Active formation of ‘chaos terrain’ over shallow …

 

 

LETTER

doi:10.1038/nature10608

Active formation of ‘chaos terrain’ over shallow
subsurface water on Europa

B. E. Schmidt1, D. D. Blankenship1, G. W. Patterson2 & P. M. Schenk3

Europa, the innermost icy satellite of Jupiter, has a tortured young
surface1–4 and sustains a liquid water ocean1–6 below an ice shell of
highly debated thickness1–5,7–10. Quasi-circular areas of ice disrup-
tion called chaos terrains are unique to Europa, and both their
formation and the ice-shell thickness depend on Europa’s thermal
state1–5,7–17. No model so far has been able to explain why features
such as Conamara Chaos stand above surrounding terrain and
contain matrix domes10,18. Melt-through of a thin (few-kilometre)
shell3,7,8 is thermodynamically improbable and cannot raise the
ice10,18. The buoyancy of material rising as either plumes of warm,
pure ice called diapirs1,9–15 or convective cells16,17 in a thick (.10
kilometres) shell is insufficient to produce the observed chaos
heights, and no single plume can create matrix domes10,18. Here
we report an analysis of archival data from Europa, guided by
processes observed within Earth’s subglacial volcanoes and ice
shelves. The data suggest that chaos terrains form above liquid
water lenses perched within the ice shell as shallow as 3 kilometres.
Our results suggest that ice–water interactions and freeze-out give
rise to the diverse morphologies and topography of chaos terrains.
The sunken topography of Thera Macula indicates that Europa is
actively resurfacing over a lens comparable in volume to the Great
Lakes in North America.

Although the settings are different, terrestrial environments can
provide critical context for Europa, particularly where water and ice
interact under pressure. Melting of ice occurs at subglacial volcanic
craters on Earth, such as Iceland’s Grimsvotn19 (Supplementary Fig. 4).
Ice covers the volcano, and as it becomes active, the ice cap melts from
below, causing surface down draw19. Water in glacial systems flows
perpendicular to the gradient of the fluid potential (Supplementary
Information section 2). In pure ice, the surface slope above a water
body is roughly 11 times as important as the basal slope in determining
flow direction19,20, and water collects where the hydraulic gradient
approaches zero, at the centre of the depression. Above the activating
volcano, the surface slope is steepest at the flanks of the crater, creating
a hydraulic seal that prevents the escape of water and drives the forma-
tion of a lens-shaped subglacial lake19,20. Because Iceland’s ice caps are
finite in width, breakout flow along the bed or floatation of the ice cap
eventually allows water to escape19,20.

Water also influences Antarctic ice shelves. Brines enter terrestrial
ice shelves through basal fractures or the front of the shelf (via a porous
layer called firn) and percolate through the ice for tens of kilometres
over many years21. Beyond enhancing its water and impurity content,
introduction of brine can weaken the ice by reducing its shear
strength21–23. Hydrofracture occurs when tidal cracks fill with water,
causing force at the crack tip, which can initiate ice shelf collapse22,23
(Supplementary Information section 2), producing tabular icebergs
surrounded by a matrix of brine-rich (‘brash’) ice.

Coupled analysis of the geomorphology of Conamara Chaos and
Thera Macula, (Figs 1 and 2, respectively) demonstrates that the two
features share a quasi-circular shape and floating blocks, whereas their

topography differs. Conamara Chaos is raised and contains raised
matrix ‘domes’. Thera Macula, however, is sunken below the sur-
rounding surface. These observations, informed by the environments
on Earth described above, suggest a four-phase ‘lens-collapse model’
for chaos terrain formation (Fig. 3). (1) Ascending thermal plumes of
relatively pure ice13 cross the eutectic point of overlying impure ice,
producing surface deflection in response to volume change associated
with pressure melting of the ice (Fig. 3a). (2) Resulting hydraulic
gradients and driving forces produce a sealed, pressurized melt lens
(Fig. 3b). (3) Extension of the sinking brittle ice ‘lid’ over the lens
ultimately generates deep fractures from below, allowing brine to both
be injected into and percolate through overlying ice, forming a fluidized
granular ice matrix and calving ice blocks (Fig. 3c). (4) Refreezing of
the melt lens and now brine-rich matrix results in topographic
heterogeneity (Fig. 3d).

