Transient Water Vapor at Europa’s South Pole In November and December 2012 the Hubble Space Telescope (HST) imaged Europa’s ultraviolet emissions in the search for vapor plume activity. We report statistically significant coincident surpluses of hydrogen Lyman-α and oxygen OI130.4 nm emissions above the southern hemisphere in December 2012. These emissions are persistently found in the same area over ~7 hours, suggesting atmospheric inhomogeneity; they are consistent with two 200-km-high plumes of water vapor with line-of-sight column densities of about 1020 m−2. Nondetection in November and in previous HST images from 1999 suggests varying plume activity that might depend on changing surface stresses based on Europa’s orbital phases. The plume was present when Europa was near apocenter and not detected close to its pericenter, in agreement with tidal modeling predictions.

We report STIS spectral images of Europa’s trailing/anti-Jovian hemisphere and leading hemisphere obtained in November and Decem-ber 2012, respectively. Previous observations, in 1999, targeted Europa’s trailing hemisphere (Table 1). The observations in 2012 were timed to coincide with the maximum variation of the Jupiter’s magnetic field orientation at Europa. With this configuration, spatially inhomogeneous yet time-variable emissions originating from the periodically changing magnetospheric conditions can be separated from time-stationary emis-sion inhomogeneities due to atmospheric anomalies.

http://www.spacetelescope.org/static/archives/releases/science_papers/heic1322a.pdfAurorae of Io and Europa: Observations and Modeling
In the present dissertation we study the auroral emissions emanating from the tenuous atmospheres of Jupiter's satellites Io and Europa. The satellites are embedded in a dense magnetospheric plasma environment. Due to Jupiter's fast rotation the corotating magnetospheric plasma particles constantly flow past Io and Europa causing a complex interaction and triggering auroral emission in the atmospheres. Therefore, aurora observations are a useful tool to explore both the magnetospheric environment and the neutral gas clouds of the satellites.
[..]
Io's aurora is characterized by bright emissions on the sub-Jovian and anti-Jovian flanks close to the equator and a fainter limb glow around the polar regions. Analyzing the STIS images we demonstrate that the variations of Io's UV aurora observed over a period of five years can be attributed to changes in the magnetospheric environment as well as to the varying viewing perspective.
[..]
Our results are a strong indication, that Io's atmosphere is driven by sublimation of SO2 frost rather than direct volcanic outgassing. The ultimate source for Io's atmospheric gas is widely debated for many years. We also investigate the observed variation or rocking of the bright auroral spots around Io's equator. The location of the spots has been shown to be correlated to the Jovian magnetic field orientation at Io. The exact correlation is, however, not 1:1, but is presumably affected by local perturbations of the magnetic field.
[…]
We find that most emission is observed on the disk of Europa rather than around the limb like in comparable observations of Io's aurora. We show that an increasing O2 density towards the sub-solar point possibly explains the observed morphology as well as previous observations. While the OI] 1356 Å emission pattern appears to vary periodically in correlation with the changing magnetospheric environment, the OI 1304 Å morphology is clearly dominated by a very bright locally confined emission in the northern, anti-Jovian quadrant of Europa's disk. The location of this anomaly coincides exactly with the longitude, where a peak in water vapor production is predicted due to increased shear heating at the surface cracks. Estimating the emission brightnesses expected for a local water plume, we find that the observed UV emission intensities are principally consistent with a locally confined abundance of water vapor. However, due to observational uncertainties and since we have neglected the effects of the plasma interaction for the approximation of the H2O abundance, our results can not be seen as prove for the existence of water plumes on Europa. To accurately determine the effects of an asymmetric O2 atmosphere and the influence of a local water plume, the plasma interaction has to be simulated.
http://kups.ub.uni-koeln.de/4894/http://kups.ub.uni-koeln.de/4894/1/RothLorenz_Dissertation_2012.pdfTHE FAR-ULTRAVIOLET OXYGEN AIRGLOW OF EUROPA AND GANYMEDE
Far-UV spectra of Europa and Ganymede, acquired by the Hubble Space Telescope Goddard High
Resolution Spectrograph, indicate that, in addition to faintly reÑected sunlight, both satellites emit O I
1304 A. and O I 1356 A. airglow radiation. The observed brightnesses of the reÑected solar C II 1335 A.
feature indicate that the disk-averaged albedos of Europa and Ganymede are about 1.5% and 2.6%,
respectively. Airglow emissions from both satellites are characterized by the Ñux ratio F(1356 A. )/F(1304
A. ) of roughly 1È2, diagnostic of dissociative electron impact excitation of O Inferred vertical 2. Ocolumn densities are in the range (2.4È14)]1014 cm~2 for Europa and (1È10)]1014 cm~2 f2or Ganymede.
