The state of the plasma sheet at Europa is
incompatible with large local mass loading from a
geophysically active source
1
2
3
D. E. Shemansky1∗ , Y. L. Yung2 , X. Liu 1 ,
J. Yoshii1 , C. J. Hansen3 , A. R. Hendrix 4
L. W. Esposito5
4
1
2
Space Environment Technologies/PSSD,
650 Alameda St, Altadena, CA 91001, USA
Division of Geological and Planetary Sciences, California Institute of Technology,
Pasadena, CA 91125, USA
3
Planetary Science Institute
109 S. Puerto Dr, Ivins, UT 84738, USA
4
Planetary Science Institute
PO Box 954, Montrose, CA 91021, USA
5
University of Colorado, Laboratory for Atmospheric and Space Physics, CO 80303, USA
∗
D. E. Shemansky; E-mail: dshemansky@spacewx.com
5
Deep EUV spectrograph exposures of the plasma sheet at the orbit of Europa,
6
obtained in 2001 using the Cassini Ultraviolet Imaging Spectrograph (UVIS)
7
experiment, have been analyzed to determine the state of the gas. The re-
8
sults are in basic agreement with earlier results, in particular with Voyager
9
encounter meaurements of electron density and temperature. Mass loading
10
rates and lack of detectable neutrals in the plasma sheet, however, are in con-
11
flict with earlier determinations of atmospheric composition and density at
1
12
Europa. A substantial fraction of the plasma species at the Europa orbit are
13
long-lived sulfur ions originating at Io, with ∼25% derived from Europa. Dur-
14
ing the outward radial diffusion process to the Europa orbit, heat deposition
15
forces a significant rise in plasma electron temperature and latitudinal size ac-
16
companied with conversion to higher order ions, a clear indication that mass
17
loading from Europa is very low. Analysis of FUV (far ultraviolet) spectra
18
from exposures on Europa leads to the conclusion that earlier reported atmo-
19
spheric measurements have been misinterpreted. The results in the present
20
work are also in conflict with a report that energetic neutral particles (ENA)
21
imaged by the Cassini INCA (ion and neutral camera) experiment originate at
22
the Europa orbit. An interpretation of persistent energetic proton pitch angle
23
distributions near the Europa orbit as an effect of a significant population of
24
neutral gas is also in conflict with the results of the present work. The general
25
conclusion drawn here is that Europa is geophysically far less active than in-
26
ferred in previous research, with mass loading of the plasma sheet ≤4.5 × 1025
27
atoms s−1 two orders of magnitude below earlier published calculations. Tem-
28
poral variability in the region joining the Io and Europa orbits, based on the
29
accumulated evidence, is forced by the response of the system to geophysical
30
activity at Io. No evidence for the direct injection of H2 O into the Europa atmo-
31
sphere or from Europa into the magnetosphere system, as has been observed
32
at Endeladus in the Saturn system, is obtained in the present investigation.
33
The most recent published observations of Europa (Roth et al 1) show images, obtained
34
using the Hubble Space Telescope (HST), describing plumes observed over a 7 hour period
35
in December 2012. This was the only reported occasion of a phenomenon forcing strongly
36
nonuniform atmospheric distribution, described as evidence for a geophysically active body
2
37
with general similarity to the behavior of Io. Stress during the orbital period was inferred to
38
be responsible. The authors (1) also reported observations in October 1999 and November
39
2012 showing atomic oxygen emission from the sunlit face, absent plume features. The general
40
conclusion from the report (1) is that Europa is a geophysically active body, emitting water
41
vapor and forming a water product atmosphere.
42
The presence of a Europa atmosphere was first reported by Hall et al (2) with detection of
43
atomic oxygen through the O I 3 P − 3s3 So and 3 P − 3s5 So multiplet transitions at ∼1304 and
44
∼1356 Å. The relative brightness of the multiplets, designated here as I(3 S) and I(5 S), is a
45
crucial quantity for the interpretation of how the upper states are excited. Hall et al (2) obtained
46
I(5 S)/ I(3 S) in the range 1.3 to 3.2. On the basis of this range of values it was concluded that the
47
emission was produced by dissociative excitation of O2 . The argument that O2 was the dominant
48
component of the Europa atmosphere and the research history leading up to this conclusion was
49
discussed at the time by Hunten (3). Subsequent observations (4) provided measurements at
50
Europa and Ganymede giving the ratio I(5 S)/ I(3 S) in the range 1.3 to 2.2. The present work
51
disputes the interpretation of this range of values as an exclusive characteristic of dissociative
52
excitation of O2 .
53
The Cassini UVIS experiment provided observations of Europa and the surrounding region
54
on January 6 and 12 of 2001, reported by Hansen et al (5). The analysis presented here is
55
confined to the 2001 DOY (day of year) 012 sequences, because the 2001 DOY 006 exposures
56
are contaminated by strong Io torus emission. Although the authors (5) did not dispute the
57
specifics of atmospheric content, this work focused on the presence of atomic oxygen at the top
58
of the atmosphere and disputed the claim (6) that a significant neutral torus was present at the
59
orbit of Europa.
60
The present work presents the results of a more extended analysis of the UVIS observations
61
at the orbit of Europa during the Cassini 2000/2001 encounter. The new analysis utilizes refined
3
62
flatfielding and calibration, allowing the extraction of deep exposures for both EUV and FUV
63
UVIS spectrographs, whereas previous work (5) was restricted to the FUV. (The Cassini UVIS
64
experiment contains two spectrographs covering the range 550 Å to 1950 Å. A single exposure
65
produces 64 spectra contiguous in space on 1 mrad centers, as described by Esposito et al (7).)
66
This research disputes the substance of most of the previous publications relative to the Europa
67
atmosphere. The establishment of the EUV spectrum has allowed an accurate quantitative de-
68
termination of the ion species in the plasma sheet at the Europa orbit, setting a strong limit on
69
the mass loading rate from the Europa atmosphere.
70
Cassini UVIS EUV spectra of the plasma sheet at Europa. The Cassini UVIS EUV spec-
71
trum extracted from the plasma sheet region at Europa is crucial to the determination of the
72
state of the plasma and in setting a limit on the composition of the Europa atmosphere. The
73
geometry of the observations is described in detail by (5). The observations were obtained
74
with the Europa image stabilized in a particular spatial pixel (pxs) on the detectors. The spatial
75
dimension of the slit in these observations was aligned parallel to the Jupiter rotational axis,
76
providing latitudinal distribution. Fig. 1 shows the integrated flatfielded signal count spectrum
77
averaged over the spatial extent (2.9 RJ ) of the plasma sheet. Fig. 1 also shows the averaged
78
spectrum above and below the plasma sheet in the same exposure sequence, containing the fore-
79
ground/background spectrum. The range to the planet was ∼16 Mkm (106 km), in 2001 DOY
80
012, giving a latitudinal image size of 14 RJ . The dominant feature in the spectrum is the com-
81
bined O II and O III multiplets near 833 Å, with a Rayleigh brightness of ∼ 1.6 R. A distinct
82
peak in the plasma emission is not recognizable at the location of Europa, but the distribution
83
shows a FWHM (Full Width at Half Maximum) of ∼1.5 RJ . The foreground/background spec-
84
trum contains two identified discrete emission lines from the Interplanetary Medium (IPM) and
85
the Local Interstellar Medium (LISM), the He resonance line at 584 Å and the H Lyβ line at
4
86
1025.7 Å. These features have identical brightnesses in both spectra, as shown in Fig. 1. The
87
discrete emission features in the plasma sheet spectrum correspond to those of the emission
88
spectrum from the Io plasma torus, with the exception of the blended transitions in the Europa
89
plasma sheet in the 600 − 630 Å region (8). Possible candidates for these transitions are K II
90
and Cl III, but these have not been investigated quantitatively. The UVIS Io plasma torus and
91
Europa plasma sheet spectra are compared in Fig. S1. The UVIS EUV spectrum of the plasma
92
sheet contains no detectable neutral species emissions.