The upper crust of Europa is rich in impurities owing to either
exogenic implantation or endogenic injection (see, for example, refs
24, 25). Models suggest that melt will be formed as warm, composi-
tionally buoyant plumes cause the cold impurity-rich ice above them
to reach its eutectic pressure-melt point, driving partial melting and ice
disruption1,2,9–15,17. Interconnected pockets of thermally stable melt
water will form above a plume and be over-pressurized by an amount
equal to the buoyant force of the plume, 104–105 Pa (ref. 13). It is
significant that previous models did not take into account the volume
change associated with melting ice15 and its ramifications 19,20. Ice
melting results in surface draw down, which hydraulically seals the
melt in place 19,20 (Fig. 3b; Supplementary Information section 2),
rather than melt draining downward as previously speculated10,13–
15,17,26. That is, water formed from melting above plumes must move
perpendicular to local hydraulic gradients, both up the plume head and
towards the centre of the depression24, and will form a lens similar to
subglacial volcanic lakes. Ultimately, the volume of melt and hydrostatic
equilibrium with the overlying depression determine the shape of the
lens; this shape will be slightly modified by any lithostatic stresses at the
edge of the lens. On Europa, the lateral continuity of the ice shell pro-
hibits horizontal escape of the water.

Pressure and fracture will contribute to the response of Europa’s ice
shell to subsurface melting. Prevalent pre-existing faults and discon-
tinuities in ice strength should be regions of localized weakness, akin to
ice shelf fractures, accumulating much of the extensional strain from
the subsidence of the lid. As cracks propagate upward from the melt
lens, hydrofracture will break up the ice (Fig. 3c) wherever high-
pressure water inflow contributes force at the crack tip22,23. Fracture
will calve steep-sided blocks and allow water to enter overlying brittle
ice.

The morphology of Conamara Chaos requires the transformation of
mostly background plains material3,11,14 into an impurity-rich matrix
(Supplementary Figs 1, 6). Background plains material is characterized
by high fracture density, and thus may be more susceptible to brine
inflow and disruption as observed11,14; the shallow subsurface may also

1Institute for Geophysics, John A. & Katherine G. Jackson School of Geosciences, The University of Texas at Austin, J. J. Pickle Research Campus, Building 196 (ROC), 10100 Burnet Road (R2200), Austin,
Texas 78758-4445, USA. 2Applied Physics Laboratory, Johns Hopkins University, 11100 Johns Hopkins Road, Laurel, Maryland 20723, USA. 3Lunar and Planetary Institute, 3600 Bay Area Boulevard,
Houston, Texas 77058, USA.

5 0 2 | N A T U R E | V O L 4 7 9 | 2 4 N O V E M B E R 2 0 1 1

©2011

Macmillan Publishers Limited. All rights reserved

a

b

800 m

N

400 m

0 m

–400 m

–800 m

c

W

N

S

N

S

N

W

W

W

E

E

E

S

20 km

LETTER RESEARCH

S

600

500

400

300

200

100

40:1
40:1

E

600

500

400

300

200

100

40:1

20

)

m

(

n
o
i
t
a
v
e
e

t
e
s
f
f

l

O

)

m

(

n
o
i
t
a
v
e
e

t
e
s
f
f

l

O

Figure 1 | Conamara Chaos is dominated by long-wavelength topography.
The region’s (8u N, 274u W) topography was analysed by filtering a DEM
(Digital Elevation Map) produced through a combined stereo
photogrammetric and photoclinometric method (Supplementary Information
section 1). The DEM has a spatial resolution of 180 m per pixel, and height
accuracy of 20 m. Colour indicates topographic heights relative to the
background terrain. a, Conamara Chaos. The area boxed in white (bottom left)
indicates the locations of the top, middle and bottom insets shown right. Insets
show, top to bottom, the short-wavelength (,20 km), absolute, and long-
wavelength (.20 km) topographic signals produced by the DEM filtering.
b, c, North–south (b) and east–west (c) topographic profiles of a typical dome-
like matrix swell. Solid lines, the DEM; dashed and dashed-dot lines, the long-
and short-wavelength topography, respectively. The profiles are offset in
elevation for clarity and vertical exaggeration is 40:1. Overall, the topography is
characterized by highs within the disaggregated ice matrix and lows at large

