More about the structure of the high latitude Jovian aurorae

Abstract

This study is based on the determination of a ‘reference’ main oval for Jupiter's aurora from a series of high-resolution images taken with the Faint Object Camera on board the Hubble Space Telescope in the H₂ Lyman bands centered near $\text{1550}\text{A}\text{̊}$. We have taken advantage of the visibility of the northern auroral oval over a large range of longitudes on June 24, 1994, especially for longitudes smaller than 170° where it is generally very faint and thus undetectable. In the south, where there is no such problem, we combined images taken from various points of view between June 1994 and September 1996. We find that the northern main oval is consistent in size and in general aspect with the footprint locus, in the Connerney (1998)'s VIP4 model, of magnetic field lines crossing the equator near 20R_J. However, the precise shape of this oval differs from the model (and from previous ‘reference’ main ovals) in that it exhibits a ‘bean-like’ aspect with excursions toward lower latitudes in the S_III longitude range 190–240° (an already reported feature), and toward higher latitudes in the poorly documented 120–150° range. In the south, our reference oval covers an area about that of the 30R_J VIP4 model. As in the north, it is shifted from a VIP4 model oval, toward lower latitudes from 110 to 200° and toward higher latitudes from 310 to 100°. The accurate definition of these (magnetically conjugate) oval loci puts additional strong constraints on magnetic field models at high latitude. Based on our reference main oval, we have then extrapolated still higher latitude ovals. Very interestingly, we find that we can fit (i) the highest latitude arc of oval detected well inside the main oval at longitudes greater than 170°, and (ii) the high latitude edge of what we had previously named the ‘transpolar emission’ at longitudes less than 170° (both also detected on images taken at other dates), by a single empirical oval. We suggest that this oval indicates the location of the northern polar cap boundary, within the uncertainty related to the temporal variability of the actual emission features. We can also fit another secondary arc of oval and a second branch of our ‘transpolar emission’ by a slightly larger empirical oval, presumably connected to the outer magnetosphere. This implies that the region of permanent high latitude diffuse emission seen inside the main oval in the dusk sector must be at the footprint of closed field lines connected all the way out from the middle magnetosphere to the magnetopause. Finally a bright spot is sometimes observed just equatorward of the northern polar cap boundary with an average brightness of 0.5-1 MR (comparing with a bright main oval). We establish that this spot does not rotate with the planet, but rather remains close to magnetic noon. We thus tentitatively identify it, by reference to the Earth aurora, with the footprint of the northern Jovian polar cusp, or with a transient dayside aurora. We also highlight the differences observed in the southern high latitude structure.