93
State of the plasma sheet.
94
are shown in Fig. S2. Models of the LISM emissions are included. The plasma model includes
95
O II, O III, O IV, S II, S III, S IV, and S V emissions calculated for an ambient electron tempera-
96
ture of Te = 250000 K (8). A model fit to the Io plasma torus East ansa obtained in the Cassini
97
encounter is shown in Fig. S3. The states of the plasmas at Io and Europa derived from the
98
model calculations are given in Table 1. This table includes results from Voyager 1 encounter
99
in 1979, showing that the plasma at Io is a variable phenomenon. Ion composition from the
100
Voyager Plasma Science Experiment (PLS) (9) is listed although ambiguous (8). The Voyager
101
PLS experiment provides valuable insight into the electron density and temperature distribution
102
between 5.9 and 9.4 RJ (10), discussed in (8). Table 1 includes the calculated mean volumetric
103
radiative cooling rates, which constitute ∼80% of the energy expended to maintain the plasma
104
in steady state (11). The quantitative significance of the energy rates is discussed in (8). Plasma
105
energy partitioning is discussed by (10, 11). Sulfur ions, originating from Io, at the orbit of
106
Europa are an expected component in the plasma (11,13,14). Evidently the plasma at the Eu-
107
ropa orbit is dominated by ions originating at the Io plasma torus. The outward diffusion of
108
ions from 5.9 RJ is characterized by substantially increased ambient electron temperatures (8)
109
accompanied by the conversion to higher order ions. The ratios of ion species densities at 5.9
Plasma model calculations fitting the EUV plasma sheet spectrum
5
110
RJ over those at 9.4 RJ , shown in Table 2, based on the present results decrease significantly
111
with increasing positive charge, indicating overall that conversion to higher order ions occurs
112
more rapidly than other loss processes. The mean density of S II at Europa is a factor of 58
113
below the value at 5.9 RJ , while S V at 9.4 RJ is only a factor of 3.6 less dense than S V at
114
5.9 RJ . These facts are indicative of low mass loading rates at Europa. The low density of the
115
plasma at 9.4 RJ indicates that electron impact is the main controlling factor in volumetric rate
116
processes. It is probable that the plasma at 9.4 RJ is never in statistical equilibrium. The Io
117
plasma torus is known to be time variable on a scale of months, but the slow diffusion rate of
118
ambient ions outward will filter the magnitude of the variation. The lifetime of S IV against
119
electron impact conversion to S V at 9.4 RJ , for example, is 120 days.
120
Energy budget. Table 1 includes calculated radiative cooling rates, which account for
121
∼80% of the energy loss from the plasma (11). The plasma is maintained primarily by Coulomb
122
electron-electron and proton-electron heating by the inward diffusion of the hot magnetospheric
123
flux passing through the local plasma sheet. Calculated rates are included in Table 1. The basis
124
for the calculations is discussed in (8). The results are consistent with the measured conditions
125
in the magnetosphere at Voyager 1 encounter, with, however, a perturbation involving the flow
126
of energy through the Io plasma torus at the time of the Cassini UVIS observations in 2001
127
(8). The results are quantitatively aligned with the argument that the heterogeneous inwardly
128
diffusing magnetospheric plasma is the primary energy source for the maintenence of the Io
129
plasma torus, and that Europa plays a geophysically passive role in the activity in the 5 RJ to 10
130
RJ region (8). The energy available for the development of the plasma sheet at 9.4 RJ is vastly
131
larger than the energy loss rate determined here from observation (8). The energy required to
132
maintain the Io plasma torus at Voyager 1 encounter, is 54 times greater than the energy required
133
for the plasma sheet at Europa (Table 1), while the primary energy pool for these two regions
6
134
at the time of Voyager differed by only a factor less than two; The energy is available at Europa
135
for a plasma on the scale of the Io torus, but Europa does not provide the required mass loading
136
(8).
Table 1: Derived plasma properties
Radial Location (RJ )a
Te (1000K)
[O I] cm−3
[O II] cm−3
[O III] cm−3
[O IV] cm−3
[S I] cm−3
[S II] cm−3
[S III] cm−3
[S IV] cm−3
[S V] cm−3
Σ[Ni] cm−3
[e] cm−3
ΣNi X Vg
Σ[Oi]/[Si]
Eradh
Ecouli
a
b
5.9b
60.
29.f
936.
33.8
0.074
3.7
287.
409
46.7
0.130
1745.
2250.
60.2
1.34
392.
399.
5.9c
75.
···
118.
148.
≤3.
···
70.
355.
71.
2.7
765.
1418.
26.4
0.53
565.
567.j
9.4d
···
···
44.
22.
···
···
6.4
8.1
5.9
···
87.2
130.
···
3.
···
···
9.4e
250.
···
8.2
13.4
4.3
···
1.2
10.7
7.6
0.75
46.1
96.
6.94
1.38
7.3
6.3
Relative to Jupiter center
Center of Io hot torus plasma; Voyager 1 encounter Mar 1979
(11)
c
Center of Io hot torus plasma; Cassini Jan 2001
Europa location, Voyager PLS Mar 1979 (9)
e
Europa location, Cassini UVIS Jan 2001
f
Theoretical physical chemistry fitting observation (11)
g
Total ion content in toroid volume (× 1033 )
h
Mean volumetric radiative cooling rate 10−15 ergs cm−3 s−1
i
Calculated Coulomb heating of ambient electrons 10−15 ergs
−3 −1
cm s (8)
j
Requires hot heterogeneous electron density inconsistent with
Voyager results (8)
d
7
Table 2: Radial ion density progression
Species
ri a
a
O II
14.4
O III
11.0
O IV
≤0.7
S II
58.
S III
33.
S IV
9.3
SV
3.6
Number density ratio Cassini UVIS 2001 ri = Ni (5.9 RJ )/Ni (9.4 RJ )
8
1.0
counts 1000 s-1 pxs-1 pxw-1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
550
650
750
850
950
1050
1150
λ (A)
Figure 1: UVIS EUV 2001 DOY 012 spectrum (red trace) integrated latitudinally over pxs 28
− 40 with Europa contained in pxs 33 (integration time ti = 148 h), compared to the spectrum
(black trace) obtained north and south of the plasma sheet in the same exposure sequence (ti
= 398 h). The spectra are in signal counts normalized to 1000 sec, flatfielded with scattered
photon background removed. Statisical error bars (1σ per pxw (wavelength dimension pixel)
before filtering) are shown for the plasma sheet spectrum. The background spectrum contains a
residual O II/O III feature at 833 Å, with two identified emission lines from the IPM and LISM,
the He resonance line at 584 Å and the H Lyβ line at 1025.7 Å. The LISM features have identical
brightnesses in both spectra. The remaining discrete features in the plasma sheet spectrum are
attributed to oxygen and sulfur ions, primarily outward diffusion products from the Io plasma
torus.