be porous. This weak ice can then be disrupted by transient pressure
from below, by diurnal tides, by wedging in freezing cracks or by
interactions with translating blocks, eventually breaking down into
the observed impurity-rich matrix1,3,7,18. This mechanism will convert
plains material into matrix without requiring thermodynamically
unlikely surface melt (see, for example, refs 10, 18).

N

C

SC

800 m

0 m

–800 m

NS

A

B

80 km

–20

–10

0

10

Distance (km)

polygonal block-like icebergs. Conamara Chaos appears to have completely
disrupted the ice fringed by its boundary scarp, and to have thickened relative to
the background terrain. Matrix ‘domes’ reach heights typically of 200 m. These
domes can entrain or tilt some smaller blocks. Large blocks represent the lowest
points within the region, with heights equivalent to or up to 100 m below the
background elevation. On average, the entire region is raised by ,100 m.
Comparing insets in a, the short-wavelength topography has little spatial
variation, while the spatial patterns of the DEM and long-wavelength
topography are similar, demonstrating that the long-wavelength signal
contributes strongly to Conamara Chaos’s heterogeneous topography. That the
highest topography is (1) rounded in shape and (2) located within or
‘controlled’ by matrix swells, indicates that water injection within the matrix
and subsequent freezing is responsible for topographic heterogeneities within
mature chaos terrain.

Conamara Chaos is not only elevated above the background terrain,
but also contains matrix domes that are in some places higher than
blocks9,18 (Fig. 1). This observation was initially interpreted as resulting
from the coalescence of multiple warm ice diapirs9. However,
Conamara Chaos’s continuity and nearly unbroken margin is more
consistent with a single source of disruption. The lens-collapse model
implies that brines are injected into former plains material as the ice
fractures (Supplementary Fig. 6), while increased water pressure from
the lens freezing raises the saturation level within the permeable matrix

Figure 2 | Thera Macula is a region of likely active chaos production above a
large liquid water lens. The image of There Macula was produced using
photoclinometry (Supplementary Information section 1) of Galileo images
with illumination from the north (N) and 220 m per pixel resolution. Colour
indicates topographic heights relative to background terrain. The region (50u S,
180u W) exhibits Conamara-like icebergs and dark matrix. The centre of Thera
Macula is sunken below the background terrain (denoted here by pale green) by
up to ,800 m just outside a large semi-circular northern subsiding province
(NS), and shows evidence for thickening of matrix in its southern chaotic
terrain (SC), which is elevated above the background terrain by up to ,800 m.
Despite the distinct difference between the morphologies of the NS and the SC,
they share a nearly continuous circular scarp boundary (red and black arrows),
suggesting that they formed from a common subsurface disruption. Blocks A, B
and C are calving at reactivated pre-existing fractures. Within the NS, the platy-
blocks appear bent, presumably by basal thinning and subsidence, and the edge
of the NS appears ready to break or is being thinned (red arrow). Tall scarps
(black arrows) either cast shadows (southern facing) within the region’s interior
or have bright faces (northern facing), demonstrating the relatively low-lying
nature of the centre of Thera Macula. To the lower right, an older band or ridge
complex interrupted by the disruption of Thera Macula appears to be swelling
along its ridge-like lineations, consistent with brine infiltration and refreeze
(blue arrow). The still sunken topography of Thera Macula is indicative of
subsurface water.

