Content uploaded by Simon A. Kattenhorn
Author content
All content in this area was uploaded by Simon A. Kattenhorn on Jan 23, 2015
Content may be subject to copyright.
Kattenhorn and Hurford: Tectonics of Europa 199
199
Tectonics of Euro p a
Simon A. Kattenhorn
University of Idaho
Te r ry Hu r f or d
NASA Goddard Space Flight Center
Europa has experienced significant tectonic disruption over its visible history. The descrip-
tion, interpretation, and modeling of tectonic features imaged by the Voyager and Galileo mis-
sions have resulted in significant developments in four key areas addressed in this chapter:
(1) The characteristics and formation mechanisms of the various types of tectonic features; (2) the
driving force behind the tectonics; (3) the geological evolution of its surface; and (4) the ques-
tion of ongoing tectonics. We elaborate upon these themes, focusing on the following elements:
(1) The prevalence of global tension, combined with the inherent weakness of ice, has resulted
in a wealth of extensional tectonic features. Crustal convergence features are less obvious but
are seemingly necessary for a balanced surface area budget in light of the large amount of ex-
tension. Strike-slip faults are relatively common but may not imply primary compressive shear
failure, as the constantly changing nature of the tidal stress field likely promotes shearing re-
activation of preexisting cracks. Frictional shearing and heating thus contributed to the mor-
phologic and mechanical evolution of tectonic features. (2) Many fracture patterns can be
correlated with theoretical stress fields induced by diurnal tidal forcing and long-term effects
of nonsynchronous rotation of the icy shell; however, these driving mechanisms alone prob-
ably cannot explain all fracturing. Additional sources of stress may have been associated with
orbital evolution, polar wander, finite obliquity, ice shell thickening, endogenic forcing by
convection and diapirism, and secondary effects driven by strike-slip faulting and plate flex-
ure. (3) Tectonic resurfacing has dominated the ~40–90 m.y. of visible geological history. A
gradual decrease in tectonic activity through time coincided with an increase in cryomagmatism
and thermal convection in the icy shell, implying shell thickening. Hence, tectonic resurfacing
gave way to cryomagmatic resurfacing through the development of broad areas of crustal dis-
ruption called chaos. (4) There is no definitive evidence for active tectonics; however, some
tectonic features have been noted to postdate chaos. A thickening icy shell equates to a de-
creased tidal response in the underlying ocean, but stresses associated with icy shell expansion
may still sufficiently augment the contemporary tidal stress state to allow active tectonics.
1. INTRODUCTION AND
HISTORICAL PERSPECTIVE
It took 369 years after the discovery of Europa, the
smallest of the Galilean moons, before humans finally
managed a close look at this icy world as the Voyager space-
craft sped by in 1979. The analysis of Voyager images of
Europa (e.g., Finnerty et al., 1981; Pieri, 1981; Helfenstein
and Parmentier, 1980, 1983, 1985; Lucchitta and Soder-
blom, 1982; McEwen, 1986; Schenk and McKinnon, 1989),
which had resolutions of 2 km to tens of kilometers per
pixel, resulted in the identification of a multitude of super-
posed crosscutting lineaments. The overall appearance,
much akin to a “ball of string” (Smith et al., 1979), spoke
of a history of intense tectonic activity (Fig. 1), but a pau-
city of large impact craters suggested a geologically young
surface. The surface deformation suggested an efficient
tectonic resurfacing process, perhaps accompanied by cryo-
volcanism, resulting in a broad classification of Europa’s
surface into tectonic terrain and chaotic or mottled (cryo-
magmatically disaggregated) terrain (Smith et al., 1979;
Lucchitta and Soderblom, 1982). Hence, competing tectonic
and endogenic processes have both been important in shap-
ing Europa’s geology. The notion of a tectonically active
world implied an effective tidal forcing of the icy shell,
leading researchers to hypothesize about the presence of a
liquid ocean beneath the icy exterior of the moon (see chap-
ter by Alexander et al.).
Detailed analyses of higher-resolution (tens to hundreds
of meters per pixel) Galileo spacecraft images 20 years after
Voyager resulted in a more precise classification of the
many types of lineaments (e.g., troughs, ridges, bands, cy-
cloids, strike-slip faults) and a reexamination of lineament-
formation mechanisms (e.g., summary in Papp al ard o et a l. ,
1999). In many studies, crosscutting relationships among
multiple episodes of lineaments allowed a complex tectonic
history to be unraveled. Cryomagmatism and chaos forma-
tion (see chapter by Collins and Nimmo) generally post-
dated tectonic resurfacing, although there was broad overlap
between tectonic and cryomagmatic processes and some
tectonic features are clearly geologically young (see chapter
by Doggett et al.). Most of the emphasis in tectonic analy-
ses of Galileo data was given to determining what caused
individual features to form in the first place, and what mech-
200 Europa
anisms resulted in their morphologic and geometric differ-
ences. Over the past dozen or so years, this work has pro-
duced a convincing framework for the role of tidal stresses
in deforming the icy shell to produce fracturing, assisted
by some amount of buildup of stress due to nonsynchronous
rotation (see chapter by Sotin et al.). These stresses can be
quantified and show a remarkable correlation to many lin-
eament orientations after accounting for the effects of lon-
gitudinal migration of the icy shell by nonsynchronous
rotation. The magnitude of the tidal distortion required for
pervasive fracturing of the icy shell strongly suggests the
presence of a tidally responding ocean beneath a relatively
thin shell that was repeatedly stressed and strained. The
presence of such an ocean is also strongly implied by Gali-
leo magnetometer measurements (Kivelson et al., 2000; see
chapter by Khurana et al.).
The goal of this chapter is to provide a thorough exami-
nation of the wide range of tectonic features that pervasively
damaged the icy shell of Europa during its visible geologi-
cal history and their implications for characterizing the
nature of the icy shell, the interior dynamics, the tidal de-
formation history, and the prospects for active tectonics. To
accomplish this, the chapter is divided into three main com-
ponents. First, we discuss the tectonic features themselves,
summarizing their likely formation mechanisms in terms of
being extensional, compressive, or lateral shearing struc-
tures. Second, we provide an overview of the causal factors
that drive tectonic deformation at all scales, focusing on the
production of stresses in the icy shell through tidal forcing
and other means. Finally, we examine the tectonic history
of Europa interpreted from the various types of tectonic
features placed in the context of the tidal stress history. In
so doing, we also address a number of topical issues such
as the evidence for diminishing tectonic activity through
time, reasons for the disparate geometries of lineaments, and
the prospect of active tectonics. We do not speculate on the
exact thickness of the icy shell in this chapter other than to
infer that it is sufficiently thin (likely <30 km) to enable a
strong tidal response in the underlying ocean, and conse-
quently inducing tectonic deformation of the shell. One
reason for this omission is that any constraints on icy shell
thickness based on tectonic features are inconclusive. Sec-
ond, fractures that have been used to deduce brittle or elastic
thicknesses do not necessarily capture the entire thickness
of the icy shell (e.g., Billings and Kattenhorn, 2005), much
of which may be behaving inelastically depending on the
timescale of deformation (see chapter by Nimmo and
Manga). Various terms have been used to describe tectonic
features on Europa. We aim to define and standardize the
relevant terms adopted in this chapter and that of Prockter
and Patterson with the glossary found in Table 1.
2. DEFORMATION STYLES
2.1. Extensional Tectonics
Analyses of low-resolution Voyager images followed by
higher-resolution Galileo images resulted in the recognition
and characterization of ubiquitous fractures in Europa’s icy
shell (Fig. 2). The prevalence of linear or curvilinear linea-
ments (e.g., Lucchitta and Soderblom, 1982) led to com-
parisons to terrestrial tension cracks, resulting in a body of
published works that interpreted the majority of europan
lineaments as tensile features (e.g., McEwen, 1986; Leith
and McKinnon, 1996). Exceptions included the identifica-
tion of lateral offset lineaments, interpreted to be strike-slip
faults (see section 2.3). The assumption that tensile failure
predominates in the icy shell is supported by numerous lines
of evidence. First, ice has been shown to be particularly
weak in tension under terrestrial conditions and in low-tem-
Fig. 1. Global mosaic of Europa, in Mercator projection, based on Voyager and Galileo imagery (courtesy of USGS, Flagstaff). The
center is at longitude 180°W and the latitude range is +57° to –57°.
Kattenhorn and Hurford: Tectonics of Europa 201
band — A general term used to describe a tabular feature that
formed by dilation (dilational band), contraction (convergence
band), and/or strike-slip motions (band-like strike-slip fault).
A dilational band is lineated (faulted or ridged varieties) or
smooth. In low-resolution images, relative albedo may be used
to distinguish between a bright band, gray band, or dark band.
band-like strike-slip fault — A type of strike-slip fault that mor-
phologically resembles a dilational band and implies an ob-
lique cumulative opening vector.
bright band — A term used to refer to high-relative-albedo, tabular
features in low-resolution imagery. These bands may have ex-
perienced some combination of dilation, strike-slip, and/or con-
vergence.
chaos — Regions of crustal disruption or disaggregation related
to an underlying endogenic process such as cryomagmatism.
complex ridge — See ridge complex. Use of this term is discour-
aged.
convergence band — a tabular zone that appears to represent miss-
ing crust after tectonic reconstructions are undertaken, implying
a localized zone of contraction.
crack — See trough.
cycloid — Curved, cuspate structures that form in chains of ar-
cuate segments linked at sharp cusps. If a central trough is
flanked by ridge edifices, the feature is called a cycloidal ridge.
Also called a flexus (plural flexu
–s).
cycloidal fracture — A type of trough that forms arcuate segments
linked at sharp cusps. They are the ridgeless progenitor of a
cycloidal ridge.
cycloidal ridge — A cycloid that has developed ridges to either
side of a central trough.
dark band — A term sometimes used to refer to a low-albedo
dilational band.
dilational band — A tabular zone of dilation in the icy shell where
intrusion of new crustal material occurred between the walls of
a crack. Also called a pull-apart band. If fine, internal linea-
tions are observed, the term lineated band may be used. The
lineations may be defined by normal faults (hence, a faulted
band) or by parallel ridges (a ridged band). If no lineations
are observable (commonly a resolution effect), the term smooth
band may be used.
diurnal tidal stress — The stress produced globally in the icy shell
in response to the oscillating tidal response of the satellite dur-
ing its eccentric orbit.
double ridge — See ridge.
endogenic fracture — A type of trough that forms above or ad-
jacent to a zone of endogenic activity in the icy shell, thus com-
monly occurring adjacent to regions of chaos.
faulted band — A type of lineated band in which the lineations
are caused by normal faults that have dissected the surface of
the dilational band.
flexure fracture — A type of trough that forms alongside a ridge
in response to flexing of the icy shell beside a ridge.
flexus — See cycloid.
fold — A rare form of contractional deformation in which the icy
shell warps into anticlinal and synclinal undulations, such as
within dilational bands.
fold hinge fracture — A type of trough that forms along the crest
of an anticline.
gray band — A term sometimes used to refer to an intermediate
relative albedo dilational band.
lineated band — A type of dilational band characterized by a fine
lineated internal texture. This term may be used generically
regardless of the inferred cause of the lineation. Two varieties
are faulted bands and ridged bands.
nonsynchronous rotation — The proposed process by which the
icy shell gradually migrates longitudinally eastward (i.e., about
the rotational poles) in response to rotational torques. As a re-
sult, all locations on the surface migrate across the tidal bulges,
resulting in a global component of stress that may contribute
to the tectonics.
normal fault — An extensional shear fracture across which a ver-
tical component of motion is inferred. Interpreted to define the
lineations within a faulted band.
protoridge — The progenitor of a ridge, composed of a central
trough flanked by poorly developed edifices. Also called a
raised-flank trough.
pull-apart band — See dilational band.
raised-flank trough — See protoridge.
ridge — The most common tectonically related feature on Europa,
comprising a central crack or trough flanked by two raised edi-
fices, up to a few hundred meters high and less than 5 km wide.