9
The UVIS
137
Emission properties of the Europa atmosphere and oxygen rate processes.
138
FUV exposures contain the only neutral gas emissions detected in the 9.4 RJ region from this
139
experiment. Fig. 2 shows the 2001 DOY 012 spectrum of the O I 3 P − 3s3 So and 3 P − 3s5 So
140
emissions at and adjacent to the position of Europa on the instrument detector. The solar phase,
141
ϕ = 116o , and consequently solar reflection features are not detected in the spectrum. The
142
exposures on 2001 DOY 006 are similar but contaminated with Io torus emission (5). The
143
image of Europa falls in pxs 32. The pxs 32 spectrum contains an indication of weak sulfur
144
ion emission from Europa, but these features will not be discussed here. The spectra in pxs
145
31, 32, and 33, with exposure times of 11 hours, are shown along with a modeled statistically
146
equilibriated spectrum of simultaneous electron and solar photon excited O I, fitting the pxs 32
147
spectrum. The O I electronic model architecture contains 54 multiplet states (details will not
148
be given here, see (8)). The O I model calculations use the combined forcing of the electron
149
environment and a high resolution solar flux model (15). The solid angle of the Europa image
150
on the UVIS detector is 50 times smaller than the pixel size, so it acts as a point source for
151
the spectrograph, giving a spectral resolution of about 2 Å. This is sufficient resolution to show
152
the optical thickness in the 3 P − 3s3 So multiplet transitions. The 3 P − 3s3 So feature in Fig. 2
153
has evident opacity, and an accurate fit to the multiplet feature has been obtained for a line-of-
154
sight (los) abundance of 4.7 × 1012 cm−2 (8). The O I (3 P − 3s3 So )/ (3 P − 3s5 So ) multiplet
155
pair is accurately modeled at an electron temperature of Te = 90000 K, as shown in Fig. 2.
156
The presence of O2 is not required to explain the O I emission in Fig. 2. The spectrum in pxs
157
31 shows a much smaller opacity and the intensity ratio I(5 S)/I(3 S) is commensurately lower.
158
The spectrum in pxs 33 shows no evidence of electron excited O I. Comparison of these results
159
with previous observations provides useful additional insight. Table 3 shows the compiled
160
observational results on the Europa atmosphere, showing 7 results from separate exposures.
161
The O I multiplet brightnesses are disk averaged values. The listed emission brightnesses in
10
162
Table 3 for the observations reported by Roth et. al. (1, Table S1) are similar but not equal
163
to the values derived from the flux information in (1), given in Table 3. The explanation is
164
provided in (8). The O I emission values do not show strong variation from 1994 to the present
165
according to this analysis. The brightness quantities in Table 3 are derived from flux values.
Table 3: Europa atmosphere observations
Expta
UVISg
HSTk
HSTn
HSTo
HST
HST
HST
Date
2001 012h
1994 153
1996 203
1996 212
1999 278
2012 313
2012 365
I(3 S)b
71. ±6.i
29. ±6.l
30. ±3.2
55. ±4.
72.p
64.
41.
I(5 S)c
67.±6.
64.±6.
64.±4.6
72.±5.2
111.
76.r
77.s
I(Lyα)d
···
···
···
···
≤30.q
≤20.
45.±31.
[O I]ℓ e
4.7j
···
···
···
···
···
···
[O2 ]ℓf
···
(240. − 1200.) m
(350. − 1100.)
(370. − 1400.)
···
···
···
[O II]ℓ
≤ 0. 31
···
···
···
···
···
···
I(5 S)/I(3 S)
0.94
2.2
2.1
1.3
1.5
1.2
1.9
a
Observatory
O I(3 P − 3s3 So ) at 1304 Å Brightness in Rayleighs
c
O I(3 P − 3s5 So ) at 1356 Å
d
H I resonance line 1216 Å
e
los abundance ( 1012 atoms or molecules cm−2 )
f
See text.
g
Present work, Trailing surface
h
read 2001 012 as 2001 DOY 012
i
average brightness over disk inside 1.15 RE
j
Inferred from I(5 S) and (16) electron environment. Opacity abundance from analysis of I(3 S) is also
4.7 × 1012 cm−2 .
k
Hall et al. (2), trailing surface; based on Table 1 of (4)
l
average brightness over disk
m
from (4)
n
Hall et al. (4), leading surface
o
Hall et al. (4), trailing surface
p
O I calculated from Fig. S5 (A) of Roth et al. (1), average brightness over 1.15 RE disk, see text.
q
Averaged H Lyα emissions as reported by (1)
r
O I calculated from Fig. S5 (B) of Roth et al. (1)
s
O I calculated from Fig. S5 (C) of Roth et al. (1)
b
166
Table 3 shows O2 abundances in column 7 required if the O I emission is produced by
167
dissociative excitation (2). These values are 100 to 1000 times larger than the present derived
168
abundance for an O I atmosphere. Given an O I atmosphere the major ionosphere ion must
169
be O II. The presence of O II in the Europa atmosphere has not been directly observed in the
170
UVIS exposures. It is clear that the plasma sheet densities in 2001 DOY 012 peak broadly
11
171
around the location of Europa, but a peak in O II emission confined to the pixel containing the
172
Europa image has not been detected above the backgound emission from the plasma sheet. The
173
upper limit in column 8 of Table 3 is close to the abundance expected if the major ion is O II.
174
The electron vertical profile has been measured in Galileo radio occultations by Kliore et al.
175
(16). The mean of 6 occultations provided a vertical profile showing electron density peaking
176
at the surface with a value [e] ∼9000 cm−3 . The electron abundance (16) is close to the limit
177
given for O II in Table 3. The neutral/ion mixing ratio at the surface from this analysis is then
178
[N]/[e] ∼15., leading to the conclusion that the Europa atmosphere is actually a weakly ionized
179
plasma with a density maximum at the surface. There are no atomic kinetic collisions in this
180
nominal atmosphere; the primary collision partners are electrons. The atmospheric plasma on
181
Europa can be compared to the state of the Enceladus plasma torus in the Saturn system. Model
182
calculations of an energy equilibrated H2 O sourced system produces [N]/[e] ∼60, with Te ∼
183
16000 K (12). The state of the gas in these systems is limited by the energy available in the
184
magnetospheric structure, and the mass insertion rate. In this case the Europa atmosphere is a
185
higher temperature, more heavily ionized plasma than the Enceladus torus.
186
Atmospheric models.
187
on the interpretation that the disk averaged emissions I(3 S) and I(5 S) are from dissociatively
188
excited O2 (17,18). A model based on an O I dominated atmosphere, developed from the Liang
189
et al. (19) model for Callisto is presented here. The source of CO2 in the model is suppressed
190
and overall source rates reduced, leaving an atmosphere dominated by O I and H I, derived from
191
H2 O photolysis. The source process is not necessarily an accurate description of how the atmo-
192
sphere has developed, but it provides a more realistic structure for comparison to observation.