2 4 N O V E M B E R 2 0 1 1 | V O L 4 7 9 | N A T U R E | 5 0 3

©2011

Macmillan Publishers Limited. All rights reserved

RESEARCH LETTER

a

Upwarping surface?

b

Deflecting surface

Hydraulic flow

Fractures

Forming melt lens

Flattening plume

d

Scarp margin

Dome margin

Brine zone

Refrozen brine

Refrozen lens

Eutectic surface

Rising thermal plume

c

~3 km

Floating blocks

Diminishing plume

Figure 3 | A new hypothesis for chaos formation. a, An ascending thermal
plume in the subsurface approaches the pressure-melting eutectic point of the
overlying impure brittle ice. b, Melting causes surface subsidence that
hydraulically confines water, and produces tensile cracks. This behaviour is
governed by the relationship19,20

+wb

~ rw
ð

{ri

Þg+zpzgri+zs

ð1Þ

where wb is the fluid potential at the lens base (initially the plume head), ri and
rw are the densities of ice and water, and zp and zs are the elevations of the base
and the surface relative to the geoid, respectively. The slope factor, ri/(rw 2 ri),
which is dependent on the impurity content of the ice and water, determines the
relative importance of the surface and basal slopes in controlling the direction
of water flow. Equation (1) requires that the melt form a lens-shaped pocket
with finite effective pressure due to the plume below and overburden from the
ice above, unless the slope over the lens is ri/(rw 2 ri) times steeper (and in the
opposite sense) than the surface slope within the depression. Ultimately the lens
top must reach hydraulic equilibrium between the melt volume and the melt-
induced depression overlying it. c, Hydrofracture from the melt lens calves ice
blocks, while fracture and brine infiltration form a granular matrix. d, Refreezing
of the melt lens and freezing of now brine-rich matrix raises the chaos feature
above surrounding terrain, and can cause domes to form between blocks and
at margins. Note also that the topography at chaos margins can depend both
on the ice type at the margin as well as the direction of the ice collapse.
Where deep fractures of blocks define the margin, steep scarps are expected,
whereas boundaries defined by matrix (or shallow fractures) will warp into a
dome.

(Fig. 3d). This now brash ice eventually swells in response to thermal
expansion of freezing brines that fill its once empty pores and cracks.
The thickening of brine-infiltrated ice will raise chaos terrains above
undisrupted blocks of background terrain and form matrix domes, as
confirmed by our analysis of the detailed topography of Conamara
Chaos (Fig. 1; Supplementary Information section 1).

Whereas block morphology and motion3,11 chronicle chaos
kinematics, matrix preserves a record of its thermodynamics and
Europa’s ice rheology. The addition of a minimum of 940 km3 of liquid
water into the matrix is required to produce the average height of
Conamara Chaos (Supplementary Information section 3); this is
probably only a small fraction of the lens required to produce such a
large feature. Also, the stability of any floating ice block to toppling
within the matrix depends strongly on its shape27,28; rectangular blocks
are stable against toppling for aspect ratios (thickness to width) below
,1.4 (ref. 28). The smallest tilting blocks at Conamara Chaos thus give
us an estimate of the depth to liquid water when they formed. The
smallest upright but tilting (conditionally stable) blocks are ,2 km
wide11, and thus ,2.8 km thick, providing an estimate of the brittle

5 0 4 | N A T U R E | V O L 4 7 9 | 2 4 N O V E M B E R 2 0 1 1

layer’s thickness and the minimum depth of water lenses. Ice block and
matrix dome heights within Conamara Chaos (Fig. 1) are consistent
with 10–35% ice porosity for reasonable brine densities (Supplemen-
tary Information section 3).

Most importantly, the lens-collapse model developed here makes
testable predictions: whereas swelling matrix indicates lens freeze-out,
liquid water may be found where the surface subsides and blocks ‘float’
above the matrix. At Thera Macula, we are probably witnessing
active chaos formation (Fig. 2). The large concentric fracture system
encircling Thera Macula resembles those of collapsing ice cauldrons19,20
(Supplementary Fig. 4), and, given the absence of a continuous moat,
suggests that subsurface melt and ice disaggregation is forming Thera
Macula, rather than the collapse of a dome29. Topographic contrast
indicates that the rapid freeze of matrix is occurring at the region’s
south, but surface slopes, ice blocks and incomplete break-up to the
north indicates that the lens below Thera Macula was liquid at the time
of the Galileo encounter. Today, a melt lens of 20,000–60,000 km3 of
liquid water probably lies below Thera Macula; this equates to at least
the estimated combined volume of the Great Lakes (Supplementary
Information section 3). Although it is unclear how rapidly the break-
up of Thera Macula took place, such a volume would take ,105–
106 years to freeze. Surface modification should be extensive just after
the collapse and persist as long as the lens is mostly liquid, such that
Thera Macula may have noticeable changes between the Galileo
encounter and the present day. Such features can be well understood
by coupled topographic and subsurface imaging30.