Also called a double ridge.
ridge complex — Several adjacent ridges that can be mutually
parallel or commonly sinuous and anastomosing. Individual
ridges in the complex are readily identifiable.
ridged band — A type of lineated band in which the lineations
are created by numerous parallel ridges that define the dila-
tional band.
ridged plains — The oldest and most expansive portions of the
surface of Europa, composed of a multitude of low, generally
high-albedo ridges, bands, and other structures that were re-
peatedly overprinted by younger features. Also called subdued
plains.
ridge-like strike-slip fault — A type of strike-slip fault that mor-
phologically resembles a ridge.
small-circle depressions — Up to 1.5-km-deep depressions in the
icy shell that form broad, circular map patterns centered ~25°
from the equator in an antipodal relationship on the leading
and trailing hemispheres.
smooth band — A type of dilational band lacking an observable
internal lineated texture, commonly in response to image reso-
lution constraints, that may include small-scale hummocks.
strike-slip fault — A lineament along which older crosscut fea-
tures were translated laterally during shearing.
subdued plains — See ridged plains.
tailcrack — A type of trough that forms where tension occurs at
the tip of a strike-slip fault in response to fault motion.
tectonic fracture — A type of linear or broadly curving trough
that forms locally or regionally, probably in response to ten-
sile tidal stress.
triple band — A now-discouraged term originally used to describe
lineaments in low-resolution imagery that appeared as a bright
central stripe flanked by dark edges. Such features were ulti-
mately identified in higher-resolution images to be predomi-
nantly ridge complexes flanked by dark materials, but may also
be double ridges or bright bands flanked by dark materials.
trough — A ridgeless fracture with a visible width that results in
a linear indentation at the surface of the icy shell. Also called
a crack.
TABLE 1. Glossary of terms used to describe tectonic features on Europa.
202 Europa
Fig. 2. (a) Conamara Chaos in the northern trailing hemisphere (orthographic mosaic E6ESDRKLIN01). The region is dominated
by the prominent tectonic lineaments Asterius and Agave Lineae and the cryomagmatically disrupted region of chaos south of where
the two lineaments cross. (b) The Bright Plains region (Galileo mosaic E6ESBRTPLN02) highlights the wide range of tectonic fea-
tures in different orientations that characterizes much of the europan surface. This region shows ridges, dilational bands, and troughs,
with the prominent Androgeos Linea cutting across the center of the image. Images courtesy of the NASA Space Photography Labo-
ratory at Arizona State University.
perature experiments, with tensile strengths of perhaps hun-
dreds of kilopascals to as much as ~2 MPa (cf. Schulson,
1987, 2001; Rist and Murrell, 1994; Lee et al., 2005). Sec-
ond, unlike Earth, Europa’s surface experiences absolute
tension on a regular basis due to oscillating tidal bulges dur-
ing the diurnal cycle (see section 4.1.5 and chapter by Sotin
et al.). Third, some lineaments of particular ages typically
form as multiple, parallel sets analogous to terrestrial ten-
sion joint sets. Fourth, the majority of lineaments on Europa
do not appear to show lateral offsets, implying only crack-
orthogonal motions (mode I cracks, in fracture mechanics
terminology). Finally, there is clear evidence of complete
dilational separation of parts of the icy shell, with infill of
material from below to create new surface area within the
dilated crack. A number of tensile tectonic features have
been identified on the basis of these lines of evidence, in-
cluding ridges, cycloids, dilational bands, and troughs. Ad-
ditionally, normal faults on Europa indicate that extensional
tectonics can occur through shear failure where there is
deviatoric tension in a compressive stress field. Inferred
extensional features, called small-circle depressions (Schenk
et al., 2008), form an antipodal pattern in the leading and
trailing hemispheres, but an explicit formation mechanism
has not been identified (see section 4.1.7).
2.1.1. Ridges. The most common lineaments on Eu-
ropa are ridges, also referred to as double ridges due to the
typical morphology in which a central crack or trough is
flanked by two raised edifices (e.g., Androgeos Linea in
Fig. 2b). As a result, ridges typically appear in medium- to
high-resolution surface images as having a central dark
stripe (the trough) flanked by two bright thicker stripes (the
raised edifices; Fig. 3c), and may extend from a few kilo-
meters to in excess of 1000 km across the surface. Ridges
are analyzed in great detail elsewhere in this book (see
chapter by Prockter and Patterson) but are also described
here due to their prominence among Europa’s tectonic fea-
tures. Not only are ridges the most common type of struc-
tural lineament on Europa, they are also the most persis-
tent, being the primary component of the very oldest ridged
plains of the icy surface (Figueredo and Greeley, 2000; Kat-
tenhorn, 2002; Doggett et al., 2007) as well as constituting
some of the youngest geological features. Hence, whatever
process is responsible for their development must have been
an ongoing process throughout Europa’s visible geological
history. Ridges of different ages commonly occur in dis-
parate orientations, resulting in a complex network of mul-
tiply crosscutting generations of ridge sets and indicating
significant changes in stress fields through time (see sec-
tions 5.1 and 5.2). Topographically, ridges range from
barely perceptible in parts of the background plains to as
high as several hundred meters (Malin and Pieri, 1986;
Greenberg et al., 1998; Greeley et al., 2000). Individual
ridge widths are typically around a few hundreds of meters
(<400 m) but higher ridges tend to be wider, with maximum
Kattenhorn and Hurford: Tectonics of Europa 203
surface widths of <5 km (Coulter et al., 2009). Kattenhorn
(2002) examined the 20 m/pixel “Bright Plains” images
(Fig. 2b) and concluded that ridges became higher, wider,
and fewer in number during the geological sequence in that
region.
Ridges do not necessarily occur in isolation but may de-
velop prominent lineaments comprising several, commonly
braided and inosculating, superposed ridges up to 15 km
wide (Fig. 3d). Such features are called ridge complexes
(e.g., Greenberg et al., 1998) but are sometimes referred
to as complex ridges (Figueredo and Greeley, 2000, 2004;
Greeley et al., 2000) and imply that an early formed ridge
may create a zone of weakness that localizes and superposes
later ridge development (Patterson and Head, in preparation),
possibly accompanied by shearing (Aydin, 2006). Promi-
nent examples include Agave and Asterius Lineae in the
Conamara Chaos region (Fig. 2a). In low-resolution images,
ridge complexes commonly appear bright with flanking
dark stripes, and were originally referred to as triple bands
(Lucchitta and Soderblom, 1982), now a defunct term as it
does not imply any explicit geological feature type. Ridges
and ridge complexes can produce a surface line-load that
impinges on the elastic portion of the icy shell, resulting in
downward flexing to either side of the ridge (Tufts, 1998;
Head et al., 1999; Kattenhorn, 2002; Billings and Katten-
horn, 2005; Hurford et al., 2005). The associated bending
stresses may induce flanking lines of tension fractures to
either side of the ridge (see section 2.1.4).
A key observation about ridges is that they appear to rep-
resent an advanced stage in a genetic sequence of lineament
development (Geissler et al., 1998a; Head et al., 1999)
(Fig. 3). This idea stemmed from the fact that some ridges
change in morphology along their trend, becoming less
prominent along their lengths with a central crack flanked
by underdeveloped edifices (Fig. 3b). This less-evolved
stage has been called a protoridge (Kattenhorn, 2002) or
a raised-flank trough (Head et al., 1999). Ultimately, the
protoridge may disappear along its trend to reveal its pro-
genitor: a ridgeless crack called a trough (Fig. 3a). This
sequence of development has been interpreted to imply that
all ridges originate as troughs that form as tension cracks,
evolving into protoridges and then finally ridges (Fig. 3c)
or sometimes ridge complexes (Fig. 3d), although the na-
ture of this evolution (i.e., the ridge growth process) is
neither straightforward nor generally agreed upon.
The majority of published models of ridge development
assume that they are fundamentally tensile features, but with
differing mechanisms to explain the manner in which the
ridge edifices are constructed. Kadel et al. (1998) attributed
ridge growth to the effects of gas-driven cryovolcanic fis-
sure eruptions. Weaknesses of this model include no expla-
nation for a subsurface source of cryovolcanic material and
the inherent difficulty in attempting to account for the re-
markably consistent morphology of ridges, commonly for
hundreds of kilometers along their lengths. A linear diapir-
ism model (Head et al., 1999) invokes buoyant upwelling
of ice beneath a tension crack near the surface, resulting in
an upward bending of the brittle carapace to either side of
the central crack, thus forming ridges. Problems with this
model include no explicit mechanism to explain why such
upwelling would occur beneath a surface crack (although
see the shear heating discussion below, for one possibility)
and it necessitates that the preexisting terrain is preserved
within the upwarped ridge slopes. This final point does not
seem to be characteristic of europan ridges even though
some examples have been described (Giese et al., 1999;
Head et al., 1999; Cordero and Mendoza, 2004). The along-
trend morphology of a well-developed ridge is invariably
unaffected by the nature of the ridged plains it crosses,
which may imply that upward warping of these plains is
not the primary process by which ridges are created. Mass
wasting could conceivably have removed the original sur-
face roughness along ridges; however, ridge slopes are typi-
cally less than 30° and on average only about 10° (Coulter
Fig. 3. Progressive evolution of fractures from (a) simple troughs to the development of (b) nascent
ridges, then (c) fully developed ridges, and finally (d) a ridge complex. The image in (a) was taken
during Galileo’s E11orbit, image 420626739; those in (b)–(d) are all from the Conamara Chaos and
Bright Plains region shown in Fig. 2, taken during Galileo’s E6 orbit. The ridge in (c) is Androgeos
Linea. The ridge complex in (d) is Agave Linea.
204 Europa
et al., 2009), which seems to be too low to be consistent
with a mass wasting process. An incremental wedging model
(Turtle e t a l. , 1998; Melosh and Turtle, 2004) explains ridge
development as being due to a gradual forcing apart of
the walls of a crack intruded by material from below that
freezes within the crack, causing upward plastic deforma-
tion at the surface to construct the ridge. In this model, it
is unclear what the source of the intruded material would
be, or why it is so readily accessible by surface cracks.
The ridge model of Greenberg et al. (1998) attributes
their formation to the cyclical extrusion of a slurry of lead
ice created by exposure and instantaneous near-surface
freezing of an underlying ocean during the opening of a
crack by diurnal tidal stresses. During subsequent closure
of the crack, the ice is squeezed onto the surface, gradu-
ally building piles of ice debris to either side of the central
crack. A caveat to this model is the need for a very thin icy
shell in order to reconcile the requirement of complete sepa-
ration of the ice layer to expose the underlying ocean (e.g.,
Rudolph and Manga, 2009) with the fact that tidal stresses
are typically very small (likely a few tens of kilopascals)
and thus would quickly be overcome by overburden pres-
sure with increasing depth. Furthermore, because of the
density contrast, ocean water can only rise to a level about
10% of the total icy shell thickness down from the surface.
It would therefore have become increasingly difficult for
material to be squeezed up to the surface if the icy shell
thickened through time, as is suggested by thermodynamic
models and the evolution of surface geological features (see
section 5.2), unless the underlying ocean is sufficiently pres-
surized to allow cryovolcanic eruption, which is unlikely to
be the case on Europa (Manga and Wang, 2007).
Gaidos and Nimmo (2000) and Nimmo and Gaidos
(2002) infer that ridge development is ultimately driven by
frictional shear heating along an existing crack (see sec-
tion 2.3.1), even for friction coefficients as low as 0.1.