193
Fig. S4 shows the vertical distribution of the primary species. The dominant ion is O II, with a
194
vertical abundance within a factor of 2 of the Kliore et al. (16) measurements. The densities
The most recent published atmospheric models for Europa are based
12
195
and scale heights are shown in Table 4, in comparison to the observational results discussed
196
here. The abundance of H I in the model would not be detectable in the observations to date.
197
The abundance of O2 (Fig. S4) in the model, however, is roughly a factor of 2 to 3 too large
198
for compatibiliy with the present observations. The major loss of O II in the model is through
199
reactions rr(9) and rr(10). An O II ionosphere is unsustainable on Callisto because of the fast
200
rr(10); This is not an issue for Europa. The scale heights in the model are much smaller than
201
the observation derived quantities, because the source function is cold. The model assumes that
202
H2 O is extracted from the surface by nuclear sputtering. This may not be an accurate description
203
if exchange processes play a significant role, in which the particles diffused into the vacuum are
204
mainly atomic (22).
205
Rate processes. The conclusion in earlier work that the Europa atmosphere was pri-
206
marily O2 was based on an observed emission ratio I(5 S)/I(3 S) consistently greater than 1.
207
The observed values for I(5 S)/I(3 S) range from 1. to 2.2 (Table 3). The reactions producing
208
emission in the two multiplets that require consideration here are as follows:
e + O2 (X) → e + O∗ + O∗
(1)
e + O2 (X) → e + O(3 P) + O(3s5 So )
(2)
e + O2 (X) → e + O(3 P) + O(3s3 So )
(3)
e + H2 O(X) → e + H(2s, 2p) + products
(4)
e + H2 O(X) → e + O(3s5 So ) + products
(5)
e + H2 O(X) → e + O(3s3 So ) + products
(6)
e + hν + O(3 P) → e + O(3s5 So )
(7)
e + hν + O(3 P) → e + O(3s3 So )
(8)
O+ + O2 → O + O+
2
13
(9)
O+ + CO2 → CO + O+
2
(10)
209
Rate coefficients critical to the present discussion are given in Fig. S5. As expected, reaction (rr)
210
rr(1) (derived from Itikawa (24)) has the most agressive rates, and it is this reaction that forces
211
high mass loading of the plasma sheet if the europa O I emissions are assumed to be caused by
212
rr(2) and rr(3). The assumption of an O2 dominated atmosphere introduces three difficulties.
213
1) The ratio I(5 S)/I(3 S) from rr(2) and rr(3) ranges from 3.5 at Te = 50000 K to 2.6 at Te =
214
250000 K, out of range of the observations (Table 3) according to the current rates shown in
215
Fig. S5. 2) Based on the Saur et al. (17) calculations using the HST observed emissions, an
216
injection rate of 2 × 1027 O I atoms s−1 is indicated for the plasma sheet. This is not compatible
217
with the state of the observed plasma. 3) The analysis of the UVIS observations described
218
above, taking opacity into account, fits the observed spectrum and I(5 S)/I(3 S) ratio with the
219
present model for rr(7) and rr(8) (Fig. 2). Further detail is given in (8). Roth et al. (1), indicating
220
emissions that are consistent with rr(4) and rr(6), is the only published work showing evidence
221
of H2 O in the near Europa environment (see Fig. S5). The rates for dissociative excitation
222
of H2 O shown in Fig. S5 are derived from (25). The observations examined here show no
223
evidence of H Lyα emission in the plasma sheet or Europa atmosphere. The stochastic process
224
consituting the observations listed in Table 3, does not show evidence of deviation from an
225
overall stationary state. The imaging of the emission on the Europa disk does show variability,
226
described as auroral activity (1), and general evidence of a variable Europa environment has
227
been discussed (20, 23), but evidence for strong localized events that significantly affect the
228
Europa global atmosphere, is absent in the present assessment. The Roth at al. analysis (1) of
229
the plume observation of 2012 DOY 365 did not develop an estimate of mass loading of the
230
magnetosphere. The plume altitude limited at 200 km, suggesting that a large fraction of the
231
gas did not did not reach escape velocity and did not significantly affect the plasma sheet.
14
232
Nominal atmosphere. Table 4 shows the nominal Europa model atmosphere emerging
233
from the present work. Examination of the details of energy balance and rate processes are
234
beyond the scope of this preliminary work. Some general conclusions, however, can be drawn.
235
The ionosphere, which Kliore et al. (16) have shown interfaces with the solid surface, will be
236
sustained by both electron impact and charge capture by plasma sheet ions, which penetrate to
237
the surface. Substantial recycling between solid state and gas is expected, with sputtering and
238
dissociative excitation to neutral atoms through multicharged particle penertration of the surface
239
(21,22). It is evident, however, that Europa does not easily deliver mass to the magnetosphere.
240
The strongest rate processes are likely to be solid to gas recycling.
Table 4: Nominal Europa atmosphere
[O I]a
1.5c
13.d
[O II]
0.1
0.23
[e]
0.1
0.23
Tg (K)
800.
150.
H(O I)b
300.
31.
H(e, O II)
240.
31.
[O I]/[e]
15.
56.
densities ([]) × 105 cm−3 at the solid surface.
Scale height (H) in km
c
Present observation based results
d
Present preliminary model
a
b
The state of the plasma sheet es-
241
Sensitivity of the plasma sheet at 9.4 RJ to neutral gas.
242
tablished here depends on a very low mass loading rate to retain significant ion charge levels.
243
A qualitative assessment from the nature of the the source of the ion population indicates that
244
in the transport time for delivery of sulfur ions from 5.9 RJ to 9.4 RJ the plasma electron
245
volumetric energy increases from 75000 K to 250000 K, against losses to collisional radiative
246
reactions, ionization, and dielectronic recombination (11). It is evident that the heterogeneous
247
magnetospheric electrons deposit a net large amount of energy into the ambient population
248
(electron-electron Coulomb interaction) to raise the energy density of the resident electrons
249
against exponentially rising loss rates (11). This is a qualitative indicator that mass loading
250
from Europa must on average be at a very low level. Without carrying out a detailed physi15
251
cal chemistry energy balanced calculation such as (11, 12) approximate limiting rates and time
252
scales can be established to provide values of core parameters. Table 5 lists the pivital quan-
253
tities. The measurement of O IV at 9.4 RJ provides one of the longer time-constants. The
254
conversion of O III to O IV by electron impact (rr(11), see (8)) has a characteristic time of 124
255
days (Table 5). This process takes place in competition with diffusive loss (rr(13)), dielectronic
256
recombination (rr(12)), and charge capture from neutral atoms (rr(14)) and lower order ions.
257
The probability for rr(11) ρ11 = 9.3 × 10−8 s−1 must be more rapid than charge capture (21) or
258
other losses in order to build population as it does during transport from 5.9 RJ . This requires
259
a mean neutral density in the plasma sheet of less than 3 cm−3 . Neutral emission upper limits
260
are at a level of ∼3 cm−3 . The rate of neutral injection at this level in the plasma sheet is ∼4
261
× 1025 atoms s−1 (Table 5). This is a crude estimate of a sustainable level of mass loading for
262
the plasma sheet. Given that the electron density and temperature measurements in the Europa
263
plasma sheet at the time of Voyager, Galileo, and Cassini constitute a stochastic process in ad-
264
dition to the Europa atmospheric emission results, the indication is that the observations on a
265
time scale of years constitute a steady state on a long time scale. The injection of atoms at a rate
266
1000 times the above mean value may take place for a limited interval on a frequency scale of
267
years, but the mean value must not exceed the ∼4 × 1025 atoms s−1 rate. The evidence for large
268
events in the Europa atmosphere does not appear in the stream of stochasic samples to date.