The existence of globally distributed chaos terrain1,2 argues that
pervasive shallow subsurface water has existed and continues to exist
within Europa’s icy shell; surface draw down at Thera Macula suggests
that it exists within 3 km of the surface. The lens-collapse model pre-
sented here explains previously discrepant observations of chaos
(refs 10, 18; Supplementary Information section 4). Our work predicts
that it is the scale of ascending plumes and local surface geology that
produces diverse chaos morphologies, and thus our model may be
extended to other features, such as Murias Chaos, as well as pits and
domes. Our analyses suggest that ice–water dynamics are active
today on Europa, sustaining large liquid lakes perched in the shallow
subsurface.

Received 17 April; accepted 4 October 2011.

Published online 16 November 2011.

2.

4.

9.

1. Pappalardo, R. et al. Geological evidence for solid-state convection in Europa’s ice

shell. Nature 391, 365–368 (1998).
Figueredo, P. H. & Greeley, R. Resurfacing history of Europa from pole-to-pole
geological mapping. Icarus 167, 287–312 (2004).

3. Carr, M. H. et al. Evidence for a subsurface ocean on Europa. Nature 391, 363–365

(1998).
Squyres, S. W., Reynolds, R. T., Cassen, P. & Peale, S. J. Liquid water and active
resurfacing on Europa. Nature 301, 225–226 (1983).

5. Cassen, P., Reynolds, R. T. & Peale, S. J. Is there liquid water on Europa? Geophys.

Res. Lett. 6, 731–734 (1979).

6. Kivelson, M. G. et al. Galileo magnetometer measurements: a stronger case for a

subsurface ocean at Europa. Science 289, 1340–1343 (2000).
7. Greenberg, R. G. et al. Chaos on Europa. Icarus 141, 263–286 (1999).
8. O’Brien, D. P., Geissler, P. & Greenberg, R. A melt through model for chaos

formation on Europa. Icarus 156, 152–161 (2002).
Schenk, P. & Pappalardo, R. T. Topographic variations in chaos on Europa:
implications for diapiric formation. Geophys. Res. Lett. 31, L16703, doi:10.1029/
2004GL019978 (2004).

10. Collins, G. C., Head, J. W. III, Pappalardo, R. T. & Spaun, N. A. Evaluation of models
for the formation of chaotic terrain on Europa. J. Geophys. Res. 105, 1709–1716
(2000).

11. Spaun, N. A. et al. Conamara Chaos region, Europa: reconstruction of mobile

polygonal ice blocks. Geophys. Res. Lett. 25, 4277–4280 (1998).

12. Rathbun, J. A., Musser, G. S. Jr & Squyres, S. W. Ice diapirs on Europa: implications

for liquid water. Geophys. Res. Lett. 25, 4157–4160 (1998).

13. Pappalardo, R. & Barr, A. C. The origin of domes on Europa: the role of thermally
induced compositional diapirism. Geophys. Res. Lett. 31, L01701, doi:10.1029/
2003GL019202 (2004).

14. Head, J. W. & Pappalardo, R. T. Brine mobilization during lithospheric heating on

Europa: implications for formation of chaos terrain, lenticular texture, and color
variations. J. Geophys. Res. 104 (E11), 27143–27155 (1999).

©2011

Macmillan Publishers Limited. All rights reserved

LETTER RESEARCH

15. Sotin, C., Head, J. W. & Tobie, G. Tidal heating of upwelling thermal plumes and the
origin of lenticulae and chaos melting. Geophys. Res. Lett. 29, 1233, doi:10.1029/
2001GL013844 (2002).