The observed double-peaked proÐle of GanymedeÏs O I 1356 A. feature indicates a nonuniform
spatial emission distribution that suggests two distinct and spatially-conÐned emission regions, consistent
with the satelliteÏs north and south poles.
Previous studies have suggested that icy satellites could possess tenuous oxygen atmospheres. The earliest theories
assumed that H sublimation, and subsequent photo-chemical processing could produce a predominantly oxygen
atmosphere on Ganymede (Yung & McElroy 1977) and Europa (Kumar & Hunten 1982). However, JupiterÏs satellites
are known to be subjected to signiÐcant Ñuxes of magnetospheric particle radiation, and charged-particle
sputtering of icy surfaces should provide a direct source of O gas equal to or exceeding that from sublimation (see Jo2hnson 1996 and references therein; Ip 1996). Charged particle irradiation is also thought to create substantial
reservoirs of O and trapped in GanymedeÏs icy surface 2 Omaterial (Spencer, C3alvin, & Persons 1995; Calvin, Johnson, & Spencer 1996; Noll et al. 1996). The 1996 Europa far-UV spectra are essentially identical to the 1994
spectrum Ðrst analyzed byHall et al. (1995), and conÐrm the existence and stability of EuropaÏs oxygen atmosphere and airglow. The 1996 Ganymede observations also reveal the existence of a thin oxygen atmosphere. However, the
double-peaked line shape of GanymedeÏs O I 1356 A. feature implies the existence of a nonuniform distribution of Ñux, with signiÐcant enhancements that coincide with the satelliteÏs polar regions. This, combined with the recent Galileo detection of a magnetosphere and charged-particle population, raises the possibility that Ganymede possesses polar auroral emissions.
Clearly, to investigate this further, the satelliteÏs far-UV emissions need to be imaged, and the newly installed Space Telescope Imaging Spectrograph is well suited to the task. Our models indicate that the O I 1304 A. ]1356 A. luminosity of each of GanymedeÏs polar hoods is roughly (5È 15)]106 watts. Using a compilation of O electron-impact cross sections (Itikawa et al. 1989), we e2stimate that the total power deposited into GanymedeÏs atmosphere is
roughly a factor of 100 greater than that emitted in the two O I multiplets. Thus the data suggest that the order of 109 watts are required to power the emissions at each pole, an electron energy Ñux of the same order as observed by Frank et al. (1997a).
http://iopscience.iop.org/0004-637X/499/1/475/pdf/36489.pdfThe plasma plumes of Europa and Callisto
We investigate the proposition that Europa and Callisto emit plasma plumes, i.e., a contiguous body of ionospheric plasma, extended in the direction of the corotation flow, analogous to the plume of smoke emitted in the downwind direction from a smokestack. Such plumes were seen by Voyager 1 to be emitted by Titan. We find support for this proposition in published data from Galileo Plasma Science and Plasma Wave observations taken in the corotation wakes of both moons and from magnetometer measurements reported from near the orbit of, but
away from, Europa itself. This lends credence to the hypothesis that the plumes escaping from the ionospheres of Europa and Callisto are wrapped around Jupiter by corotation, survive against dispersion for a fairly long time and are convected radially by magnetospheric motions. We present simple models of plume acceleration and compare the plumes of the Europa and Callisto to the known plumes of Titan.
3. Atmospheres and ionospheres
The tenuous atmospheres and ionospheres discovered by HST and Galileo observations are the source of the
plumes described above at Europa and possibly at Callisto. A detailed aeronomic study of these atmospheres and
ionospheres lies outside the scope of this paper. We shall, however, summarize some results from the literature and
point out the incompleteness of the present state of understanding of the atmospheres.