Shearing is driven by the ever-rotating diurnal tidal stress
field (see section 4.1.5) with inferred diurnal timescale shear
velocities of 10–6–10–7 m s–1. Subsequent heating results in
an almost sevenfold increase in the local surface heat flux
causing buoyant rising of warm ice toward the surface and
resultant upwarping of a near-surface brittle layer to form
a ridge. For shear velocities >~10–6 m s–1, local melting
along the shear zone will cause downward draining of any
melt products and the production of void space that may
promote sagging or lateral contraction. This model places
only mild depth constraints on the preexisting cracks that
get sheared (i.e., the cracks do not need to penetrate the icy
shell to an underlying ocean), it is based on a proven de-
formation mechanism on Europa (i.e., shearing; see sec-
tion 2.3), and is a process that has likely been occurring on
Europa as long as there has been a tidally responding ocean
beneath the icy shell. Caveats to the model include the lack
of observable lateral offsets along the majority of ridges and
the lack of a clear mechanism for how the buoyant upwell-
ing ultimately constructs the ridges. The mechanism appears
to resemble the linear diapirism model described above and
the two ideas are probably not mutually exclusive and thus
share similar caveats regarding the ridge construction proc-
ess. However, a source of heating may also perhaps explain
the relatively low outer slopes of ridges if viscoplastic flow
is the dominant form of ridge slope modification (Coulter
et al., 2009). Han and Showman (2008) explicitly model a
linear zone of thermal upwelling beneath a frictionally shear
heated zone, producing narrow, laterally continuous, ridge-
like features with heights of up to 120 m. Ridges are known
to attain heights of more than twice this amount; however,
this small disparity may perhaps be circumvented if there
is a component of fault-perpendicular contraction during the
shear process that contributes to the construction of ridges
(Nimmo and Gaidos, 2002; Vetter, 2005; Ayd in, 2006; Pat-
terson et al., 2006; Kattenhorn et al., 2007; Bader and Kat-
tenhorn, 2007) or additional buoyancy mechanisms such as
depletion of salts during heating (cf. Nimmo et al., 2003).
Possible evidence for contraction across ridges is described
in more detail in section 2.2.3, but essentially allows ridge
development to accommodate convergence across weak
zones caused by shear heating along cracks. The associated
contraction of the brittle icy shell at the surface may de-
velop a permanent positive relief structure, perhaps negat-
ing the problem of expected relaxation of buoyant-up-
welling-induced ridges over the thermal diffusion timescale
of ~107 yr (Han and Showman, 2008).
Regardless of the precise formation mechanism for
ridges, the fact remains that ridge building has been an ef-
fective geological process on Europa. Considering the rela-
tively young surface age based on crater densities (40–
90 m.y.) (Zahnle et al., 2003; see chapter by Bierhaus et
al.) and the great number of ridges of different ages and
orientations, each individual ridge must form in a relatively
short period of time. Greenberg et al. (1998) suggest that
ridges may form within 30,000 yr based on their tidal pump-
ing model; however, that estimate is necessarily based on
a number of uncertain assumptions regarding crack dilation
and spacing, as well as the effectiveness of the ice extru-
sion process. Melosh and Turtle (2004) estimate 10,000 yr
to form a ridge through an incremental ice-wedging pro-
cess. Gaidos and Nimmo (2000) estimate that shear heat-
ing may result in buoyant uprising of warm ice at a rate that
could conceivably construct a ridge in only ~10 yr. Presum-
ably, the amount of time during which a crack may remain
active (whether in dilation, shearing, or contraction) with
the capability of developing a ridge is ultimately controlled
by the amount of time over which the global tidal stresses
drive crack activity. This timing may be controlled by the
rate of nonsynchronous rotation, if present, relative to a
tidally locked interior (see sections 4.1.6 and 5.1), which
would eventually rotate an active crack away from a stress
field conducive to crack activity. One cycle of nonsynchro-
nous rotation takes in excess of 12,000 yr (Hoppa et al.,
1999b) and perhaps as much as 1.3 m.y. [cf. Hoppa et al.
(2001), accounting for the revised surface age of Europa
in the chapter by Bierhaus et al., and a factor of 3 uncer-
tainty], suggesting that an estimate of a few tens of thou-
Kattenhorn and Hurford: Tectonics of Europa 205
sands of years during which a crack can remain active and
thus form a ridge may be reasonable.
2.1.2. Cycloids. Europa also exhibits unique features
morphologically similar to ridges called cycloids, also re-
ferred to as cycloidal ridges or flexu
–s. Cycloids, first de-
scribed from Voyager data (Pieri, 1981; Lucchitta and
Soderblom, 1982; Helfenstein and Parmentier, 1983), are
curved, cuspate cracks that form chains of multiple, con-
catenated segments extending hundreds to thousands of
kilometers across the surface (Fig. 4). Each curved segment
of a cycloid chain is linked to the adjacent one at an abrupt
kink called a cusp. Individual segments are typically tens
of kilometers long, measured in a direct line from cusp to
cusp, but are locally up to several hundred kilometers long.
The central crack is commonly, but not necessarily, flanked
by ridges, implying that progressive ridge development
occurs to either side of an initial crack, analogous to linear
ridges described in section 2.1.1. In some cases, ultimate
dilation of the cracks occurs to form cycloidal bands (see
chapter by Prockter and Patterson). Cycloids are distinctly
different from other tectonic cracks on Europa and, other
than one possibly analogous feature in the south polar re-
gion of Enceladus, are unique in the solar system. There-
fore, their mode of formation must also differ from other
tectonic lineaments on Europa, implying significant vari-
ability in crack-driving processes during the long-term tec-
tonic history.
Hoppa and Tufts (1999) and Hoppa et al. (1999a) pro-
posed that cycloidal cracks form as a result of tensile crack-
ing in response to diurnally varying tidal stresses produced
by Europa’s orbital eccentricity (these stresses are described
in detail in section 4.1.5; also see chapter by Sotin et al.).
As Europa orbits Jupiter, it is constantly being reshaped as
the size and location of the tidal bulges change, producing
a time-dependent diurnal tidal stress field. At any location
on the surface during the orbital period, the orientations of
the principal stresses rotate and the magnitudes oscillate.
The stresses rotate counterclockwise in the northern hemi-
sphere and clockwise in the southern hemisphere, 180° each
orbit (while only magnitudes change along the equator). The
individual cycloid segments are hypothesized to grow as
tension fractures that propagate perpendicular to the rotating
direction of maximum tensile stress, resulting in curved seg-
ments. Cracking begins when the tensile strength of europan
ice is overcome and ceases when the tensile stress drops
below the crack propagation strength, assumed to be less
than the tensile strength. These parameters are unknowns
on Europa; therefore, Hoppa et al. (1999a, 2001) estimated
them in such a way so as to provide the best match between
observed cycloid shapes and theoretical stress fields. This
growth model for cycloids predicts that all northern hemi-
sphere cycloids that are concave toward the equator grew
from east to west, whereas all cycloids concave toward the
poles grew from west to east. Similarly, northern hemi-
sphere cycloids that are concave to the east are predicted
to have grown from north to south, whereas cycloids that
are concave to the west grew from south to north. The
opposite sense of growth is true in all cases in the southern
hemisphere. Also, because the diurnal stress characteristics
are dependent on latitude and longitude, expected cycloid
shapes are correspondingly variable across the surface of
Europa, ranging from broadly curved to squarish (Bart et
Fig. 4. (a) Voyager 2 image (c2065219) showing several chains
of cycloidal ridges. (b) Example of a cycloidal fracture in the
northern trailing hemisphere with ridge development along a small
part of the arc (Galileo image 449961879, E15 orbit).
206 Europa
al., 2003; Hurford et al., 2007a). Cycloid chains commonly
exhibit a large-scale curvature due to regional variability in
global tidal stress fields. The growth of these broadly curved
cycloid chains ultimately ceases when they propagate into
an area where critical stress values are not attained, result-
ing in some regions where cycloids theoretically never form
(Fig. 5).
The growth of a single cycloid arc is assumed to be a
continuous process during the course of a europan day and
must occur at a low crack propagation speed (~3–5 km/h)
in order for curved segments to produce a match to the pre-
dicted stresses. Actual growth speeds may be much faster,
albeit occurring in short, discrete spurts to produce a much
lower effective propagation speed (Lee et al., 2005). After
cessation of growth of a cycloid arc, the next arc starts
growing when the tensile strength of the ice is again over-
come in the subsequent orbit; however, because the stresses
rotate during the period of no crack growth, the new cyc-
loid arc propagates at an angle away from the tip of the
previous one, forming a sharp cusp. These rotated stresses
resolve a component of shearing along the arrested cycloid
segment immediately prior to the development of the cusp,
inducing a concentration of stress at its tip that drives the
development of the new cycloid arc (Marshall and Katten-
horn, 2005; Groenleer and Kattenhorn, 2008). In this way,
cycloid cusps form identically to features called tailcracks
that develop at the tips of strike-slip faults on Earth and
elsewhere throughout the solar system, including Europa
(see section 2.3.3). The tailcrack then continues to grow in
tension to form a new cycloid arc, driven by the diurnal
stresses. Ongoing shearing near cycloid cusps has also been
suggested to be the cause of multiple tailcrack-like splays
of fractures emanating from cusp regions, producing com-
plex cusps (Marshall and Kattenhorn, 2005) that resemble
horsetail fracture splays along terrestrial strike-slip faults.
2.1.3. Dilational bands. Also referred to as pull-apart
bands (see chapter by Prockter and Patterson), dilational
bands represent clear evidence of prolonged dilation in the
icy shell and hence a resurfacing process on the icy moon.
A dilational band is a tabular zone of new crustal material
that intruded between the progressively dilating walls of a
tension fracture (Fig. 6a). The surface of this material ap-
Fig. 5. Theoretical distribution of cycloids that grow from east to west (Hurford et al., 2007a) in a
stress field composed of a diurnal component plus stress due to 1° of nonsynchronous rotation. East-
growing cycloids will have an opposite curvature in each hemisphere but a similar overall distribution.
Kattenhorn and Hurford: Tectonics of Europa 207
pears featureless in lower-resolution images, in which case
the term smooth band can be used as a descriptor; however,
if an internal geometry of fine lineations is observable (usu-
ally at medium to high resolution, as in Fig. 6a), the term
lineated band may be used. Complete separation and infill
of the surface is evidenced by the fact that the ridged plains
to either side of a dilational band typically match up, im-
plying that the dilational band material represents new sur-
face area (Schenk and McKinnon, 1989; Sullivan et al.,
1998). The exact source of the dilational band material is
unclear. Some have suggested that the dilation occurred
across cracks that fully penetrated the icy shell, with the
smooth material representing frozen portions of an exposed,
underlying ocean (Tufts et al., 2000). Alternatively, dila-
tional bands may represent regions where the brittle portion
of the icy shell dilated slowly enough that ductile ice un-
derwent buoyant upwelling from deeper portions of the icy
shell to fill in the dilational gap (Pappalardo and Sullivan,
1996; Sullivan et al., 1998; Prockter et al., 1999, 2002).
The low albedo of most young dilational band material has
been suggested to represent the effects of magnetospheric
particle bombardment of endogenic sulfur-bearing com-
pounds (Kargel, 1991; Noll et al., 1995); however, contin-
ued exposure at the surface ultimately results in a bright-
ening of the dilational band such that the oldest dilational
bands have a similar albedo to the surrounding ridged plains
(Pappalardo and Sullivan, 1996; Geissler et al., 1998a;
Prockter et al., 1999, 2002), which historically resulted in
use of the descriptor bright bands. Although tectonic plates
(in the terrestrial sense) do not exist in the icy shell (see
section 3), dilational bands represent a spreading phenom-
enon analogous to mid-ocean ridge spreading centers, mak-
Fig. 6. (a) A dilational band (provisionally named Phaidra Linea) in the
equatorial trailing hemisphere. Note how features can be matched up to ei-
ther side of the band. Although the dilational band is one of the younger
features in this region, a number of younger ridges and troughs can be seen
to crosscut the band (from Galileo observation 11ESREGMAP01). (b) A
17-km-wide ridged band in the northern leading hemisphere (from Galileo
image mosaic 11ESMORPHY01). (c) The dark dilational band Yelland Linea
in the “Wedges” region, Argadnel Regio (from Galileo image mosaics
12ESWEDGE_01/02/03 superimposed on C3ESWEDGES01).