269
The influence of plasma sheet ions on the generation of atmospheric gas is a factor to con-
270
sider, given that multiply charged ions are involved. The sputtering of solid surfaces is deter-
271
mined by momentum transfer, electronic excitation, and exchange reactions. The last process
272
has generally received little attention because the process has a small role in singly ionized ion
273
impacts. Multiply charged ions, however, can contain a significant amount of internal energy
274
that is delivered to the solid molecules in charge capture and subsequent emission in multiple
275
x-ray photon deposition into the solid (21,22). Charge capture processes in the solar system
16
276
have been discussed in (22). The plasma sheet ion impact energies at the surface of Europa are
277
only at the low keV level, and only exchange reactions can contribute significantly to sputtering.
278
The exchange sputtering process is one of ionization by the multiply charged ion with simulta-
279
neous photon emission. The photo electrons released produce atomic products in dissociative
280
recombination, with enough kinetic energy to diffuse into the vacuum. O IV has enough energy
281
to contribute to the sputtering process. Magnetospheric protons, at much higher kinetic energy
282
sputter through momentum transfer. Efficiencies are discussed in (22). The calculated rates
283
for the O IV sputtered contribution given in Table 5 do not include kinetic energy distributions,
284
but the values indicate that plasma sheet ions may have a significant contribution to the Europa
285
atmosphere. The current calculated population of neutrals in the plasma sheet can be compared
286
to the Mauk et al. (6) result from ENA imaging, and the Saur et al. (17) model source rates,
287
given in Table 5. The rates by (17) are much larger because of the assumption (Fig. S5) that the
288
atmosphere was dominated by O2 .
Table 5: Critical plasma sheet volumetric and gas-surface reaction rates
ρO23 a
94.
ρO32 b
3.2
[Nℓ ]c
3.
Nℓ Td
4.6
Nℓ Te
90.
S(Nℓ )f
4.5
S(Nℓ )g
200.
Y¯S h
3.0
F O3 i
0.45
S¯Y j
1.3
Probability (ρ (× 10−9 sec−1 )) for electron impact ionization of O III, rr(11)
Probability for dielectronic recombination of O IV, rr(12)
c
Limiting neutral density (cm−3 ) to allow steady state of plasma sheet O IV population
d
Limit to total number of neutral atoms in the plasma sheet ( × 1032 )
e
Mauk et al. (6)
f
Limiting source rate of neutral particles maintaining the plasma sheet (× 1025 s−1 )
g
Saur et al. (17)
h
Sputtering yield of neutral atoms by plasma sheet O IV ions (× 1025 atoms s−1 )
i
Flux of O IV ions into the Europa surface (× 108 cm−2 s−1 )
j
Flux of surface sputtered atoms into the vacuum (× 108 . cm−2 s−1 ) (8)
a
b
289
Conclusions and discusssion
290
1. Deep exposures using the Cassini UVIS EUV spectrograph have revealed a low density
291
plasma at the Europa orbit, composed mainly of ions originating in the Io plasma torus.
17
292
This is a predictable finding, because the radial electron density distribution measured
293
previously using the Voyager, and Galileo missions produced similar densities and tem-
294
peratures to the values in the present analysis. No neutral atomic emissions were detected,
295
and upper limits from approximate plasma analysis and the emission results indicate a
296
mean neutral density of 3 cm−3 in the plasma sheet. There is an indication that Europa
297
contributes oxygen to the plasma. Mass loading of the plasma sheet is estimated to be ∼
298
5 × 1025 atoms s−1 , a factor of 40 below the Saur et al. (17) O2 model (Table 5).
299
2. The atomic oxygen multiplet emissions from Europa identified in previous observations
300
beginning in 1994, have been misinterpreted as dissociative excitation of an atmosphere
301
dominated by O2 . The consequence of this earlier conclusion was the adoption of a
302
model that inserted significant mass loading into the plasma sheet at the Europa orbit.
303
The present analysis of the Cassini UVIS FUV measurements indicates that the emis-
304
sion spectrum averaged over the Europa disk, is electron excited atomic oxygen. The
305
atmosphere at Europa in the present analysis is atomic oxygen at a density two orders of
306
magnitude below the previous O2 dominated model.
307
3. The heterogeneous hot component of the magnetospheric plasma deposits most of the
308
energy required to maintain the Io plasma torus. The energy density at the orbit of Europa
309
is at a similar level and given mass loading, could support a plasma torus on the same
310
scale. This has never been observed. The energy passing through the plasma at the Io
311
torus (Table 1) is two orders of magnitude greater than in the plasma sheet at Europa
312
at the time of Voyager encounter in 1979, and at Cassini encounter in 2001. Given the
313
stochastic remote sensing evidence, mass loading at the Europa orbit is not consistent
314
with a geophysically active body.
315
The overall impact of the present work on the assessment of the level of geophysical activity is to
18
316
indicate that the evidence for vents releasing gas into the atmosphere is at best tenuous. Russell
317
(23) has pointed out that periodic plasma enhancements from Pioneer 11 measurements were
318
interpreted by Intriligator and Miller (20) as a continuous plume at Europa. The phenomenon
319
was, however, found to be non-repeatable in work with Galileo (23). These (20) results could be
320
interpreted as temporal variations in the density of the corotating plasma (23). Europa has the
321
property of a conducting sphere with a radius close to the solid surface radius. Given the low
322
density of the kinetically collision free atmosphere derived here, the conducting sphere could
323
be the highly developed ionosphere rather than a subsurface conductive fluid.
324
The observational evidence in the region of the Galilean satellites shows consistency in
325
magnitudes. Sittler and Strobel (10) from Voyager measurements show plasma sheet electron
326
densities decreasing outward from 5.9 RJ with a sharp drop at 7.2 RJ coincident with a sharp
327
rise in temperature, an indicator of a distinct decrease in mass loading. The evidence in the
328
present work is consistent with this structure in showing a plasma at Europa that has converted
329
the ion population to higher charge species at rates faster than dielectronic recombination and
330
charge capture on neutral gas (Table 2). At 9.4 RJ during Voyager encounter the density was
331
∼40 cm−3 , and Te ∼250000. K.