In contrast to Ganymede the equatorial regions of which are shielded from the magnetospheric sputtering flux by an
intrinsic magnetic field (Kivelson et al., 1998), the entire surfaces of both Europa and Callisto are exposed to energetic particles. Orton et al. (1996) observed Europa from Galileo and found a subsolar temperature of 128 K and a terminator temperature of about 90 K. Europa appears according to
Orton et al. (1996) to exhibit a considerably lower thermal inertia than the other icy satellites and to be considerably
colder on the night side than expected. Shematovich and Johnson (2001) have simulated the creation of the tenuous
molecular oxygen atmosphere and show how it evolves from a near equilibrium configuration near the surface to a highly non-equilibrium atmosphere at higher altitudes. The global 3-D simulation of Saur et al. (1998) provides a satisfactory description of the neutral atmosphere in the equilibrium region
near the surface, but becomes a poor approximation at higher altitudes. The observations of Hall et al. (1995, 1998)
were not space-resolved on Europa and the satellite was taken to be a uniform emitting disk. They also derived their
abundances from Voyager PWS electron data.
In view of these resolution limitations, we consider the atmosphere of Europa to be global and uniform, although,
in reality, the composition of the atmosphere is latitudedependent, since the polar cap temperatures, which are too
low to allow effective sublimation, will cause any sputtered water vapor to recondense before it can dissociate. Molecular oxygen, on the other hand, can maintain a vapor pressure in the solid phase even at polar cap temperatures with the result that there will be gaseous O2 available for interaction
with the plasma, energetic particles and the solar UV flux (Johnson, 1996). In the low-latitude regions of Europa,
the dayside temperature has been observed to be as high as 128 K (Orton et al., 1996) and a vapor pressure
of water can be maintained above the ice. Bar-Nun et al. (1985) find that the sputtering yields of both water vapor
and atomic hydrogen are temperature independent and of comparable magnitude which they interpret to imply that H is created by the breaking of water molecule bonds at the surface at a rate comparable with the sputtering of water.
The hydrogen thus released will escape and the hydroxyl remaining behind will be photodissociated within a matter
of days (Johnson and Quickenden, 1997), with the hydrogen again escaping. We expect, therefore, that there will
be no detectable flux of protons in the escaping plasma plume, because of the fast escape in a time short compared
to the ionization time, as was also shown with respect to the polar wind of Ganymede (Vasyli¯unas and Eviatar, 2000;
Eviatar et al., 2001b). For the case of Europa, the simulation of Shematovich and Johnson (2001) predicts an atmosphere dominated everywhere by O2. A similar conclusion was reached with respect to both Europa and Ganymede from UV (Hall et al., 1995, 1998) and radio occultations (Kliore et al., 1997, 2002). In analogy with the analysis of Eviatar et al. (2001b) for an atmosphere over a sputtered ice surface, the dominant ion in the atmosphere is expected to be O+2 .
Atomic oxygen created by dissociative recombination of O+2 escapes and is, therefore, unavailable for local ionization by either electrons or UV photons. A significant shortfall in the predicted model density was
found for Europa by Shematovich and Johnson (2001) and it is clear that an additional source of sputtered matter is required, especially, as Shematovich and Johnson (2001) point out, the yield of molecular oxygen in the ion sputtering flux is no greater than 20%. Shematovich and Johnson (2001) suggest electron sputtering which while inefficient is capable of deep penetration and of bringing out molecular oxygen. Paranicas et al. (2001) estimate the energy flux of electrons into Europa to be of the order of 1011 keVcm−2 s−1, which can provide a sufficiently large source strength of O2, of order 2 × 1010 cm−2 s−1, if the G value (O2 emitted per 100 eV of electron energy deposited) is as large as 0.02. Shematovich and Johnson (2001) used a somewhat lower estimate of the electron dose and found G>0.03 to be the required constraint.
The situation at Callisto is even more complex and puzzling. The occultation data from Galileo cited above are
consistent with a density about two orders of magnitude greater than that of Europa and Ganymede. Despite the resulting high column density no corresponding aurora or air glow emissions above an upper limit of 15 R were observed (Strobel et al., 2002), which deprives the model of the ultraviolet observational constraints on the densities available for Europa and Ganymede (Hall et al., 1995, 1998).