208 Europa
ing europan dilational bands the only other known feature
in the solar system where complete lithospheric separation
has occurred. Similar to mid-ocean ridges, dilational bands
stand higher than the surrounding plains by up to 200 m
(Prockter et al., 1999, 2002; Nimmo et al., 2003) and may
have a relatively reduced elastic thickness (in the range of
a few hundred meters to a few kilometers) (Prockter et al.,
2002; Billings and Kattenhorn, 2005; Stempel et al., 2005).
Opening vectors across dilational bands indicate that both
orthogonal and oblique dilations are common.
Similar to ridges, dilational bands can be hundreds of
kilometers in length, implying a regional process respon-
sible for the recorded period of spreading. Dilational band
widths generally do not exceed ~30 km and are commonly
only a few kilometers, suggesting that the process respon-
sible for driving the dilation is unable to sustain spreading
beyond a certain time and/or width limit. Nonetheless, a di-
lational band formation event likely represents a prolonged
period of uninterrupted tectonic extension. Discrete episodes
of dilation are evidenced by an internal fabric of fine linea-
ments within lineated bands (Fig. 6a) that commonly form a
bilateral symmetry about a central axis and exhibit a con-
sistent spacing on the order of ~500 m (Sullivan et al., 1998;
Prockter et al., 1999, 2002; Stempel et al., 2005). If these
lineaments are normal faults, creating a type of lineated band
called a faulted band, they may perhaps form a graben-like
system centered about the middle of the dilational band (see
section 2.1.5), with older faults being translated successively
further away from the actively spreading central axis over
time. At slow spreading rates, the faulted surface of the dila-
tional band may become rugged, with tilted fault blocks cre-
ating repeating valleys and ramparts (Prockter et al., 2002;
Stempel et al., 2005).
Some dilational bands have been noted to exhibit a maxi-
mum dilation near the center of the length of the band, with
dilation decreasing toward either tip, such as Thynia Linea
(Pa pp a la rd o a n d S ul l iv a n, 1996) and Yelland Linea (Fig. 6c)
(Stempel et al., 2005). Such dilational bands resemble typi-
cal cracks in an elastic layer that are dilated by a regional
tensile stress acting perpendicular to the feature. Nonethe-
less, some dilational bands are more rhomboidal where they
occur in extensional stepover zones (i.e., pull-aparts) along
strike-slip faults (see section 2.3.3) such as Astypalaea Linea
in the southern antijovian region (Tufts et al., 1999, 2000;
Kattenhorn, 2004a) and wedge-shaped bands in Argadnel
Regio (Schulson, 2002; Kattenhorn and Marshall, 2006),
implying a localized tectonic phenomenon.
Although it is possible that the driving stress responsible
for a dilational band also created the original crack across
which dilation ensued, some dilational bands show evidence
of raised flanks that appear to be the two halves of an erst-
while ridge. Hence, some dilational bands dilated preexist-
ing cracks that had already developed ridges along their
rims. Considering that ridge formation is not instantaneous
but may take tens of thousands of years and involve a range
of mechanisms (dilational, contractional, and shearing), the
process responsible for dilational band formation is unlikely
to be a natural endmember of the ridge formation process.
Ridges simply provide weakness zones within the icy shell
that may be utilized when the conditions conducive to dila-
tional band formation occur. The actual spreading across
dilational bands may be driven by underlying convection in
a region of locally high heat flow (Prockter et al., 2002),
perhaps originally initiated by the cracking of the icy shell
(cf. Han and Showman, 2008). This upwelling may explain
why portions of ridged plains immediately adjacent to some
dilational bands are also slightly elevated relative to the sur-
roundings (Nimmo et al., 2003). Comparisons to terrestrial
midocean-ridge spreading models suggest strain rates in the
range 10–15–10–12 s–1 at europan dilational bands (Nimmo,
2004a,b; Stempel et al., 2005), implying spreading rates
of 0.2–40 mm yr–1, similar to terrestrial mid-ocean ridge
spreading rates. Nimmo (2004b) estimates a maximum dura-
tion of 10 m.y. for dilational band formation, with the ob-
served narrow rift geometry of dilational bands implying
a high strain rate or a relatively thick shell (but probably
<15 km) at the time of dilational band formation. Stempel
et al. (2005) estimate the duration of spreading to be in the
range 0.1–30 m.y. (with lower estimates representing lower
coefficients of friction of the ice), which is well within the
limits of surface ages deduced from the cratering record
(Zahnle et al., 2003; see chapter by Bierhaus et al.). None-
theless, the great number and different ages of dilational
bands and other tectonic features on Europa suggest that
dilational band activity is likely to be at the lower end of
this duration estimate. This deduction is also supported by
the requirement for self-similar stress conditions during the
development of dilational bands that exhibit a uniform
lineated texture, although some dilational bands may record
a more complicated dilational/stress history. The needed
stresses to drive spreading are on the order of a few mega-
pascals (Stempel et al., 2005), which are likely provided by
nonsynchronous rotation (see section 4.1.6). Considering
that nonsynchronous rotation would result in a gradually
changing stress field in any area of dilational band forma-
tion, the duration of band formation must be less than the
time it would take for nonsynchronous rotation to move the
developing dilational band into a new stress state.
Some cycloidal cracks also show evidence of having di-
lated to form cycloidal dilational bands (Marshall and Kat-
tenhorn, 2005), including Thynia Linea (Papp al ardo and
Sullivan, 1996; Tufts et al., 2000), wedge-shaped bands in
Argadnel Regio (Prockter et al., 2002), and the prominent
example of the “Sickle” (provisionally named Phaidra Linea)
in the equatorial trailing hemisphere (Tufts et al., 2000;
Prockter et al., 2002) (Fig. 6a). In these examples, the open-
ing vector across each dilational band is constant, result-
ing in portions of the band having undergone oblique dila-
tion relative to the margins.
An apparent type of lineated dilational band morphologi-
cally similar to a faulted band, but composed of ridges, is
a tabular spreading zone referred to here as a ridged band
(cf. Figueredo and Greeley, 2000, 2004) (Fig. 6b). Stempel
et al. (2005) used this same terminology to refer to faulted
Kattenhorn and Hurford: Tectonics of Europa 209
bands; however, we abandon that use of the nomenclature
because ridges (which have an explicit meaning on Europa)
are not the dominant feature in faulted bands. In low-reso-
lution imagery, ridged bands may be mistaken for smooth
bands if the internal lineated texture is not observable; how-
ever, ridged bands are essentially up to 60-km-wide com-
plexes of multiple adjacent ridges and furrows (i.e., essen-
tially dilational bands composed of ridges). Ridged bands
appear to be analogous to the “Class 2 ridges” of Greenberg
et al. (1998) and Tufts et al. (2000) and have also been clas-
sified as “complex ridges” (Kattenhorn, 2002; Spaun et al.,
2003). This latter terminology causes confusion and should
be avoided in light of previous usages and to avoid confu-
sion with the definition of a ridge complex as a zone of mul-
tiple superposed and interweaving ridges (see section 2.1.1).
Ridged bands also represent zones of spreading in the icy
shell but the infill process differs from smooth bands and
faulted bands, although hybrid features may exist that take
on the appearance of both smooth bands and ridged bands
(Tufts et al., 2000). Instead of upwelling of ductile mate-
rial from below, ridged bands may represent a different form
of spreading. One model invokes injection of material from
below into discrete fractures, analogous to a terrestrial dike
swarm. At the surface, each of these cracks may then take
on the form of a ridge through some ridge-forming mech-
anism (see section 2.1.1). Like a dike swarm, there is no
necessity for symmetry about a central crack, as the se-
quence of intrusion through vertical dike-like features may
be somewhat random across the spreading zone (Tu fts et
al., 2000). Nonetheless, bilateral symmetry is possible (e.g.,
Kattenhorn, 2002).
2.1.4. Troughs. Considering the lack of raised edifices
along the crack margins, as typifies ridges, troughs are eas-
ily overlooked in images of the surface. They likely repre-
sent tension fractures in the ice shell that never developed
into a more evolved landform, such as a ridge, and have
been referred to simply as cracks. They have been acknowl-
edged (Figueredo and Greeley, 2000, 2004; Kadel et al.,
2000; Prockter et al., 2000; Kattenhorn, 2002) but have
received little attention, despite being relatively common in
high-resolution images of the surface. The majority of the
tectonic features on Europa likely owe their beginnings to
tension fracturing at the surface of the icy shell, starting as
troughs.
Confining stress considerations dictate that most frac-
tures are likely tension fractures that are initiated at the
surface and then propagate downward. Whether or not the
fractures completely penetrate the icy shell is dependent on
the availability of stresses to overcome the overburden,
which becomes increasingly difficult for thicker icy shells.
Diurnal tidal stresses (see section 4.1.5) are rapidly over-
whelmed by the overburden in the icy shell, which, for an ice
density of 0.91 g/cm3, increases at a rate of about 1.2 MPa/
km. Hoppa et al. (1999a) suggest that fracture depths should
thus not exceed ~65 m; however, consideration of crack-tip
stresses increase this depth to a few hundred meters (Lee et
al., 2005; Qin et al., 2007) and perhaps up to several kilo-
meters when ice porosity effects are considered, allowing
complete penetration of the icy shells if its thickness does
not exceed 3 km. The addition of stress due to nonsynchro-
nous rotation (see section 4.1.6) allows penetration depths of
several kilometers (Panning et al., 2006) to approaching
10 km (Golombek and Banerdt, 1990; Leith and McKinnon,
1996), perhaps allowing complete penetration of an icy shell
in the thickness range 6–13 km for a plausible range of ice
porosity (Lee et al., 2005). If there has been more than
10 km of thickening of the icy shell through time (see sec-
tion 4.1.8), sufficient stress may have been produced at the
base of the icy shell to initiate cracking and permit complete
penetration through the shell (Manga and Wang, 2007), with
upward propagation aided by fluid pressure (cf. Crawford
and Stevenson, 1988), in which case it would not be unrea-
sonable to assume that tension fractures provide potential
connection pathways between the surface and an underly-
ing ocean. Nonetheless, viscoelastic relaxation in the lower
part of the icy shell over the shell thickening timescale likely
limits complete penetration by tension fractures to icy shells
≤2.5 km thick (Rudolph and Manga, 2009).
Troughs constitute the youngest tectonic features on Eu-
ropa based on crosscutting relationships. Hence, they pro-
vide the most promising indicator of recent to current tec-
tonic activity (see section 5.3). Kattenhorn (2002) showed
that troughs make up a significant portion (the youngest
20%) of the tectonic history of the Bright Plains region.
Although the surface geological history appears to record
a transition from principally tectonic to predominantly cryo-
magmatic activity, with associated formation of regions of
chaos and lenticulae (see section 5.2 and chapter by Doggett
et al.), troughs have been noted to cross regions of chaos
and thus may be very recent features.
Troughs vary greatly in geometry, scale, orientation, and
location, reflecting differences in causal mechanisms driv-
ing fracturing (e.g., Figueredo and Greeley, 2004). A classifi-
cation scheme for troughs is suggested here to address these
differences (Fig. 7): (1) Tectonic fractures range in length
from <10 km to hundreds of kilometers (Fig. 7a). They are
presumably the result of global tidal stresses plus any other
global, regional, or local contributing stress components that
may drive global tectonics (see section 4). They exhibit
a range of orientations and are characterized by linear or
broadly curving geometries. They are sometimes segmented
along their lengths or may show evidence of having formed
by the coalescence of numerous segments, analogous to ter-
restrial joints. (2) Cycloidal fractures (Figs. 4b and 7b) are
also tectonic fractures but are specifically the ridgeless pro-
genitors to cycloidal ridges. They are distinct in their curved
and cuspate nature, probably reflecting the dominant control
of the diurnal stress field in their formation. Although cy-
cloidal ridges are more common, relatively younger cyc-
loidal fractures have also been identified (e.g., Marshall and
Kattenhorn, 2005), indicating that cycloid development has
persisted until at least geologically recent times. (3) Tail-
cracks emanate from the tips of strike-slip faults (Fig. 7b)
and represent brittle accommodation of fault motion in the
210 Europa
tensile quadrant of a fault tip (Schulson, 2002; Kattenhorn,
2004a; Kattenhorn and Marshall, 2006). Also referred to
as horsetail splays or wing cracks in terrestrial analogs, they
have been recognized in many locations on Europa such
as the impressive example at the southeast end of Agenor
Linea (Prockter et al., 2000; Kattenhorn, 2004a). Tailcracks
form in response to the locally perturbed stress field at the
tip of a fault and thus do not provide a direct indicator of
the global stress field at the time of their formation, except
where the tailcracks extend far enough from the fault tip
that their orientations are controlled by regional stresses.
