1. Cassini UVIS Results on the Enceladus Plume and Spacecraft Safety
Larry W. Esposito
3 May 2007
Enceladus SDT

2. Cassini clipped edge of plume: INMS, CDA in situ Results
~1 minute before closest approach the Cosmic Dust Analyzer detected a peak in the number of small particles (blue diamonds), 460 km altitude
35 seconds before closest approach the Ion Neutral Mass Spectrometer measured a large peak in water vapor (yellow diamonds), 270 km altitude
Gas and dust plumes are decoupled at these altitudes
Cassini actually flew through the outer edge of Enceladus’ plume
Red line is closest approach to hot spot
Green line is c/a to Enceladus
Coma observed by INMS out to 4000 km. Densities in this extended region may be indicative of flux that varies on time scales of < 1 hr
Hill sphere radius = 948 km

3. CDA Peak
INMS Peak
Outline of hot spot

4. Enceladus Stellar Occultation Geometries
February Lambda Scorpii Occultation
July Gamma Orionis Occultation
Positive detection of plume on ingress!
Tick marks every 5 sec
Egress
Ingress

5. Localization of Enceladus’ Plume (Not a global atmosphere)
Ray intercepts were at latitude / west longitude:
15 / 300 Lambda Sco ingress (non-detection)
-31 / 141 Lambda Sco egress (non-detection)
-76 / 86 Gamma Ori ingress
-0.2 / 28 Gamma Ori egress (non-detection)
Consistent with localized plume or jet:
Enceladus’ gravity insufficient to hold gravitationally bound sputtered atmosphere
Also, the combination of other Cassini data sets are consistent with a plume of water vapor coming from Enceladus’ “Tiger Stripes” driven by the hot spot at the south pole detected by CIRS

6. FUV Data: Comparison of Occulted to Unocculted Spectra
FUV configuration:
spectral channels binned by 2 (512 spectral channels from nm to 191 nm)
low resolution slit width
5 sec integration time
full spatial resolution
Time record 33, the last full 5 sec integration prior to ingress, shows the deepest absorption. The ray altitude above Enceladus’ surface corresponding to time record 33 ranged from 30 to 7 km.
Clear signature of an atmosphere is present – both relatively narrow and broad absorption features

7. Composition of Plume is Water Vapour
I=I0 exp (-n*)
I0 computed from 25 unocculted samples
n = column density
 = absorption cross-section, function of wavelength
Water spectrum from Chan et al, Data taken at 298 K.
The absorption spectrum of water (pink line) is shown compared to Enceladus’ plume spectrum (I/I0) for a column density of n = 1.5 x 1016 cm-2

8. Structure of the Plume
The increase in water abundance is best fit by an exponential curve – a comet-like evaporating atmosphere (1/R2) does not fit the data well, nor do global hydrostatic cases
The best fit scale length is 80 km

9. Neutral Species in Saturn’s System
The Saturnian system is filled with the products of water molecules:
H detected by Voyager in 1980, 1981
OH detected by Hubble Space Telescope in 1992
Atomic Oxygen imaged by UVIS in 2004
The water budget

10. “Search for the Missing Water Source”1
Neutral Species
Water and its products are lost from the system by collisions, photo- and electron- dissociation and ionization
Estimates of required re-supply rates, water molecules/sec:
2.8 x Shemansky, et al.
3.75 x Jurac, et al.
Jurac and Richardson
2 x Shemansky, et al.
E Ring
Saturn’s E ring, composed primarily of 1 micron particles, is also subject to erosion and loss due to sputtering of water from the surface of the E ring’s dust particles and collisions of particles with Saturn’s moons
Estimate of required re-supply rate:
1 kg / sec Juhasz and Horanyi
E ring particles are 0.3 to 3 microns
Lifetime of 1 micron grains is ~50 yrs

11. Estimation of Water Flux from Enceladus
S = flux
= N * h2 * v
= n/h * h2 * v
= n * h * v
Where
N = number density / cm3
h2 = area
v = velocity
n = column density measured by UVIS
The biggest uncertainty is what to use for h
Estimate h from plume dimension, = 80 (from scale length) or 175 km (from horizontal distance traversed)
Estimate v from thermal velocity of water molecules in vapor pressure equilibrium with warm ice (41,200 at 145 K or 46,000 cm/sec at 180 K – note that escape velocity = 23,000 cm/sec) h v
S = 1.5 x 1016 * (80 or 175) x 105 * (41 or 46) x 103 = 0.5 to 1.2 x H2O molecules / sec
= 150 to 360 kg / sec

12. New Plume Model
A new model has been developed for Enceladus’ plumes by Tian, Toon, Larsen, Stewart and Esposito, paper to appear in Icarus
Monte Carlo simulation of test particles given vertical + thermal velocity, particle trajectories tracked under influence of gravity and collisions
Assumes source of multiple plumes added together along each tiger stripe
UVIS ray path across tiger stripes

13. Monte Carlo model results - Predicted Plume Shape
Simulations by Teddy Tian. Paper submitted to Icarus
10,000 test particles given a vertical velocity and a thermal velocity. Thermal velocity direction is chosen randomly and speed satisfies a Boltzmann distribution for Tthermal. Tthermal of 140 to 180K not much affect on results (these temps from Spencer). If T=273 then crack must be very narrow.
Inferred surface density is 1010 to 1011 cm-3
Trajectories of particles tracked under the influence of gravity and collisions. Mean free path is calculated, increases with altitude. Number density near surface is high enough for collisions. Model which takes into account viewing geometry of tiger stripes suggest 10 to 100 % of stripes are venting
Level 1.0 => 1017 cm2; Level 0.1 => 1016 cm2