332
The Roth et al. (1) HST observation in Dec 2012 is unique in two respects:1)Two plumes
333
are identified in the south polar region, giving emission in H Lyα and the O I 1304 Å multi-
334
plet, indicators of dissociative excitation of H2 O. 2) The emission from the disk does not have
335
verified H Lyα emission, but shows atomic oxygen emission at a mean brightness similar to
336
all other reported observations (Table 3). Roth et al. (1) report the emissions against the disk
337
as auroral phenomena with microscale variability, but with brightness averaged over the disk at
338
the same level as the other reported observations. Variability is certainly present in the system
339
involving Europa, but the present research indicates it is a magnetospheric phenomenon, rather
340
than geophysical instability. The Io plasma torus is well known to be variable, with significant
19
341
changes reported on time scales of months to years. As a primary forcing function on the sys-
342
tem Io will impose variability on the Europa environment (23). An event on Io propogates to
343
the Europa orbit, but filtered by the long physical chemistry and diffusion time-constants. The
344
general conclusion is that Europa does not show evidence of internal geophysical activity. The
345
Saur et al. (17) model for Europa has mass loading at a rate similar to that at Io. The mass
346
loading rate at the Io torus is ∼7 × 1027 atoms s−1 (11,18). The (17) rate at Europa, ∼2 × 1027
347
atoms s−1 , is not plausible given the known conditions.
348
The calculations of O IV exchange sputtering (Table 5) indicate that this process, which
349
diffuses dissociated atomic species into the vacuum, could be a significant contributor to the
350
maintenence of the tenuous atmosphere.
351
The work presented here provides approximate quantities relevant to the state of the Europa
352
environment. Model calculations on the detailed level of the energy balanced plasma calcula-
353
tions such as (11, 12) for the Io and Enceladus environments will provide further insight on
354
system forcing and the formation of the Europa atmosphere.
20
counts 1000s pxs pxw
1
-
1
-
-1
1.2
2001 012
1.0
Te = 90000. K
0.8
pxs 32
0.6
pxs 31
pxs 33
0.4
[O I] = 0.0 cm-2
[O I] = 4.7 X 1012 cm-2
[e] = 14000. cm-3
{O I (3Pg - 3s 5So)} = 67 R
0.2
0.0
1270 1280 1290 1300 1310 1320 1330 1340 1350 1360 1370
λ (A)
Figure 2: UVIS accumulated FUV spectra (2001 DOY 012) at the Europa ansa, in the 1270. −
1370. Å region, compared to model simulations, showing the O I 3 P − 3s3 So and 3 P − 3s5 So
emission multiplets. Spectra in pxs 31 (blue trace), 32 (light red trace), and 33 (green trace) are
shown, with Europa contained in pxs 32. Integration time is 11 hours. The solid black trace is a
model calculation with solar and electron forcing, fitting the spectrum in pxs 32 with an opacity
abundance of 4.7 × 1012 cm−2 . The optically thin model is shown as a dotted black trace. The
model calculation shown fitting both O I multiplets requires an electron temperature of Te =
90000. K. The electron density in the model is [e] = 14000. cm−3 , and the gas temperature Tg
= 1000. K. The brightness of the O I 3 P − 3s5 So forbidden multiplet is 67 R as indicated on
the plot. An upper limit non-LTE model calculation for the excitation of H2 is shown as a light
blue trace on the plot.
21
355
356
357
358
359
References
1. L. Roth, J. Saur, K. D. Retherford, D.l F. Strobel, P. D. Feldman, M. A. McGrath, F.
Nimmo, Science 343,171(2012) doi:10.1126/science.1247051.
2. D. T. Hall, D. F. Strobel, P. D. Feldman, M. A. McGrath, H. A. Weaver , Nature 373, 677
(1995).
360
3. D. M. Hunten, Nature 373, 654 (1995).
361
4. D. T. Hall, P. D. Feldman, M. A. McGrath, D. F. Strobel, Astrophys. J. 499, 475 (1998).
362
5. C. J. Hansen, D. E. Shemansky, A. R. Hendrix, Icarus 176, 305 (2005)
363
364
365
doi:10.1016/j.icarus.2005.02.007
6. B. H. Mauk, D. G. Mitchell, S. M. Krimigis, E. C. Roelof, C. P. Paranicas, Nature 421,
920 (2003).
366
7. L. E. Esposito, C. A. Barth, J. E. Colwell, G.M. Lawrence, W. E. McClintock, A. I. F.
367
Stewart, H.U. Keller, A. Korth, H. Lauche, M. C. Festou, A .L. Lane, C. J. Hansen, J.
368
N. Maki, R. A. West, H. Jahn, R. Reulke, K. Warlich, D. E. Shemansky, Y. L. Yung,
369
SpaceSci.Rev. 115, 299(2004).
370
8. See supporting online material.
371
9. F. Bagenal, , J. Geophys. Res. 99, 11043(1994)0148-0227/93JA-02908$05.00
372
10. E. J. Sittler Jr, D. F. Strobel, J. geophys. Res. 92, 5741 (1987).
373
11. D. E. Shemansky, J. Geophys. Res. 93, 1773(1988).
374
12. J. Yoshii, Univ. Southern California PhD Disseration, Sep(2011).
22
375
13. D. E. Shemansky, G. R. Smith, J. Geophys. Res. 86, 9179(1981).
376
14. D. E. Shemansky, Astrophys. J. 242, 1266(1980).
377
15. R. Killen, D. Shemansky, N. Mouawad, Astrophys. J Supp. 181, 351(2009) doi:10.1088/0067-
378
0049/181/2/351
379
16. A. J. Kliore, D. P. Hinson, F. M. Flasar, A. F. Nagy, T. E. Cravens, Science 277, 355(1997).
380
17. J. Saur, D. F. Strobel, F. M. Neubauer, J. Geophys. Res. 103, 19947 (1998) doi:10.1029/97je03556.
381
18. W. H. Smyth, M. L. Marconi, Icarus 181, 510 (2006) doi:10.1016/j.icarus.2005.10.019
382
19. M. C. Liang, B. F. Lane, R. T. Pappalardo, M. Allen, Y. L. Yung, J. Geophys. Res. 110,
383
(2005)
384
20. D. S. Intriligator, W. D. Miller, J. Geophys. Res. 87, 8081(1982)doi:10.1029/JA087iA10p08081.
385
21. F. Aumayr, H. Winter, Phil. Trans. R. Soc. Lond. A 362, 77 (2004) doi:10.1098/rsta.2003.1300.
386
22. D. E. Shemansky, Rarefied Gas Dynamics 32rd Int. Symp., AIP Conf. Proc. 663 Ch. 9,
387
687(2003). doi:10.1029/2004je002322.