Introduction

The Jovian auroral emissions, which result from the precipitation of energetic charged particles into the planet's upper atmosphere along magnetic field lines surrounding the poles, represent the observable signature of auroral magnetospheric processes. As such, they have been studied for about 50 years in the radio frequency range, and for almost two decades at shorter wavelengths where spatial information is available. The far ultraviolet (FUV) auroral emissions, discovered in 1979 with Voyager and IUE (Broadfoot et al., 1979; Clarke et al., 1980), and due to collisional excitation of atmospheric species, have been the only direct signature of magnetospheric precipitating beams until Galileo discovered auroral emission at visible wavelengths with the solid state imaging instrument (SSI) (Ingersoll et al., 1998). FUV images were not obtained until the early 90s, with imagers on board the Hubble space telescope (HST), first with the faint object camera (FOC) (Dols et al., 1992), then the wide field planetary camera 2 (WFPC2; Clarke et al., 1996), and now with the space telescope imaging spectrograph (STIS). These instruments have provided high-resolution images with which to study the morphology of Jupiter's aurorae. This in turn allows to identify active regions in the magnetosphere by mapping the auroral emission loci up the magnetic field lines to the equatorial plane. Previous studies have revealed that the Jovian aurorae consist of a complex, but rather stable, pattern of features illustrated on Fig. 1, at the footprints of magnetic shells crossing the equatorial plane at various distances, ranging roughly form the orbit of Io at 5.9 Jovian radii (R_J) from the planet center in the inner magnetosphere, out to the magnetopause (e.g. Clarke et al., 1996; Grodent et al., 1997; Prangé et al., 1998). Among these features, the ‘main oval’ seems to be the most permanent feature, and very often the brightest one, i.e. the one corresponding to the most efficient precipitation process in the magnetosphere. It has also been the most thoroughly studied from the analysis of various FUV datasets, and of near infrared (IR) images of the thermal H₃⁺ emission at somewhat lower spatial resolution, where it is the dominant feature (Kim et al., 1994; Connerney et al., 1996). The main oval is a very narrow structure, $\text{50–500}\text{km}$ across at half-maximum (Tsurutani et al., 1997; Prangé et al., 1998; Ingersoll et al., 1998) which surrounds each pole of Jupiter approximately along the footprints of magnetic field lines crossing the equator in the middle magnetosphere near 20–30R_J on the dayside (Gérard et al., 1994a; Ballester et al., 1996; Clarke et al., 1996; Satoh et al., 1996; Prangé et al., 1998), and maybe somewhat closer in on the nightside (Vasavada et al., 1999). It has long been recognized that the locus of the main ovals slightly departs from the surface footprint of a constant model magnetic shell, especially in the north where the oval extends at lower latitude in the System III (1965) longitude range 180°⩽λ_III⩽300° (e.g. Gérard et al., 1994a; Connerney et al., 1996), even using the latest improved magnetic field models (Connerney et al., 1998). This departure has been identified as a longitude, rather than a local time, effect by various independent studies, including Galileo observations from the nightside (Satoh et al., 1996; Prangé et al., 1998; Vasavada et al., 1999). The ovals also exhibit longitudinal structures at various spatial scales (Ballester et al., 1996; Prangé et al., 1998). At the largest scale, it has long been noted that the north oval was generally bright and well defined for longitudes λ_III⩾170° and that it became faint or even disappeared at smaller longitudes (e.g. Gérard 1993, Gérard 1994a). More systematic studies have identified a longitudinal asymmetry both in the north and in the south, with the brightest parts of the northern and southern ovals being out of phase (Satoh et al., 1996; Prangé et al., 1998; Satoh and Connerney, 1999). This was interpreted as the combined effects of a particular precipitating process in the magnetosphere, pitch angle scattering which produces energetic charged particle losses toward the field line footprint where the surface field strength is smallest, and of the azimuthal asymmetry of the Jovian magnetic field structure (Prangé and Elkhamsi, 1991). This effect is best visible in the north and the locus of the main north oval is most often impossible to define below λ_III∼170°. Although very stable in shape, the main oval is subject to significant temporal intensity variations. Some seem to affect the feature as a whole (Prangé et al., 1998), accounting for the variability of the global auroral activity over a few days identified with IUE, and to be related to dynamical processes in the middle magnetosphere (Prangé and Livengood, 1998; Prangé et al., 2001). Others, observed a few times so far, consist of sudden brightenings of the main oval, very localized in azimuth near the dawn limb, and fading within a few hours (Gérard et al., 1994b; Ballester et al., 1996). Finally, it happens at times that the dim-to-undetectable segment of the north oval, eastward of 170° brightens temporarily. An event of this type has been discussed in Dougherty et al. (1998), and we have shown that the brightened part of the oval was magnetically connected to a transient layer of field-aligned currents detected by the magnetometer on board the Ulysses spacecraft. Such transient brightenings provide the only opportunities to map this part of the northern main oval with some degree of accuracy. It has not been possible to use the event mentioned above to map this region of the main oval, because it was obtained in 1992 during a pioneer observing program with the HST/FOC which suffered from a rather poor signal-to noise ratio and spatial resolution (Dols et al., 1992).

In 1994, just after the HST spherical aberration was corrected, we performed a new series of observations with FOC when its nominal resolution was restored. The images, taken in support of a series of spectra of the auroral emissions at central meridian longitudes (CML) spanning the range 105–227°, had not been analyzed so far because the auroral emission was not very bright. Indeed, the auroral activity at that date, as derived from simultaneous IUE observations (see Fig. 3 in Prangé et al., 2001) was below average. However, we noted that the λ_III⩽170° segment of northern main oval was brighter than the rest of the oval and clearly visible, in particular in a few images in which it was enhanced by line of sight effects (CML∼150°). The analysis of this set of images allows us to derive a reference northern auroral oval. As we had only a single image of the southern oval at that time, we have used a series of data taken at other dates and covering a large range of CML to similarly derive a southern reference main oval.

In Section 2, we describe the data, the observing conditions, and the procedures used to improve the visibility of auroral features. In Section 3, we explain how our reference main ovals are defined, we establish their geometry, and compare them with previous determinations of the auroral ovals. Our reference main oval is then used in Section 4 to derive a best fit oval to higher latitude features, which we suggest could be the polar cap, and we infer a possible signature of the northern polar cusp. Finally, the results are summarized and discussed in Section 5.