile range.