388
23. C. T. Russell, Planet. Space Sci. 53, 473 (2005) doi:10.1016/j.pss.2004.05.003.
389
24. Y. Itikawa, J. Phys. Chem. Ref. Data 38, 1 (2003) DOI:10.1063/1.3025886
390
25. H. D. Morgan , J. E. Mentall, J. Chem. Phys. 60, 4734 (1974) doi: 10.1063/1.1680975
391
26. X. Liu, D. E. Shemansky, J. T. Hallett, H. A. Weaver, Astrophys. J. Supp. 169, 458(2007).
392
27. W. H. Smyth, D. E. Shemansky, Astrophys. J. 271, 865 (21983) doi:10.1016/j.icarus.2005.10.019
393
28. M. Paonessa, J. Geophys. Res. 90, 521(1982)0148-0227/085/004A-1123$02.00
23
394
29. A. Lagg, N. Krupp, J. Woch, Geophys. Res. Lett. 30, 1556(2003)doi:10.1029/2003GL017214
395
30. W.-H. Ip, D. J. Williams, R. W. McEntire, B. H. Mauk, Geophys. Res. Lett. 35,
396
829(1998)0094-8534/98/98GL-00472$05.00
397
31. T. A. Cassidy, C. P. Paranicas, J. H. Shirley, J. B. Dalton III, B. D. Teolis, R. E. Johnson ,
398
L. Kamp, A. R. Hendrix, Planet. Space Sci. 77, 64 (2013)doi:10.1016/j.pss.2012.07.008
399
32. J. F. Cooper, R. E. Johnson, , B. H. Mauk, JB. H. Garrett, N. Gehrels, Icarus 149, 133
400
(2001)doi:10.1016/j.pss.2012.07.008
401
33. E. C. Sittler Jr., J. F. Cooper, R. E. Hartle, W. R. Paterson, E. R. Christian, A. S. Lipatov,
402
P. R. Mahaffy, N. P. Paschalidis, M. A. Coplan, T. A. Cassidy, J. D. Richardson, B. Fegley
403
Jr., N. Andre, Planet. Space Sci. 88, 26 (2013)doi.org/10.1016/j.pss.2013.01.013
404
34. A. L. Lane, R. M. Nelson, D. L. Matson, Nature 292, 38 (1981).
405
35. A. Eviatar, Y. Meklar, J. Gephys. Res. 89, 1496 (1984).
406
36. A. Eviatar, G. L. Sisco, T. V. Johnson, D. L. Matson, Icarus 47, 75 (1981).
407
Acknowledgements DES, XL, JY and were partially supported by the Cassini UVIS pro-
408
gram through contract to SET. YLY was supported in part by the Cassini UVIS program via
409
NASA grant JPL.1459109 to the California Institute of Technology. CJH, and ARH were par-
410
tially supporedby the Cassini UVIS program through contract to PSI.
24
411
Supporting online material
413
The state of the plasma sheet at Europa is incompatible with a
geophysically active source
414
Shemansky et al.
412
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
Unidentified emission in the EUV spectrum. The plasma sheet spectrum contains emission
in the the 600 − 630 Å region that is found spatially mixed with the known emission species.
Emission from the minor species K II and Cl III is possible, but investigation is needed. This
region of the spectrum also contains transitions in O II, O III, O IV, and O V, but the model
calculations produce emission levels at least an order of magnitude too weak to account for
these features as normal electron impact excitation.
Plasma sheet model temperatures. The electron temperature used in fitting the observed
spectrum at 9.4 RJ is the same value as that derived in the Voyager 1 PLS measurement at
this location in the magnetosphere (10). Relative strengths of emission lines at temperatures
this high lose sensitivity to the exact value and uncertainty in temperature from emission line
relative strength is ∼ ±30000K. Atomic emission properties in the temperature range below
100000 K show much higher sensitivity to the exact temperature value.
The Voyager PLS measurement of the ion mass spectrum. The ion spectrum produced
by this instrument was on an energy per charge scale, so that for example O II and S III were
coincident in the signal output and therefore not uniquely extractable.
The UVIS EUV and FUV spectra. The model calculations in the reduction of the spectra
shown in Figs. S2 and 2 are based on the methodology described in (13) using updated physical
parameters where necessary. The model code calculates statistical equilibrium in the gas with
the applied fixed forcing of the system. The collision strengths are established on the basis of
the principle of detailed balance so that at very high density, the state populations approach a
Maxwell-Boltzmann distribution. The calculations here are in non-LTE. The O I architecture
used in fitting the spectrum of Fig. 2 contains 54 states; calculations take place at the electron
orbital level, predicting emission in 887 transitions. Other species in the model calculations are
developed to a similar level (13). Emission in the O I system depends on the combined effects
of solar flux and electron impact. In general there is dependence on electron density. The rate
data given in Fig. S5 for O I are for electron density [e] = 104 cm−3 . The solar flux model,
which combines observations with solar plasma modeling, is required to be at high resolution
(∼200000 − ∼600000) in order to aviod errors in the fluorescence calculations (15).
Limits on neutral species based on spectral analysis The limits on H I and H2 in the Europa
plasma sheet and atmosphere are less definitive than those determined from the plasma sheet
properties, but these quantities contribute to the accumulated evidence. O I or other neutrals
have not been detected in emission from the plasma sheet. The impact of these results has
relevance to previous model calculations of atmospheric structure and consequent rates of mass
loading at 9.4 RJ . The atomospheric model calculations by Saur et al. (17) is compared to
the present results in Table 5. The Saur et al. (17) model is based on an assumed atmosphere
dominated by O2 . Smyth and Marconi (18) provide a model based on proton surface sputtering,
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
which removes molecules into the vacuum by the inefficient process of nuclear momentum
transfer, whereas multiply charged ions sputter by the exchange and dissociative recombination
process(22). The physical chemistry applied by (18) produces an atmosphere dominated by O2
, with H2 the second most abundant species. Species limits from the present work are compared
to previous model results in Table S1. The limit on H2 abundance established here is ∼100 times
below the (18) model calculations in the plasma sheet at Europa (Table S1). For future reference
Fig. S6 shows the upper limits on H2 brightness in the EUV/FUV calculated for the volumetric
conditions in the plasma sheet and the Europa atmosphere. The H2 model calculations include
rotational levels extending to J = 25 (26). The calculations are carried out for the combined
forcing of solar flux and electron impact, in statistical (non-LTE) equilibrium. Differences in
the spectra are caused by electron density and temperature, and the interaction of solar flux with
significantly different ground state vibration-rotation populations. Outward diffusion of atomic
oxygen from the Io Torus does not reach the orbit of Europa in measureable quantities (27). The
excitation of neutrals by the inwardly diffusing hot magnetosphere population (28) is negligible,
although ambient electron heating by this flux is significant. Another investigation (29) has
reported finding evidence of significant neutral populations near Europa in perturbations of pitch
angle distributions of magnetosphere ions. The production of a surface sputtered atmosphere at
Europa has been discussed by (30, 31, 32, 33).
Table 1: Limits to neutral species abundance (× 1012 cm−2 )
species
[O I]ℓ
[H I]ℓ
[H2 ]ℓ
[O2 ]ℓ
[H2 O]ℓ
a
b
Present
Plasma Europa
≤0.23
4.7
≤0.18
≤ 1.
≤ 4.5
···
···
···
···
Smyth (18)
Plasma Europa
5.
1.2
···
0.076
100.a
77.
···
450.
···
2.3
Saur (17)
Europa
···
···
···
500.
···
Ip (30)
Europa
···
···
···
200. − 2000.
···
b
Roth (1)
Europa
1.
···
···
350.
17.
los through Europa in orbital plane
Produced by heterogeneous magnetosphere surface sputtering, source rate 3 × 1026 O2 s−1
Critical plasma sheet physical chemistry
e + O2+
e + O3+
O3+
O3+ + N∗
469
470
471
→
→
→
→
O3+ + 2e
O2+
dif f usive loss
O2+ + N+∗
(11)
(12)
(13)
(14)
where N∗ refers to a neutral atom, atomic ion, or neutral molecule.