(4) Endogenic fractures are most commonly associated with
Fig. 7. Troughs (indicated by white arrows). (a) Tectonic fractures refer to any generic troughs induced by global deformation of the
icy shell, such as due to tidal forcing (from Galileo mosaic 15ESREGMAP01). (b) Tailcrac ks form at the tip of a strike-slip fault (in
this case, Agenor Linea) where localized tension is induced by fault motion. One tailcrack here is also a cycloidal fracture (see Fig. 4).
From Galileo images 466664413 and 466665378, E17 orbit. (c) Endogenic fractures may be irregular and typically form adjacent to
chaos, as occurs here in Galileo image mosaic 11ESMORPHY01. (d) Fo ld h in g e fr a ct u re s are caused by tension along anticlinal fold
hinges. This example is within Astypalaea Linea, from Galileo image 466670113, E17 orbit. A fifth fracture type is a flexure fracture
as occurs to either side of Androgeos Linea (see Fig. 2b) in response to elastic bending of the icy shell.
Kattenhorn and Hurford: Tectonics of Europa 211
regions of chaos (Fig. 7c) and thus reflect a local process
likely driven by thermal upwelling or diapirism in a con-
vecting icy shell, which in turn disrupts the brittle carapace
(Collins et al., 2000; Figueredo and Greeley, 2004; Schenk
and Pappalardo, 2004; Mitri and Showman, 2008). These
fractures may also be associated with broad scale warping
at the surface of the icy shell (e.g., Fig. 3b in Prockter and
Papp al ar do, 2000), which is also likely driven by an en-
dogenic process. Thus, the orientations of endogenic frac-
tures are not controlled by a global stress field. (5) Flexure
fractures occur along the flanks of many large ridges (e.g.,
Fig. 2b) that have caused elastic flexing of the adjacent icy
shell, probably due to the loading caused by their weight
(Tufts et al ., 2000; Hurford et al., 2005) and/or by with-
drawal of material from beneath the ridge flanks (Head et
al., 1999). Flexing of the elastic portion of the icy shell
alongside a ridge induces bending stresses that can exceed
the tensile strength of the ice, at which point one or more
fractures develops (Billings and Kattenhorn, 2005). (6) Fold
hinge fractures (Fig. 7d) form in response to tensile bend-
ing stresses along the hinge lines of surface anticlines (see
section 2.2.1), although such features appear to be relatively
rare (Prockter and Pappalardo, 2000).
2.1.5. Normal faults. Although tensile fracturing is
dominant, some of the extension of the icy shell occurred by
normal faulting, evidenced by fault scarps that show a com-
ponent of vertical motion. Tensile stresses are common in
the shell (see section 4.1); however, normal faulting does not
occur under conditions of absolute tension, only deviatoric
tension. Normal faulting is possible when the minimum
horizontal compressive stress is less than the overburden
compressive stress by a sufficient amount to overcome the
internal friction (Beeman et al., 1988; Pappalardo and
Davis, 2007). Compressive stresses are common in the icy
shell during the tidal cycle, therefore adequate conditions
for extensional shear failure are likely to be common.
The great majority of normal faults are inferred to con-
stitute the fine striations within lineated bands (see sec-
tion 2.1.3) that parallel the band boundaries, contributing
to the similarity between dilational bands and terrestrial
mid-ocean ridge spreading centers. Rare normal faults with
significant vertical displacements (~300 m) have also been
noted to crosscut ridged plains (Nimmo and Schenk, 2006),
and it is possible that up to 1.5-km-deep troughs in the icy
shell that define antipodal small-circle depressions in the
leading and trailing hemispheres are also normal fault re-
lated (Schenk et al., 2008). In lineated bands, bilateral sym-
metry about a central trough is reminiscent of repeating
pairs of inward-dipping, graben-bounding faults that move
progressively outward from the spreading axis with time.
In high-resolution images (less than a few tens of meters
per pixel), the striations in the lineated bands are clearly
normal faults, showing smooth and highly reflective fault
planes bounding rotated blocks of band material with low-
albedo tilted upper surfaces. Mechanical interactions are
evident in fault trace patterns, as well as relay ramps in
overlap zones, analogous to terrestrial normal fault systems
(Kattenhorn, 2002). Normal faults in lineated bands do not
necessarily dip inward towards an axial trough but may dip
consistently in one direction across the entire width of a
band or be restricted to certain parts of the band only (Kat-
tenhorn, 2002; Prockter et al., 2002). This pattern of fault-
ing is distinctly different to mid-ocean ridge rift zones and
may imply that some dilational bands were extended by nor-
mal faulting at some time after initial band formation, per-
haps due to the reduced brittle thickness of the dilational
band relative to the surrounding ridged plains, causing the
bands to act as necking instabilities that focused extension.
It is unclear, however, why some dilational bands localize
normal faulting (forming faulted bands) rather than localiz-
ing tension fractures (in which case, ridged bands may ulti-
mately develop).
2.2. Compressive Tectonics
An apparent consequence of the great amount of exten-
sion on Europa is the need for corresponding contraction
to provide a balanced surface area budget. Although Europa
has likely undergone some amount of expansion due to
cooling and thickening of the icy shell, the maximum likely
extensional strain after even 100 m.y. of cooling would be
~0.35% (Nimmo, 2004a), which is insufficient to account
for the amount of new surface area created at spreading
bands, which occupy ~5% of the surface area in the pole-to-
pole regional maps of the leading and trailing hemispheres
(Figueredo and Greeley, 2004). Hence, some amount of
contraction must have occurred during the visible tectonic
history in order to create a balanced surface area budget.
Given this need, contractional features might be expected
to be as common on the surface as extensional features. In
actuality, contractional deformation on Europa is visibly
sparse and by no means obvious. Greenberg et al. (1999)
contemplated whether much of the contraction may have
been accommodated within areas of chaos terrain, which
covers ~10% of the surface; however, no ultimate geologi-
cal evidence arose to imply that chaos represents sites of
contraction.
Part of the reason for an apparent lack of contractional
features may be a historical emphasis on the many forms
of extensional structures such as ridges and dilational bands,
which are morphologically dominant; however, a potentially
greater hindrance to the identification of contraction has
been a failure to clearly recognize the manifestation of such
deformation. In terrestrial settings, contraction may be ac-
commodated variably over a range of scales. At the outcrop
scale, small amounts of contraction may occur along de-
formation bands (tabular zones of cataclasis and porosity
reduction in granular rocks) or along pressure-solution sur-
faces (sometimes called anticracks, where part of the rock
is removed in solution resulting in a volume loss). At the
regional scale, contraction is accommodated in brittle ma-
terials through the development of thrust faults (with asso-
ciated mountain building) or elastic warping, and in ductile
materials through folding. At the planetary scale, broad-
212 Europa
scale convergence occurs at subduction zones to accommo-
date the creation of new oceanic lithosphere at spreading
ridges. Hence, any contraction on Europa is likely mani-
fested through one or more of these mechanisms. There is
no evidence for global plate tectonics on Europa (see sec-
tion 3) and thus no removal of surface area along subduc-
tion zones; however, all other forms of contraction described
above may be viable on Europa.
2.2.1. Folding. The first documented evidence for con-
traction on Europa was the identification of several parallel
folds within a dilational band associated with the right-lat-
eral strike-slip fault Astypalaea Linea (Prockter and Pappa-
lardo, 2000). The folds are visible due to subtle differences
in surface brightness and have a wavelength of ~25 km and
crest-to-trough amplitudes of 250 ± 50 m (Dombard and
McKinnon, 2006). In adjacent ridged terrain, the folds grad-
ually disappear or are not resolvable, suggesting that di-
lational band material is more easily folded, perhaps due
to a localized high heat flow (Prockter, 2001). Other tell-
tale clues for the presence of the folds include clusters of
bending-induced tensile cracks along the hinge lines of an-
ticlines and compressive crenulations along the hinge lines
of synclines (Fig. 7d). The high heat flow in a dilating band
(see section 2.1.3) relative to adjacent ridged terrain, com-
bined with a reduced ice thickness (Billings and Kattenhorn,
2005), causes dilational bands to become localized zones
of crustal weakness that may be able accommodate con-
traction by folding. Prockter and Pappalardo (2000) sug-
gest that sufficient compressive stress may accumulate to
drive folding over <40° of nonsynchronous rotation of the
icy shell, and also suggest that fold axis orientations are
consistent with predicted stress fields. In contrast, Mével
and Mercier (2002) suggest that late transpressive motion
along Astypalaea Linea may have caused the localized fold-
ing. Over time, folds may relax away due to ductile flow
of deeper, warm ice; however, Dombard and McKinnon
(2006) suggest that such relaxation would occur so slowly
relative to the age of Europa’s surface (e.g., as little as 4%
relaxation over 100 m.y.) that all folds that formed during
the geologically visible past should still be apparent. Hence,
the distinct scarcity of visible folding on Europa suggests
that it is not a significant accommodator of crustal contrac-
tion and/or is difficult to recognize in existing images, and is
certainly insufficient to balance out the amount of extension.
2.2.2. Convergence bands. A second candidate feature
for localized contraction, first identified by Sarid et al.
(2002), is a convergence band, across which a tectonic re-
construction reveals a zone of “missing” crust. These bands
superficially resemble dilational bands caused by plate
spreading in that they form broad zones that may be many
kilometers wide and perhaps tens of kilometers long, and
appear to disrupt the surrounding ridged terrain (Greenberg,
2004) (Fig. 8). They differ from dilational bands, however,
in that their edges may be nonlinear or even inosculating,
and one edge of the band is not necessarily a mirror im-
age of the other side, unlike dilational bands that form by
icy shell separation and spreading. Convergence bands are
among the least-studied features on Europa; therefore, much
work is still needed to fully characterize their broad-scale
geometries, internal morphologies, and mechanical evolu-
tion. Nonetheless, existing studies imply two varieties of
convergence bands: (1) those driven by motion along strike-
slip faults, with resultant convergence adjacent to the fault
in the tip-region compressional quadrants; and (2) those that
develop along zones of preexisting weakness in the icy
shell, such as dilational bands and dilational strike-slip faults.
Sarid et al. (2002) describe two sites in the trailing hemi-
sphere where convergence is driven by motions along ad-
jacent strike-slip faults (type 1 above). One example in the
southern trailing hemisphere (in the Galileo regional im-
age mosaic 17ESREGMAP01), between Argadnel Regio
and Castalia Macula (Fig. 8a), is suggested to be a zone of
convergence related to right-lateral motion along an ap-
proximately north-south-oriented strike-slip fault north of
the zone of convergence. Patterson et al. (2006) describe
evidence of dilational band development in the same vicin-
ity. We infer that both processes may have occurred here
at different times (early convergence with later dilation).
Strike-slip fault driven convergence is suggested to have
occurred in the marginal blocks alongside Astypalaea Linea
(Mével and Mercier, 2002), resulting in up to 55% contrac-
tion along numerous distributed ridge-like crenulations (per-
haps a related contractional mechanism to convergence
bands). Evidence for strike-slip fault-related convergence
bands is also presented by Kattenhorn and Marshall (2006),
who characterize compressive stress concentrations in the
tip regions of strike-slip faults in Argadnel Regio (see sec-
tion 2.3.3).