It is not immediately obvious why Callisto should have a denser atmosphere in view of the above mentioned fact
that the particle energy flux available for sputtering delivered to Callisto and Ganymede’s equator is about 300 times
smaller than that delivered by Jupiter to Europa, while an intermediate flux is delivered to the polar cap of Ganymede (Cooper et al., 2001). As mentioned above, the detailed comparative aeronomy of the Galilean satellites lies outside the scope of this paper and is worthy of a separate study. We note that the paucity of available input energy can readily explain the absence of air glow emissions at Callisto. Aurorae
at Ganymede are excited by the Birkeland current acceleration of electrons into the polar cap along the flux tubes of
the intrinsic magnetic field (Eviatar et al., 2001a). No such mechanism capable of compensating for the reduced electron energy input exists at Callisto. It is thus to be expected that air glow and auroral UV emissions will be absent at Callisto without invoking a high conductivity ionosphere as proposed by Strobel et al. (2002). If indeed access to the surface of Callisto is denied by strong currents in the ionosphere, then it is not clear how a dense atmosphere and concomitant ionosphere can be created out of matter sputtered from the surface.
4. Acceleration of the plume
A major difference between the plumes of Europa and Callisto, and those of Titan, in addition to the density, is in
the velocity of the plume. At Titan, Voyager 1 detected an outflow speed of about 10 km/s in the PLS data (Hartle et
al., 1982) while the Low Energy Charged Particle (LECP) detector was unable to detect any flow in the wake of Titan
(Maclennan et al., 1982). In the wake of Europa, on the other hand, the PLS measured velocity was at least 70% of the corotation velocity outside the wake (Paterson et al., 1999). Unfortunately, no information relating to the velocity of the plasma downstream of Callisto is available.
This difference between Europa and Titan is to be expected in view of the vast difference between the densities of the two atmospheres, the intensities of the interactions with the respective magnetospheres and the fundamental differences between Jupiter and Saturn. One such difference is the ability of the ionosphere of Jupiter to impose nearly rigid corotation, by means of its Pedersen conductivity, in the inner magnetosphere where the Galilean satellites orbit (Vasyli¯unas, 1983), in contrast to the inability of that of Saturn to do so at the orbit of Titan. For the case of Titan, Eviatar et al. (1982) found that the only mechanism capable of accelerating ionosphere plasma to the observed velocity was a standing Alfvén wave configuration as modeled for Io by Neubauer (1980). Birkeland currents were found to be inadequate for the task. It is shown in Eviatar et al. (1982) that the rate of acceleration by Birkeland currents of the plasma swept out of a satellite ionosphere is given by the expression:
[..]
where Σp is the Pedersen conductivity in the planet ionosphere, B0 is the surface equatorial magnetic field, ρ the
ambient plasma density and Hm the scale height in the satellite ionosphere. For Titan where L = 20 and the conductivity is low, it is found that the plume cannot be accelerated in this manner.
For Europa, located at much lower L and in the magnetosphere of a planet with a stronger magnetic field
and a much larger Pedersen conductance in the ionosphere [as high as 0.5 mho (Hinson et al., 1998)], the acceleration rate will be much greater than at Titan. The acceleration by an Alfvén wave (Neubauer, 1980) is
also significantly more effective for Europa than for Titan where it is shown to work (Eviatar et al., 1982). Consider the ponderomotive force exerted by the draping of the magnetic field around a satellite as shown in Fig. 4. The acceleration away from the satellite will take place in the region of the atmosphere in which the plasma density is fairly high but ion-neutral collisions can be ignored. The equation of motion is
[..]
The acceleration rate is proportional to the Alfvén velocity and this quantity will be significantly larger in the environment of Europa than in that of either Callisto or Titan because of the weaker magnetic field and lower ambient density at Callisto and Titan compared to these values at Europa. A comparison of the Alfvén velocities at the three satellites, for nominal parameter values is shown in Fig. 5.
http://www.igpp.ucla.edu/public/mkivelso/refs/PUBLICATIONS/Eviatar%20EuropaPlumes.pdf