14. Monte Carlo Model - Fit to Data
Zero Vz => sublimation -> column density declines more rapidly than observed
Mass deposition is 2 orders of magnitude < escape rate, still adequate to resurface and account for high albedo
Density distribution in plume as f(z): n(z) = nsurf * (1 + (z/(1+Vz/Vth))-2
Number density of plume consistent w/ column density = 1.5 x 1016 cm-2 is cm-3, depending on assumed line of sight altitude range. If vertical velocity is same order of magnitude as thermal velocity the surface density is cm-3
Best fit to UVIS column density as a function of altitude requires a vertical velocity of 300 to 500 m/sec
Water flux is x 1027 molecules/sec = kg/sec
(consistent with initial estimate)

15. Detecting Temporal Variability
The water budget derived from the water vapor abundance shows Enceladus supplies most if not all of the OH detected by HST, atomic oxygen in the Saturn system detected by UVIS
Implies activity for > 15 years, since HST observed OH in 92
Since the oxygen in the system comes from Enceladus UVIS may be able to remotely monitor Enceladus’ activity levels by monitoring the system oxygen level

16. Changes in System Oxygen Content
Units are Rayleighs

17. Weekly O1304 Trend

18. Enceladus Summary
UVIS measures water source large enough to create neutral oxygen cloud and to re-supply E ring
UVIS column density equal to about a single 1mm ice grain per square meter
A uniform source could loft ice particles of radius about 1 micron; higher density jets could loft particles dangerous to Cassini

19. Plume physical explanations
Models
Fumarole model. Misty vapor cools as it expands; ice particles condense. T ~ 170K.
Geyser model. Local heating gives boiling water at depth, vent geometry gives vertical velocity, collimation; bubbles form and liquid freezes, effectively lofting larger particles to high speeds. T ~ 270K.
Comet model. Sublimating vapor lifts ice grains from vent interface and carries them away. T ~ 200K.

20. Comparable mass
In all these models, there is a close coupling between the ice and vapor
Growth, lofting and/or evaporation involve an interchange between water molecules and solid ice particles
For any significant interchange of mass or momentum, the column of water vapor incident on an ice grain’s surface area must have a comparable mass to the grain mass

21. Mass Balance N0 *  * a2 * H * mH20 =  * 4/3 *  * a3
For H ~ 40km,  ~ 1, we solve for a (in microns) a ~ N0/ (1012 cm-3)
Thus, high pressure vents could loft or grow big particles, potentially dangerous to Cassini

22. Number of expected collisions
Ncolumn/1.5 x 1016 f1 a0-3 A
where a0, the particle radius, is measured in mm; and A, the Cassini sensitive area, is measured in m2. Ncolumn is the column density of gas along Cassini’s path through the plume. For E3, this would give about
10-3 f 1 hits (from INMS results). consistent with our safe passage.

23. Observational constraints
The shape of the observed plumes shows V0 > Vth
Tian etal can match the UVIS results with V0 ~ 400 m/s and N0 ~ 1010 – 1012 cm-3
This gives typical grain sizes a ~ 0.01–1, roughly consistent with photometry and CDA measurements
Present analysis of Cassini data would give no indication of dangerous particles, by orders of magnitude

24. Hazard calculation: Approach and assumptions
Plume has cylindrical symmetry about pole
Estimate plume density integrated along Cassini path from water column measured by UVIS star occultation
See Spencer and Hansen figures: UVIS had a measurement at location of highest density for rev 61

26. Rev 61 c/a >

27. Calculation
If all water vapor along this line of sight to star (Ncol = 2E15/cm2) were swept up by Cassini’s sensitive area (0.8 m2), this would form a solid ice sphere of radius 500 microns
Assume measured solid particle size distribution can be extended as a power law in radius to sizes dangerous to Cassini
CDA: q = 4
RPWS: q = 6.4 (radius power law)

28. Number of dangerous particles
Calculate the predicted number of hits by dangerous particles (r > 900 microns, Dave Seal) if Cassini flew a path with same minimum altitude:
ND = f1* (4-q)/(q-1) *
a03/(amax4-q - amin4-q) *
(a*1-q - amax1-q)

29. Key parameters a*: dangerous particle radius, 900 microns
a0: equivalent ice radius, 500 microns
amin, amax: size range, radius microns
f1: ratio of solid ice mass to water vapor
q: power law size index

30. Results
ND = 3E-9 f1 for q = 6.4
ND = 2E-3 f1 for q = 4

31. Values for f1, mass ratio solid/gas
Simple physical arguments of mass balance, force balance, growth of solids from vapor give f1 < 1
Comparing mass loss of solids estimated by ISS, CDA to vapor by UVIS gives f1 ~ 0.01

32. But, what about small, high pressure vents
But, what about small, high pressure vents? They could loft dangerous particles. Signal more variable within plume …
Outside
Within plume

33. Same number of high and low outliers

34. Entire Enceladus ingress cut (binned data)
3 features near limb found by ring feature search algorithm, but below significance threshold

35. Conclusions from 2 independent searches
Sensitive to events as small as 50m; opacity as small as 10%
We see no significant deviations from smooth variation
Outlier events have width less than 1km and opacity less than twice mean

36. Conclusions
Extrapolating Cassini plume measurements to rev 61 and to radius dangerous to Cassini, using the most optimistic size range, provides a conservative estimate of the number of hits expected
The value is 0.2% or less
No evidence for high pressure vents
Better measurements of the size distribution and its opacity would improve the model