Convergence bands that localize along sites of crustal
weakness such as preexisting dilational bands (type 2 above)
may also be driven by nearby strike-slip fault motions. This
phenomenon appears to have occurred in the northern trail-
ing hemisphere (Fig. 8b) example described by Sarid et al.
(2002), although they did not recognize it as such. Several
major strike-slip faults on Europa dilated during their de-
velopment to become band-like (see section 2.3.3) and may
have subsequently acted as loci for a component of con-
vergence if conditions changed from transtensional to trans-
pressional. Candidate site examples include the bright bands
Corick Linea, Katreus Linea, and Agenor Linea (see chap-
ter by Prockter and Patterson). Agenor Linea was interpreted
by Prockter et al. (2000) to contain contractional features
related to shear-related duplexing, suggesting transpression.
Greenberg (2004) inferred 20 km of convergence across
Corick Linea in order to match up offset features, resulting
in the formation of a 5-km-wide convergence band, imply-
ing 75% shortening.
It is unclear how inferred contraction is physically ac-
commodated within a convergence band. Sarid et al. (2002)
describe an internal fabric of numerous subtle parallel stria-
tions. Conceivably, these could be the traces of thrust faults.
In the 15ESREGMAP01 example (Fig. 8b), 8 km of con-
vergence is accommodated within a band that is only 2.7 km
wide, indicating 66% shortening. Even with a perhaps un-
Kattenhorn and Hurford: Tectonics of Europa 213
realistically small fault spacing of 100 m, each fault would
need to accommodate 300 m of horizontal shortening, re-
sulting in 170–300 m of vertical relief across each fault (de-
pending on the fault dip, assumed to be in the range 30°–
45°). These numbers would increase as the fault spacing
increases. The convergence bands do not appear to display
such rugged relief, perhaps casting doubt on thrust fault-
ing as a mechanism for contraction. A related complicat-
ing factor is the relatively large amount of differential stress
(up to 10 MPa by a depth of 3 km) required for thrust faults
to develop on Europa (Pappalardo and Davis, 2007), which
would hinder thrust fault development in the icy shell ex-
cept at very shallow depths (perhaps less than a few hun-
dred meters) due to the typical magnitudes of tidal stresses
Fig. 8. Two sites of potential contraction in the icy shell, manifested as convergence bands. (a) Possible convergence bands where
~15 km of convergence may have occurred in the equatorial trailing hemisphere, near Castalia Macula. The inferred convergence bands
resemble dilational bands but have irregular margins and do not have matching geology to either side of the band. This particular
example is superimposed with a later dilational band that passes across the center of the convergence band (Pat te rs on et a l. , 2006).
From Galileo mosaic 17ESREGMAP01. (b) An irregular convergence band (white arrowed zone between dashed lines) where ~8 km
of crust appears to be missing, resulting in mismatches of older, crosscut features (from Galileo mosaic 15ESREGMAP01).
214 Europa
(unless some other source of compressive stress exists).
Other possible contractional mechanisms include actual
volume loss within convergence bands due to remobili-
zation of warm ice toward deeper portions of the icy shell,
or compaction through porosity reduction (e.g., Aydin,
2006).
The documented large values of contraction at conver-
gence bands should make them easy to identify through dis-
ruption of the continuity of the ridged plains, except per-
haps where convergence occurs within preexisting dilational
bands and the convergence is less than the original dilation.
Convergence band-like features might be common in Eu-
ropa images but may have been previously misidentified as
dilational bands; therefore, they are potentially important
tectonic features for balancing Europa’s surface area bud-
get. Additional work will be needed to test this hypothesis.
2.2.3. Contraction across ridges. A third possibility
for sites of convergence on Europa is the ubiquitous double
ridges that dominate the tectonic fabric of the icy shell, first
suggested by Sullivan et al. (1997). The origin of ridges is
an equivocal source of great dispute (see section 2.1.1).
Some ridges show evidence for lateral offsets along them
but the majority of ridges do not. Nonetheless, recent mod-
els (Nimmo and Gaidos, 2002) hypothesize that ridge for-
mation may be driven by lateral shearing and frictional
heating along cracks. An analysis of ridges containing rela-
tively offset features (i.e., where relatively older lineaments
have undergone apparent lateral displacements along a
ridge) reveals that the offsets cannot be related to lateral
shearing alone (Patt ers on and Pap palardo, 2002; McBee et
al., 2003; Vet t e r, 2005; Patterson et al., 2006; Bader and
Kattenhorn, 2007; Kattenhorn et al., 2007). Although pure
dilation across any lineament (e.g., a dilational band) cannot
produce lateral offsets of features split apart by the dilation,
convergence across a lineament will result in the produc-
tion of apparent lateral offsets of all nonorthogonal cross-
cut features along the lineament that have nothing to do
with lateral shearing.
Aydin (2006) identified ridge-like contraction lineaments,
which he referred to as compaction bands or compactive
shear bands in the case where lateral shearing also occurred.
As used by Aydin (2006), the word “band” has no associa-
tion or morphological resemblance to either dilational bands
or convergence bands described earlier, but rather originates
from terrestrial analogs called deformation bands, across
which compaction can occur through the comminution of
granular materials. We therefore avoid use of “compaction
band” to describe these features, which morphologically
resemble ridges. The amount of apparent lateral offset along
a ridge is controlled by the amount of convergence as well
as the relative angle α measured clockwise from the ridge
toward any offset feature (Fig. 9). It is possible to differ-
entiate between apparent offsets caused by convergence and
those caused by true lateral shearing by examining the dis-
tribution of normalized separation as a function of α, where
separation is the orthogonal distance between a linear fea-
ture on one side of a ridge and the projection of that same
feature on the other side of the ridge (i.e., the conventional
structural geologic definition of separation). Separation is
uniquely related to α and the amount of convergence rela-
tive to the amount of true lateral shear motion (Ve tt e r, 2005;
Bader and Kattenhorn, 2008).
Apparent lateral offsets along ridges related to conver-
gent motions raise the possibility that ridge development
may, in some cases, be partially driven by contraction across
sheared lineaments. Such a notion is congruent with the
Nimmo and Gaidos (2002) frictional shear heating model
for ridge development. Such heating may result in a more
mobile wedge of ice alongside a shearing lineament, caus-
ing a localized zone of weakness in the icy shell where
contraction may be accommodated. In this model, part of
the loss of volume along the sheared lineament occurs by
remobilization of warm ductile ice into the deeper parts of
the shell or by localized melting and downward draining
within the frictionally heated zone. Part of the near-surface
convergence, however, is manifested through the construc-
tion of ridge edifices along the crack. The ridges could re-
sult from buoyant upwelling above the frictionally heated
zone, buckling, or porosity reduction in near-surface ices,
assuming an initially high enough porosity to allow volu-
metric compaction. Thus, the possibility exists that this
is a plausible mechanism for ridge development, particu-
larly because it is likely that all active lineaments on Europa
are subject to shearing and thus heating during the euro-
pan day in response to the constantly rotating diurnal tidal
stresses (see section 2.3.4). Those ridges that have been
shown to exhibit a component of contraction across them
invariably show a larger amount of strike-slip motion ac-
companying the contraction (Ve tt e r, 2005), providing a
strong argument that the contraction is related to the proc-
ess of shearing. The implication of this mechanism for ridge
development is that the majority of the elusive contraction
in Europa’s icy shell may, in fact, be accommodated across
its most common type of feature (i.e., ridges), with only
a minimal amount of convergence needed across any one
ridge to allow for a balanced surface area budget. Unfortu-
nately, other than a few prominent ridge examples, this
hypothesis has not been convincingly demonstrated, plau-
sibly because apparent offsets along ridges are below ex-
isting resolution limits, perhaps due to minimal amounts of
convergence and/or lateral offsets. Therefore, any inferences
about ridges as important accommodators of convergence
remain mostly conceptual for the time being. An added
complication is that many sheared ridges have been shown
to exhibit evidence of dilation using the same technique (α
vs. separation graphs) described above (Bader and Katten-
horn, 2008), indicating that ridges may be subject to a com-
bination of dilational, convergent, and shearing motions dur-
ing their history.
2.3. Lateral Shearing
2.3.1. Shear failure of ice. Lateral shearing refers to
strike-slip motions along lineaments. Voyager images indi-
Kattenhorn and Hurford: Tectonics of Europa 215
cated the presence of lateral offsets in the europan icy shell
at scales of tens of kilometers (Schenk and McKinnon,
1989). On Earth, such motions may occur for two reasons:
(1) through primary shear failure of rock in a stress field
where the intermediate compressive principal stress (σ2) is
vertical and the horizontal differential stress (σ1 – σ3) ex-
ceeds the rock frictional strength; and (2) through reacti-
vation of existing structures (faults and fractures) in re-
sponse to a temporal change in the stress field relative to
when the existing structures first formed. In the first case,
approximately vertical strike-slip faults are produced on
Earth and include transform plate boundary faults (such as
the San Andreas fault) as well as intraplate faults. In the
second case, any of the principal stresses could be vertical,
resulting in oblique-slip motion in cases where the faults
are not vertical.
On Europa, the second of these situations most certainly
exists. The stress field changes constantly due to variations
Fig. 9. See Plate 12. (a) Analytical curves can be used to differentiate the relative amounts of strike-slip motion (whether right-
lateral or left-lateral) and contraction across a ridge/strike-slip fault based on the amount of separation between two halves of an ap-
parently offset feature (inset, upper left). α is the angle measured from the ridge/strike-slip fault to the offset feature in a clockwise
sense. A separate set of curves exists for the case of dilation plus strike-slip motion (Ve t t er, 2005). (b) Illustration of various configu-
rations of mismatches of crosscut features in response to pure right-lateral motion, convergence only, or a combination (motion is
along the thicker black line, representing a ridge or strike-slip fault). If α is 90°, offsets only occur if there is lateral motion. Pure
convergence can produce a mixture of left- and right-lateral offsets, depending on α.
216 Europa
in the tidal figure during each orbit (see section 4.1); there-
fore, any active lineament on Europa necessarily experi-
ences shear stresses that could induce lateral shearing. If a
crack is dilated during shearing (i.e., the normal stress is
tensile), there is no frictional constraint on shear motion.
For the case of sliding of a closed crack, the limiting factor
is the coefficient of friction, µ, based on the Coulomb fail-
ure criterion: τ = S0 + µσn at failure, where τ is shear stress,
σn is normal stress, and S
0 is the inherent shear strength
(cohesion) in the absence of normal stress.
The value of µ for europan ice is unknown but is almost
certainly less than occurs in rocky materials and is com-
monly approximated on the basis of comparisons to terres-
trial observations and laboratory experiments (Beeman et
al., 1988; Rist and Murrell, 1994; Rist, 1997; Schulson,
2009). Complications exist in that µ is likely affected by
ice chemistry, sliding speed, and temperature, with the fric-
tion increasing as temperature decreases. Based on low-tem-
perature (77–115 K) experiments on sawcut pure water ice,
Beeman et al. (1988) provide perhaps the most appropri-
ate estimate of µ applicable to europan conditions in the
upper part of the icy shell, but even so, µ is variable de-
pending on loading conditions. At low confining pressures
(P ≤ 5 MPa; or down to a depth of ~4 km on Europa), µ =
0.55 and the inherent shear strength, S0 = 1 MPa. Where
P ≥ 10 MPa (deeper than ~8 km), µ = 0.2 and the inherent
shear strength, S0 = 8.3 MPa. Potential caveats to the labo-
ratory technique include the 2–5 orders-of-magnitude-faster
sliding velocities used than may occur on europan faults,
which underestimates the friction coefficient (e.g., Kennedy
et al., 2000; Fortt and Schulson, 2004); a time-dependency
to crack strength, increasing with the number of slip events
or due to development of fault gouge; the usage of pure
water ice; the possibility that frictionally heated ice along a
shearing crack may experience a decreased friction coeffi-
cient (as has been suggested to occur on Enceladus) (Smith-
Konter and Pappalardo, 2008); and the assumption of a pre-
existing fault.
