Steve Desch

1
Cryovolcanism on Charon and other Kuiper Belt
Objects
Steve Desch Jason Cook now at SwRI, Wendy
Hawley, Thomas Doggett School of Earth and Space
Exploration Arizona State University
2
Can KBOs experience cryovolcanism?

• A few words about cryovolcanism.
• A description of our model to calculate the
  thermal evolution of KBOs
• Results for Charon, including analysis of the
  physics
• Likelihood of subsurface liquid on other KBOs.
• Outline of a process for bringing liquid to the
  surface.

KBOs the size of Charon or larger can retain
subsurface liquid to the present day, and may
even be experiencing cryovolcanism, provided they
formed with moderate amounts of ammonia.
3
Crystalline Water Ice Cryovolcanism?
Crystalline water ice observed on many large
KBOs Crystalline water ice is expected to be
amorphized by cosmic rays doses of 2-3
eV/molecule (Strazzulla et al. 1992 Mastrapa
Brown 2006), which takes lt 3 Myr in Kuiper Belt
(Cooper et al. 2003). Once amorphized, KBO
surfaces stay amorphous because of low
temperatures. Cook et al. (2007) reviewed
annealing mechanisms. Most favorable was
micrometeorite impacts, but all of them were
found unable to compete with cosmic-ray
amorphization.
4
Crystalline Water Ice Cryovolcanism?

• Cook et al. (2007) intepreted crystalline water
  ice as diagnostic of cryovolcanism on KBOs. This
  would be incorrect IF
• Dust fluxes were gt an order of magnitude larger
  than interplanetary dust flux, as is possible in
  planetary environments. (2003 EL61 collisional
  family, too?)
• Real ices dont conform to experiments of
  amorphization

5
Cryovolcanism?
Still, cryovolcanism does exist. Ariels surface
lt 100 Myr old (Plescia 1989), Tritons even
younger (Schenk Moore 2007)
Are these objects tidally heated, or are young
surfaces common on KBOs, too??
6
Cryovolcanism needs ammonia
X NH3 / (H2ONH3). Maximum cosmochemical value
is X 15 (Lodders 2003). Models of
molecular cloud chemistry predict N2 is
efficiently dissociated, converted into NH3
(Charnley Rodgers 2002). Depletion of N2
recently confirmed observationally (Maret et al.
2007). Models predict 25 of all N in NH3
ices, for X 5 Observations of 9.3 micron band
of ammonia ice suggest X 5 - 10 (Gibb
et al. 2001, Gurtler et al. 2002), but are
disputed (Taban et al. 2003). Comets show X lt
1.5, but may be devolatilized. Ammonia content
of KBOs is unknown, but X 5 is not unreasonable
7
Description of Model
Model updates internal energy in zone i
Qi(t) rate of heating by long-lived
radionuclides Fluxes into zone i (Fi-1) and out
of zone i (Fi) found assuming thermal conduction
Equation of state is used to convert E back
into temperature
8

• Ammonia
• We use simplified phase diagram to include
  following phases
• Solid water ice
• Solid ammonia dihydrate (ADH)
• Liquid water
• Liquid ammonia
• Rock (analogs being ordinary chondrites)

9
Ammonia
10
Ammonia Energy added to each zone goes into
heating components via heat capacity, or into
latent heats due to phase transitions. Each
shell with mass M has energy E at the end of each
timestep. We then find temperature T and fraction
of mass in each (non-rock) phase that is
consistent with this E
k refers to regime in phase diagram
11
Ammonia For example, in regime 1 (Tlt 176-dT K),
Similar (but much more complicated) expressions
apply to other regimes
12
Ammonia For example, in regime 3 (176dT lt T lt
Tliq),
13
Ammonia
Hunten et al (1984)
Just a few ammonia drastically lowers the
viscosity, especially once ADH melts.
Limit for meter-sized rocks to slip 10 km/Myr
Arakawa Maeno (1994)
14
Differentiation
If the ice contains a few ammonia,
differentiation can occur wherever T gt 176
K Maximum radius at which T176 K ever
Rdiff Within Rdiff, we separate into rocky
core, then ADH ammoniawater slush layer,
then water ice on top. Undifferentiated rock-ice
crust lies outside Rdiff. ADH denser than its
melt, so slush layer well mixed we mix
compositions and internal energies after each
timestep (this mimics convection).
15
Radiogenic Heating We consider heating by
long-lived radionuclides 235U, 238U, 232Th and
40K only.
Avg heating during first 1 Gyr 5 x Avg
heating during last 1 Gyr!
16
Thermal conductivities Rock We use values
measured for ordinary chondrites at low
temperatures (100 - 500 K) by Yomogida Matsui
(1983) k 1.0 W/m/K, independent of
temperature Water Ice k 567 / T W/m/K (Klinger
1980) Ammonia Dihydrate (ADH)k 1.2 W/m/K
(based on Lorenz Shandera 2001) Water /
Ammonia Liquids assumed to be convecting k set
to high value k 40 W/m/K ConvectionWe check
for convection in water ice layer, but Ra ltlt 1000
in all models we ran no convection.
17
Thermal conductivities Conductivities of non-rock
components combined using geometric mean, using
volume fractions Conductivities of rock and ice
components combined using percolation theory
formula of Sirono Yamamoto (2001)
Conductivity of undifferentiated rock-ice mixture
on Charon well described by k(T) 3.21 (T/100
K)-0.73 W/m/K
18
Thermal conductivities
19