Energy budget: Table 1 notes. The radiative cooling rates calculated here consitute most of
the energy required to maintain the plasma (11, 12). The calculated energy deposition rates
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
shown in the Table occur through Coulomb interaction of the heterogeneous inwardly diffusing
hot plasma with the ambient electrons. This process dominates the energy input required to
maintain the resident plasma (11, 12). The results showing the state of the plasma at Voyager
1 encounter are model calculations fitting the Voyager UVS spectrum. For this reason neutral
atom densities are model projections and not measured in the data. the ion densities and electron
temperatures are derived from the spectral data; the electron density and temperature from this
analysis agree with the insitu results from the Voyager PLS experiment (10). The inbound
electron temperature and densities along the S/C trajectory extracted from (10) but converted to
an RJ scale from L-shell are given here in Fig. S7. The important properties of Fig. S7 are the
agreement with the UVS results in Table 1 at the Io torus, and the distribution of the ambient
electron temperatures and densities between 5.9 and 9.4 RJ . From the Io torus outward, the
electron temperature rises and the electron density declines monotonically to 7.4 RJ , where
the temperature rises and the density decreases sharply to 8.4 RJ . Following a region of lower
temperature to 8.8 RJ , the temperature rises sharply again coincident with a decrease in density
exptending to 10.RJ . These characteristics are interpreted here a decreases in mass loading
moving outward, allowing the heterogeneous plasma environment to heat the ambient plasma
diffusing outward from the Io torus. Comparing the state of the Io torus plasma from the spectral
data at Voyager encounter in 1979 (col 2) and the result from Cassini UVIS in 2001 (col 3) it
is evident that this region is subject to significant variability. The Io torus in 1979 was colder
and more dense, than in 2001. Radiative cooling in 1979 was lower than in 2001 because of
lower temperature. Col 4 of Table 1 shows the analysis of the Voyager PLS ion partitioning
(9) measurement. These results are less reliable because the PLS experiment measured the
mass per charge spectrum, so that O II and S III, for example are coincident. Nevertheless the
Voyager results in total mass ratio comparing 5.9 and 9.4 RJ , are similar to the Cassini UVIS
results in 2001. Col 5 of Table 1 shows the Cassini UVIS result at 9.4 RJ , where compared
to 5.9 RJ the temperature is very much higher, but because of low density, raditaive cooling
is reduced by a factor of 54. The last row of Table 1 shows the calculated energy deposition
into the ambient electrons by Coulomb energy transfer (11, 12). The heterogeneous plasma
quantities for the Voyager results at the Io torus (col 2, Table 1) are from the PLS analysis of
(10). The calculations for the Io torus in 2001 (col 3, Table 1), however, require that the hot
plasma density be increased by a factor of 2.4 in order to approximate the energy deposition
rate required to maintain the plasma. For the Cassini UVIS result at 9.4 RJ , the deposition
rate calculated using the heterogeneous plasma density from Voyager (10) provides sufficient
energy to maintain the observed ambient plasma radiative loss.
2000
1800
1600
counts
1400
LISM HLyβ
β
LISM He 584 A
1200
1000
800
600
400
200
0
550
650
750
850
950
1050
1150
λ (A)
Figure 1: UVIS EUV spectrum of Fig. 1 (brown trace) compared to UVIS spectrum of the Io
plasma torus eastern ansa (red trace). The spectra are normalized at the O II and O III multiplets
near 833 Å. The LISM neutral lines appear in the Europa spectrum and not in the Io torus spectrum because of relative brightness. The electron temperature difference between the plasma
volumes affect the emission line intensity distributions.
1.0
counts 1000 s pxw pxs
1
-
-1
-1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
550
650
750
850
950
1050
1150
λ(A)
Figure 2: UVIS EUV spectrum of Fig. 1 compared to a plasma model calculation including the
species O II, O III, O IV, S II, S II, S IV, S V, for an electron temperature Te = 250000 K. The
reduced data is the red trace, the total model is the light green trace, the O IV component is
shown in light red, and the LISM emission model is in light blue.
1600
1400
counts
1200
1000
800
600
400
200
0
550
650
750
850
λ (A)
950
1050
1150
Figure 3: UVIS spectrum of the Io plasma torus eastern ansa (blue trace) compared to a plasma
model calculation including the species O II, O III, S II, S III, S IV, S V, for an electron temperature Te = 75000 K. The total model is the red trace, the S IV component is shown in light green.
100
OI
HI
O II
O2
O2+
90
80
h (km)
70
60
50
40
30
20
10
0
10-1
100
101
102
103
104
105
106
107
[N] (cm-3)
Figure 4: Europa model atmosphere based on the Liang et al. (19 ) model for Callisto, in which
the CO2 source is suppressed to reflect conditions at the surface of Europa. The oxygen in this
model is derived from H2 O extracted from the surface cold, with recycling, leaving a cold O I
atmosphere. The analysis of observations to date would not be in conflict with the model. The
legend identifies the species in the plot, O I, H I, O II, O2 , and O+
2.
10-7
4
2
10-8
4
2
kij (cm s )
110-9
4
3
2
10
-10
10-11
ν
ν
4
2
10-12
4
2
10-13
→
→
→
e + O2 2O
e + O2 O(3P) + O(3s 5So)
e + O2 O(3P) + O(3s 3So)
e+h +OI
O(3s 5So)
e+h +OI
O(3s 3So)
e + H2O HLy
4
2
104
2
3 4 5 6 7 105
2
→
3 4 5 6 7 106
→
→
α
2
3 4 5 6 7 107
Te (K)
Figure 5: Rate coefficients for reactions rr(1) thru rr(4), rr(7) and rr(8) (see text). The forcing
in rr(7) and rr(8) includes a high resolution solar flux model (15). The latter two reactions have
electron density dependence; these calculations are for [e] = 104 cm−3 , where [e] refers to the
ambient electron population. A hot heterogeneous electron population, [eh] = 2. cm−3 (Teh =
107 K) is included in rr(7) and rr(8), and rates are for optically thin gas. The solar flux and [eh]
acquire greater influence at low ambient electron temperatures.
1.2
I (10 R A )
1.0
Europa H2 plasma sheet limit
Europa H2 atmosphere limit X 10-2
1
-
2
-
0.8
0.6
0.4
0.2
I (10 R A )
0.0
800
1
-
2
-
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1200
900
1000
1100
1200
Europa H2 plasma sheet limit
Europa H2 atmosphere limit X 10-2
1300
1400
1500
1600
1700
λ (A)
Figure 6: Model spectra of electron and solar forced H2 emission in the EUV/FUV region,
using parameters for the Europa atmosphere and plasma sheet. The calculations are made for
a nominal 2 Å FWHM gaussion psf, giving absolute scale upper limits in differential brightness (R/Å). The model includes rotational levels limited only by the dissociation limit in most
electronic state potentials used in the calculation (26). Details of the solar spectrum used in the
calculation are provided by (15).
3
2
Te (K)
106
7
5
4
3
2
105
7
5
4
3
2
104
4
3
2
[e] (cm-3)
103
6
4
3
2
102
6
4
3
2
101
4
5
6
7
8
9
10
11
12
13
RJ
Figure 7: The ambient electron temperatures and densities measured on the inbound leg of
the Voyager 1 spacecraft in 1979 at Jupiter encounter (from 10) transformed to an RJ scale.
These results show a general compatibility with the spectral measurements of the Voyager UVS
experiment, as well as with the results from the Cassini UVIS experiment in the 2001 encounter.