The assumption of a preexisting fault is only problem-
atic in attempting to make predictions about the orientations
of primary shear fractures relative to the principal stresses.
In laboratory experiments on water ice, this angle is com-
monly reported to be ~45° (Durham et al., 1983; Sammonds
et al., 1989; Rist and Murrell, 1994; Rist et al., 1994; Rist,
1997). Considering that the angle, θ, between a develop-
ing shear fracture and the maximum compressive princi-
pal stress, σ1, is given by θ = 0.5 tan–1 (1/µ), an angle of
45° implies µ = 0, which disagrees with deduced values of
µ described above. Schulson et al. (1999) suggest that the
45° shear fracture angles were an artifact of the loading
apparatus. They produced shear fractures in ice where θ =
30° (i.e., µ ≈ 0.6), more congruous with the low confining
stress results of Beeman et al. (1988). If low sliding veloci-
ties (10–6–10–7 m s–1) (Nimmo and Gaidos, 2002) occur on
Europa, µ may actually be in the range 0.6–0.8 (Kennedy
et al., 2000), creating shear fractures at θ = 25°–30° to the
orientation of σ1, unless the ice is sufficiently warm (e.g.,
due to shear heating) that µ is reduced.
2.3.2. Lateral shear failure on Europa. Normal faults
have already been shown to be a common form of exten-
sional deformation (see section 2.1.5), therefore lateral
shearing would tend to produce oblique-slip motions along
them. Kattenhorn (2002) describes evidence of en echelon
breakdown zones at the tips of normal faults within dila-
tional bands, implying oblique motions during fault tip
propagation. Nonetheless, vertical fractures predominate on
Europa due to tensile failure perpendicular to the horizontal
maximum extension direction. As a result, troughs, ridges,
and dilational band margins are all likely to be vertical and
prone to reactivation through lateral shearing as long as the
fractures have not yet healed. Strike-slip motions along these
features thus also fall into category 2 faults described in sec-
tion 2.3.1. Whether or not category 1 strike-slip faults exist
on Europa (i.e., due to primary shear failure) is still unclear,
especially seeing as they may be impossible to decipher
from the myriad other vertical cracks that underwent lateral
shearing due to reactivation. Nonetheless, primary shear
failure is ultimately dependent on whether or not the stress
conditions favor strike-slip fault formation, exactly as it
would on Earth. It should also be noted that lateral shear-
ing has been noted in terrestrial ice shelfs (Wilson, 1960)
and glaciers (e.g., during the rupture of the Denali fault in
Alaska in 2002), but typically takes on the form of en ech-
elon fracture arrays, with individual fractures oriented ob-
liquely to the trend of the fault zone (Haeussler et al.,
2004). Very few analogous en echelon crack geometries
have been described in europan strike-slip fault zones (e.g.,
Prockter et al., 2000; Michalski and Greeley, 2002), and
appear to be more commonplace within shear-reactivated
dilational bands, perhaps implying that strike-slip motions
typically reactivate existing cracks that initially formed in
tension (cf. Greenberg et al., 1998; Tufts et al., 1999). In
some instances, shearing across regions of closely spaced
tension cracks may result in fragmentation along the de-
veloping shear zone (Aydi n, 2006).
Any evaluation of the formation mechanisms for strike-
slip faults on Europa must be placed within the context of
the mechanics of shear failure of ice and must incorporate
a candid analysis of the pros and cons of shear failure vs.
tensile failure interpretations for europan lineaments show-
ing apparent offsets (e.g., Kattenhorn, 2004b). For example,
Spaun et al. (2003) suggest that northeast- and northwest-
oriented ridges with strike-slip offsets in the equatorial trail-
ing hemisphere formed as primary shear fractures because
they form X-shaped patterns that superficially resemble
conjugate shear sets, and because of their locations rela-
tive to expected nonsynchronous rotation stresses. Although
the X-shapes are reminiscent of conjugate shear fractures,
shear offsets are either absent or inconsistent along ridges
of a particular orientation and no explicit crosscutting re-
lationships exist, probably implying that the ridges are dis-
tinctly different in age. Also, potentially conjugate ridge sets
Kattenhorn and Hurford: Tectonics of Europa 217
(i.e., X-shapes for ridges of similar stratigraphic age) show
no consistent conjugate angle, 2θ, between them. This
would not be expected for true conjugate sets as 2θ is ex-
plicitly controlled by the coefficient of friction of the ice,
µ (see section 2.3.1), which should be somewhat consis-
tent. Hence, there is currently no convincing geological evi-
dence that primary shear failure of the icy shell, analogous
to strike-slip faulting, produced global lineaments. None-
theless, given that extensional shear failure is likely respon-
sible for normal faults in the icy shell, the possibility that
some lineaments formed through strike-slip shear fractur-
ing cannot be dismissed.
2.3.3. Strike-slip fault morphologies. Lateral offsets
can potentially occur along all lineament types (i.e., troughs,
ridges, cycloids, and dilational and convergent bands), re-
gardless of how they initially formed. Nonetheless, cumu-
lative offsets that are large enough to be resolvable in Gali-
leo images (generally hundreds of meters or more) typically
occur along ridges and dilational bands. Accordingly, Kat-
tenhorn (2004a) distinguishes two predominant types of
strike-slip faults on Europa: ridge-like and band-like, based
on their morphologic similarity to ridges and dilational
bands (Fig. 10). The mechanical evolution of the two types
of strike-slip faults is distinctly different. Ridges likely ini-
tiated as troughs (see section 2.1.1); therefore, any lateral
shear motions along them probably happened later, poten-
tially contributing to the process of ridge development. In
the case of dilational bands (see section 2.1.3), the ques-
tion arises as to whether strike-slip offsets along them hap-
pen prior, during, or after dilation. Strike-slip offsets across
dilational bands are evidenced by the need for oblique clos-
ing to reconstruct older features affected by the dilation.
Sigmoidal lineations within dilational band material may
imply oblique dilation (i.e., concurrent dilation and strike-
slip motion), as occurs along Astypalaea Linea; however,
the timing of strike-slip motions along dilational bands
cannot necessarily be determined based on band morphol-
ogy alone (Prockter et al., 2002).
Kattenhorn (2004a) and Kattenhorn and Marshall (2006)
suggest that secondary cracks at fault tips, called tailcracks,
provide insights into the mechanics of ridge-like vs. band-
like strike-slip fault development. Tailcracks are secondary
tension fractures commonly observed at the tips of strike-
slip faults on Earth, and which have also been documented
on Europa (see sections 2.1.4 and 4.2). The intersection of
the fault and its associated tailcrack is manifested by a sharp
kink, with an angle described by linear elastic fracture me-
chanics theory as being controlled by the ratio of shear
stress to normal stress at the instant of tailcrack development.
Hence, it is possible to determine whether or not a fault is
dilating at the instant it is shearing laterally based on the
geometry of its tailcracks. Using this line of reasoning, Kat-
Fig. 10. Examples of strike-slip faults. (a) Unnamed right-lateral, ridge-like strike-slip fault cutting Yelland Linea in Argadnel Regio,
from Galileo mosaic 12ESWEDGE_03. (b) A portion of the band-like strike-slip fault Astypalaea Linea in the southern antijovian
region. White solid lines are fault segments that slipped right-laterally, producing dilational pull-aparts along linking cracks (white
dashed lines). A ridge that was offset by 77 km (gray) illustrates the fault kinematics. From Galileo mosaic 17ESSTRSLP02.
218 Europa
tenhorn (2004a) showed that ridge-like faults (Fig. 10a)
commonly undergo lateral shearing while slightly dilated
or frictionally closed (see also Aydin , 2006), whereas the
band-like faults they examined (Fig. 10b) accrue strike-slip
offsets concomitant with significant dilation. Hence, many
band-like strike-slip faults are essentially transtensional,
oblique spreading zones. They exhibit typical lateral offsets
of tens of kilometers (e.g., ~77 km along Astypalaea Linea)
(Kattenhorn, 2004a) and lengths of up to ~1500 km (e.g.,
Agenor Linea) (Prockter et al., 2000).
Despite these interpretations, band-like strike-slip faults
should not be considered to be simply types of dilational
bands, as some strike-slip motions almost certainly occur
after dilation has ceased and may even be associated with
some amount of later contraction. For example, the bright
bands originally described from Voyager imagery, such as
Agenor Linea (Fig. 7b) (Schenk and McKinnon, 1989), may
reflect geologically recent reactivation of older dilational
bands to create transpressional strike-slip faults. The interior
structure of Agenor Linea is convoluted (unlike simple dila-
tional bands) with some portions being reminiscent of trans-
pressional duplexing along terrestrial fault zones (Prockter
et al., 2000). This duplexing probably occurred during
shearing of the original band, with restraining bends devel-
oping because of the irregular geometry of the band mar-
gins. Tectonic annihilation of the margins of Agenor Linea
during shearing may have resulted in ridged plains that
cannot be matched from one side of the fault zone to the
other, unlike dilational bands.
We i nfer t hat so me b an d-li ke s trik e- slip fa ul ts a re e ss en-
tially ridge-like/band-like hybrids, in that the band portions
of the fault formed as dilational pull-aparts along segmented
ridge-like faults, where the sense of step between adjacent
fault segments was the same as the sense of slip along them.
A prominent example is the right-lateral fault Astypalaea
Linea in the south-polar region (Tufts et al., 1999; Katten-
horn, 2004a), where rhomboidal pull-apart bands formed
between right-stepping ridge-like fault segments (e.g., Cy-
clades Macula). As a result of this formation mechanism,
the interior striations of these bands (which trend perpen-
dicular to the inferred spreading direction) are highly ob-
lique to the overall trend of Astypalaea Linea.
As with distinctions between dilational bands and band-
like faults, ridge-like strike-slip faults should not be con-
sidered to be simply types of ridges. Crosscut features that
can be matched to either side of ridges (producing so-called
piercing points that used to be together) typically show zero
lateral offsets, indicating that measurable cumulative lateral
offsets are not a requirement for ridge development. None-
theless, some ridges do show appreciable lateral offsets,
implying a long-term process of shearing and strike-slip
accumulation, whereas troughs never show lateral offsets
(Hoppa et al., 1999c), suggesting that lateral motions may
be an important aspect of the ridge building process. Ridge-
like fault offsets are typically in the range of hundreds of
meters to several kilometers (Hoppa et al., 2000) but have
been measured as high as 83 km (Sarid et al., 2002). An
important observation is that the lateral offset along a ridge-
like fault does not scale with the length of the ridge, as
would be true of terrestrial strike-slip faults and theoreti-
cal predictions for shear fractures based on linear elastic
fracture mechanics (Pollard and Segall, 1987). For exam-
ple, Agave Linea, which passes north of Conamara Chaos
(Fig. 2a), is at least 2000 km long, yet shows lateral off-
sets of ~5 km, implying that it did not initiate or grow as a
primary shear fracture but rather accrued a minor amount
of lateral offset during subsequent movement. Therefore,
models that favor ridge formation through primary shear
failure of the icy shell do not seem justified in these cases.
Nonetheless, frictional shearing may very well contribute
to the ridge developmental process (Nimmo and Gaidos,
2002) (see section 2.1.1). Accordingly, tailcrack geometries
along ridge-like faults confirm that frictional sliding of
closed cracks commonly did occur to form these features
(Kattenhorn, 2004a). The caveat to this model is that not
all ridges show lateral offsets, implying two possible sce-
narios: (1) repeated forward and backward frictional shear
motion along cracks during the diurnal cycle resulted in
zero or unresolvable cumulative strike-slip offset but still
produced enough total heat to drive ridge construction; or
(2) frictional shearing is just one of perhaps several con-
tributing factors to ridge development (see section 2.1.1),
such that the absence of strike-slip motions does not pre-
clude ridge formation. Neither scenario can be proven; how-
ever, the creation of apparent lateral offsets through a com-
ponent of convergence, if present (see section 2.2.3),
implies other processes may also contribute to ridge devel-
opment. The upshot is that strike-slip motions are relatively
common on Europa and that the lateral motions commonly
utilize ridges and dilational bands, which must therefore act
as planar weaknesses in the icy shell along which lateral
shearing is accommodated.