• Results
• Canonical case, a Charon-like body
• R 600 km
• ? 1.7 g cm-3 (rock fraction 63)
• X 5
• Differentiation starts at t65 Myr, reaches
  fullest extent by 100 Myr
• Rdiff 474 km... half the mass differentiates

20
t1.74 Gyr
t2 Gyr
slush layer
t1 Gyr
t3 Gyr
water ice layer
t4 Gyr
t4.6 Gyr
icerock crust
rocky core
t0 Gyr
21
water ice layer
rocky core
slush layer
icerock crust
H2O(s)
rock
rock
H2O(s) ADH
H2O(l) NH3(l)
H2O(s)
ADH
22
All ammonia within Rdiff leads to liquid. No
additional liquid is created without ammonia
antifreeze.
Temperatures in slush layer drop below 176 K
freezing starts at t 4.3 Gyr
Differentiation takes place within 100 Myr
23
Present-day steady-state radiogenic heat flux at
surface would be F 1.28 erg cm-2 s-1.
Analytical estimate of temperature at base of
ice shell would be T 100 (0.993)3.704
exp(0.287) 129 K. Flux is enhanced over
steady-state radiogenic heat flux by amount ? F
by release of heat from rocky core. Temperatures
in ice shell and in undifferentiated crust
explained to within 1 by model with ?0.28.
Temperature at base of ice layer predicted to be
T 100 (0.9930.172?)3.704 exp(0.2870.581?)
182 K. Release of heat from core predicted to
enhance flux by amount ?0.33 Release of stored
heat from core is significant!
24
Release of latent heat is also significant! Freezi
ng commences at t4.3 Gyr Mass of water/ammonia
liquid that freezes 4 x 1022 g Latent heat
released during freezing 5 x 1033 erg Release
of this latent heat would enhance surface flux by
a whopping 0.4 erg cm-2 s-1 if released in just
0.1 Gyr. Release of latent heat buffers
freezing, prolongs it to take gt 0.6 Gyr
Doubling ammonia (X10) creates more liquid,
and also prolongs it to take 1.5 Gyr!
25
(No Transcript)
26

• Our model is highly favorable to maintenance of
  subsurface liquid
• Undifferentiated crust containing half the rock
  (as well as ADH) is thermally insulating
  (compared to pure water ice).
• Core containing the other half of the rock---and
  its radionuclides---concentrates and stores heat
• Release of stored heat and latent heat of
  freezing is significant, and demands a time
  evolution model.
• These physical effects would not be captured in a
  steady-state, fully differentiated model.

27
chondrite melting point
P gt 200 MPa
Bigger is better... but beware
R 800 km
28
Recipe for present-day liquid X gt 5 M gt 1024 g,
? gt 1.3 g cm-3 (R gt 500 km, f_rock gt 40)
29
(No Transcript)
30
How does subsurface liquid surface?
Crawford Stevenson (1988) use linear elastic
fracture mechanics to show that the stress
intensity at the tip of a fluid-filled crack of
length l, extending from base of ice layer (top
of subsrface ocean), is
If this exceeds Kc 6 x 108 dyne cm-3/2, the
crack will self-propagate.
31
How does subsurface liquid surface?
On Europa, ?? 1.00 g cm-3 - 0.92 g cm-3 gt 0,
and tension T is needed to initiate a crack. The
crack has a maximum possible length. In our
models, ?? 0.88 g cm-3 - 1.71 g cm-3 lt 0, and
buoyancy can drive the crack all the way to the
surface. Cracks will propagate at several m/s
(Crawford Stevenson 1988), reaching the surface
in 1 day.
32
How does subsurface liquid surface?
Cracks as small as 0.8 km can become
self-propagating within Charons ice
layer. Cracks are likely to be initiated during
freezing of slush layer, when its volume must
increase by 7. Displacement of 7 of ADH over
0.6 Gyr would coat Charons surface with
water-ammonia ices to depth 1 m / Myr 1 mm /
kyr ( 0.6 km total). Heat flux carried to
surface only 0.004 erg cm-2 s-1, too small to
affect thermal evolution.
33
cracks form here
Conclusions
Basic structure of KBOs 400-800 km in radius
thermally insulating, undifferentiated rock-ice
crust
pure water ice layer, does not convect
ADH - ammonia - water layer buffered near 176 K
hot rocky core
34
Conclusions

• Our models include time evolution, ammonia and
  differentiation. These are significant factors
  for thermal evolution of KBOs, and their effects
  are favorable for maintaining subsurface liquid.
• Rule-of-thumb for subsurface liquid today
• M gt 1024 g, ? gt 1.3 g cm-3, X gt 5
• Charon and Orcus likely to have subsurface
  liquid.
• Liquid could be brought to surface via cracks,
  especially as bodies freeze (which is now for
  Charon)
• Obvious astrobiological implications can
  bacteria live in water thats 32 ammonia, and
  near -100ºC ??