16. McKinnon, W. B. Convective instability in Europa’s floating ice shell. Geophys. Res.

17. Han, L. & Showman, A. P. Coupled convection and tidal dissipation in Europa’s ice

Lett. 26, 951–954 (1999).

shell. Icarus 207, 834–844 (2010).

18. Collins, G. C. & Nimmo, F. in Europa (eds Pappalardo, R. T., McKinnon, W. B. &

Khurana, K.) 259–282 (Univ. Arizona Press, 2009).

19. Bjo¨ rnsson, H. Subglacial lakes and jo¨kulhlaups in Iceland. Glob. Planet. Change 35,

255–271 (2003).

20. Gudmundsson, M. T., Sigmundsson, F., Bjornsson, H. & Hognadottir, T. The 1996
eruption at Gjalp, Vatnajo¨ kull ice cap, Iceland: efficiency of heat transfer, ice
deformation and subglacial water pressure. Bull. Volcanol. 66, 46–65 (2004).
21. Kovacs, A. & Gow, A. J. Brine infiltration in the McMurdo Ice Shelf, McMurdo Sound,

Antarctica. J. Geophys. Res. 80, 1957–1961 (1975).

22. Scambos, T. et al. Ice shelf disintegration by plate bending and hydro-fracture:

satellite observations and model results of the 2008 Wilkins ice shelf break-ups.
Earth Planet. Sci. Lett. 280, 51–60 (2009).

23. MacAyeal, D., Scambos, T. A., Hulbe, C. L. & Fahnestock, M. A. Catastrophic ice-shelf
break-up by an ice-shelf fragment-capsize mechanism. J. Glaciol. 49, 22–36
(2003).

24. McCord, T. B. et al. Hydrated salt minerals on Europa’s surface from the Galileo

NIMS investigation. J. Geophys. Res. 104 (E5), 11827–11851 (1999).

25. Carlson, R. W. et al. Sulfuric acid production on Europa: the radiolysis of sulfur in

water ice. Icarus 157, 456–463 (2002).

26. Gaidos, E. & Nimmo, F. Tectonics and water on Europa. Nature 405, 637 (2000).
27. Nye, J. F. & Potter, J. R. The use of catastrophe theory to analyse the stability and

toppling of icebergs. Ann. Glaciol. 1, 49–54 (1980).

28. Guttenberg, N. et al. A computational investigation of iceberg capsize as a driver of

explosive ice-shelf disintegration. Ann. Glaciol. 52, 51–59 (2011).

29. Mevel, L. & Mercier, E. Large-scale doming on Europa: a model of formation of

Thera Macula. Planet. Space Sci. 55, 915–927 (2007).

30. Blankenship, D. D., Young, D. A., Moore, W. B. & Moore, J. C. in Europa (eds

Pappalardo, R. T., McKinnon, W. B. & Khurana, K.) 631–654 (Univ. Arizona Press,
2009).

Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.

Acknowledgements We thank D. Young, K. Soderlund, A. Barr, J. Greenbaum, J. Leisner
and D. MacAyeal for comments and discussions on the development of these concepts.
B.E.S. was supported by a fellowship from the Vetlesen Foundation and the Institute for
Geophysics of the Jackson School of Geosciences, University of Texas at Austin (UTIG).
D.D.B. was supported by NASA, NSF and UTIG. NASA supported the work of G.W.P. and
P.M.S.

Author Contributions B.E.S. and D.D.B. conceived of and actively discussed this
project. B.E.S. analysed Galileo imaging data, found and studied terrestrial analogue
information, analysed results, formulated the model, calculated values, and wrote the
paper. D.D.B. provided discussion and direction, and edited the paper. G.W.P.
performed the FFT analysis of Conamara Chaos topography data. P.M.S. produced the
original DEM of Conamara and the photoclinometry of Thera Macula.

Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of this article at
www.nature.com/nature. Correspondence and requests for materials should be
addressed to B.E.S. (britneys@ig.utexas.edu).

©2011

Macmillan Publishers Limited. All rights reserved

2 4 N O V E M B E R 2 0 1 1 | V O L 4 7 9 | N A T U R E | 5 0 5