2.3.4. Tidal walking. True strike-slip motions along
faults are suggested to be driven by diurnal tidal stresses
through a process informally referred to as tidal walking
(Hoppa et al., 1999c). In response to the constantly chang-
ing diurnal tidal stress field (see section 4.1.5), faults re-
peatedly experience tension then compression out of phase
with left- and right-lateral shear stresses, because of the
manner in which the tidal stress tensor resolves normal and
shear stresses onto the fault surfaces. As with any fault or
fracture, certain failure criteria rules apply. Where tension
opens a crack, there would be no frictional resistance to
shear motion; therefore, concomitant shearing should pro-
duce lateral offset. If a crack is closed, shear motion is lim-
ited by the frictional strength; therefore, lateral offsets can
only be produced if the shear stress exceeds the normal
stress multiplied by the coefficient of static friction, µ (i.e.,
the Coulomb failure criterion). Hence, the tidal walking
theory implies that when the stress normal to a fault is ten-
sile, the fault opens at the surface, allowing shear stress to
produce a small amount of offset. The net sense of shear
during this dilational phase, which depends on the extent
to which the normal and shear stress curves are out of phase
with each other (Hoppa et al., 1999c; Groenleer and Kat-
tenhorn, 2008), controls the sense of strike-slip offset. As
Kattenhorn and Hurford: Tectonics of Europa 219
the normal stress acting across the fault changes from ten-
sion to compression about half an orbit later, the fault sur-
faces must be in frictional contact. The friction along the
fault then limits the ability of a changing sense of shear
stress to completely remove the recently accrued offset.
Consequently, after each diurnal cycle, the fault accumu-
lates a small net sense of strike-slip offset along its length.
Over many successive cycles of this process, the accumu-
lation of a multitude of small strike-slip offsets produces a
visible amount of strike-slip displacement.
The sense (left- or right-lateral) of strike-slip displace-
ment along a fault depends on the orientation and location
of the fault (Fig. 11). Poleward of 45° latitude, regardless
of fault orientation, tidal walking must result in only left-
lateral faults in the northern hemisphere and right-lateral
faults in the southern hemisphere. A mixture of left- and
right-lateral displacements is predicted in the midlatitudes,
depending on fault orientation. Observations of many strike-
slip fault offsets appear to support this theoretical expecta-
tion (Hoppa et al., 2000). In the leading and trailing hemi-
sphere Galileo regional mapping observations, Sarid et al.
(2002) observed left-lateral faults in the high northern lati-
tudes and right-lateral faults in the high southern latitudes;
however, the transition to orientation-controlled offset sense
was not observed to occur at the predicted 45° latitude.
They posited that a small amount of polar wander may be
responsible for the mismatch between tidal walking theory
and their strike-slip fault observations.
Tidal walking provides a simple mechanism for driving
strike-slip motions on Europa, with a similar phenomenon
now suggested to be driving fault activity in the south po-
lar region of Enceladus (Hurford et al., 2007b; Smith-Konter
and Pappalardo, 2008), possibly responsible for plume
eruptions. Nonetheless, while band-like faults show consis-
tent agreement with the predictions of tidal walking, there
are exceptions for ridge-like faults (such as left-lateral ridge-
like faults in the south-polar region) (Kattenhorn, 2004a).
A potential caveat is that the theory does not incorporate
the longer-term effects of nonsynchronous rotation stresses
on strike-slip motions, which are expected to be important.
For example, Schulson (2002) uses a frictional sliding cal-
culation to infer that a significant amount of nonsynchro-
nous rotation stress buildup would be needed to drive mo-
tion along a strike-slip fault with sufficient force to allow a
developing tailcrack to penetrate down to a depth of 1 km
into the icy shell. Considering that some tailcracks have de-
veloped into dilational bands, and thus must have penetrated
sufficiently deep to reach viscoplastically deformable ice,
nonsynchronous rotation stresses may indeed have been an
important contributor to this process. A final caveat is that
the tidal walking theory is currently conceptual, as it does
not incorporate a quantitative test of shear motions using
frictional failure criteria. In fact, the recent work involving
tidal stresses acting on faults on Enceladus has produced a
conceptual model for slip along faults that is based on the
mechanics of strike-slip faults on Earth (Smith-Konter and
Papp al ard o, 2008). This newer model of tidally driven slip
along faults might still be considered a form of tidal walking
but its predictions of the sense of slip may be different in de-
tail from the Hoppa et al. (1999c) model shown in Fig. 11.
2.3.5. Lateral shearing and cycloid growth. Although
cycloids are probably tension fractures (see section 2.1.2),
Fig. 11. Global pattern of strike-slip motion sense of faults of all orientations as a result of diurnal tidal stresses and the process of
tidal walking. Fault orientations are indicated by the circular rose diagrams where the top of each circle is the 0° or north-south ori-
entation. There is a prevalence of left-lateral motions (black) in the northern hemisphere and right-lateral motion (white) in the south-
ern hemisphere.
220 Europa
lateral shearing is likely to play an important part in their
developmental mechanics. Marshall and Kattenhorn (2005)
highlighted geometrical similarities between cycloid cusps
and tailcracks and developed a conceptual model that as-
cribed cusp formation to lateral shearing and tailcrack ini-
tiation in the tip region of a recently formed cycloid arc in
response to the rotating diurnal stresses. The tailcrack then
continues to grow in tension, forming the next cycloid arc
in the chain. This model was bolstered by the demonstra-
tion of an excellent match between cusp angles and theo-
retical tailcrack angles predicted using linear elastic frac-
ture mechanics equations and calculated tidal stress fields
(Groenleer and Kattenhorn, 2008). The point in the orbit
at which cusp development and new arc initiation occurs
is therefore controlled by the interplay between resolved
shear stress and normal stress in the tip region of the pre-
viously formed arc, not the timing of the maximum tensile
principal stress during the orbit. This point in the orbit
changes for each cycloid arc, depending on the orientation
of the end region of the most recently formed arc.
3. LATERAL MOTIONS WITHOUT
GLOBAL PLATE TECTONICS
Plate tectonics, in the terrestrial sense, implies that the
entire brittle outer layer (lithosphere) is broken into numer-
ous discrete fragments (plates) that move across the surface
relative to each other in response to a combination of three
types of motions along their boundaries: divergent, conver-
gent, and transform. Europa contains a highly fractured
outer layer (cryosphere), raising the issue of whether it too is
broken into discrete plates with discernable boundaries. The
existence of dilational bands on Europa (see section 2.1.3)
indicates that new surface area was created through the pro-
cess of complete separation of the icy shell and concurrent
filling of the resultant lithospheric gap by material from be-
low, analogous to mid-ocean ridge spreading centers. How-
ever, these tabular dilational gashes in the icy shell tend to
vary in width along their lengths and may have discrete
ends, or tips, indicating that they do not facilitate rigid mo-
tions of the two opposing sides away from the spreading
site as typifies tectonic plates on Earth. Terrestrial plate tec-
tonics has created a balance between surface area creation
at mid-ocean ridges and surface area removal at subduction
zones. In contrast, there are no known subduction zones on
Europa (Sullivan et al., 1998; Prockter et al., 2002) and no
obvious global pattern of tectonic plate boundaries that
localize deformation. Rather, brittle deformation of the icy
shell is globally pervasive, although the intensity of defor-
mation may be localized (e.g., at dilational bands and ma-
jor strike-slip faults). The locations of tensile fracturing or
lateral shearing at any point in time is likely driven by the
characteristics of the global tidal stress field, which is geo-
graphically variable (see section 4). However, any long-term
reorientation of the icy shell, such as by nonsynchronous
rotation (section 4.1.6) or polar wander (section 4.1.7), could
move all locations on the moon’s surface into optimal stress
zones for brittle deformation at some point, allowing for the
global pervasiveness of brittle fracturing. The upshot is that
there is no indication of a terrestrial-like, global plate tec-
tonic system on Europa. Accordingly, there is no known (or
likely) driving mechanism for global plate tectonics either.
The problem remains that the repeated creation of tens-
of-kilometers-wide dilational gashes in the icy shell must
be somehow accommodated without global plate tectonics.
Localized contractional features apparently do exist on
Europa to accommodate some of the spreading occurring
at dilational bands (see section 2.2), including convergence
bands, ridges, and to a much lesser extent, folds. It is also
possible that much of the opening across a band is accom-
modated elastically in the icy shell at timescales less than
the Maxwell relaxation time. In a linear elastic half-space, a
tension fracture causes significant elastic deformation within
the material within which it is embedded out to a distance
that scales with the smallest dimension of the crack (e.g.,
Pollard and Segall, 1987). On Europa, this dimension is
likely to be the fracture depth for the case of fractures that
do not fully penetrate the icy shell. For a very long tension
fracture or dilational band, elastic deformation will mostly
dissipate within about three crack depths perpendicularly
away from the structure, which should be in the range of a
few kilometers to a few tens of kilometers. If the crack fully
penetrates the shell (a situation that has been proposed but
never confirmed on Europa), elastic deformations will ex-
tend away from the crack out to a distance that scales with
the crack length, as long as this length exceeds the shell
thickness (also see Sandwell et al., 2004). Hence, shell-
penetrating cracks that extend hundreds of kilometers across
the surface would cause elastic deformations in the adja-
cent icy shell out to distances of hundreds of kilometers.
The point here is that dilations could conceivably be ac-
commodated by elastic contractional strains alongside a
dilating crack without the need for visible contractional
deformation. Because dilational bands maintain a perma-
nent dilation, the elastic deformation of the icy shell can
only be dissipated by long-term viscous relaxation within
the icy shell. Considering that the relaxation time for the
cold outer brittle portion of the icy shell (~80 G.y.) (Nimmo,
2004a) exceeds the age of the solar system, ongoing open-
ing of a dilational band should ultimately cause stored elas-
tic energy within any elastically perturbed zone around the
dilational band to be released through brittle deformation,
such as shear failure (e.g., strike-slip faulting) or conver-
gence along existing structures. Hence, some motions along
structures may be driven by locally perturbed stresses along-
side dilating bands rather than global tidal stresses.
Ter re str ia l strike-s lip f aults can be isola ted or trans curr ent
intraplate faults or transform plate boundary faults. Isolated
faults have distinct tips at the end of the fault trace, toward
which the lateral offset decreases to zero from some maxi-
mum along the fault trace (commonly near the geometric
center of the fault). In contrast, transcurrent and transform
faults may be unconstrained by fault tips as a result of the
presence of deformation belts, spreading centers, or sub-
duction zones at either tip, allowing the fault to behave
analogous to a rigid-body translation. In these cases, the
Kattenhorn and Hurford: Tectonics of Europa 221
total offset along the fault may be approximately constant
along the fault length. In the absence of global plate tec-
tonics on Europa, one might expect only nontransform
faults; however, palinspastic reconstruction of fault offsets
can commonly be achieved via a rigid-body technique (e.g.,
Astypalaea Linea) (Tufts et al., 1999; Kattenhorn, 2004a).
This is possible because even in the absence of plate tec-
tonics, fault-tip motions can potentially be accommodated
by dilational bands where extension is needed and by con-
vergence bands where contraction is needed (Sarid et al.,
2002; Schulson, 2002; Kattenhorn and Marshall, 2006).
Nonetheless, it is not always possible to match up features
across a palinspastically restored strike-slip fault using rigid
translations because of variability in the amount of strike-
slip displacement associated with elastic displacement
gradients (e.g., Ve tt e r, 2005). This behavior may suggest that
isolated strike-slip faults exist on Europa with slip gradi-
ents along their lengths and should not be treated as rigid-
body transforms in tectonic reconstructions.
Although the preceding discussion highlights the lack of
global plate tectonics on Europa, rigid translations and rota-
tions appear to have occurred locally along circumferen-
tially detached portions of the icy shell (i.e., areas that are
completely surrounded by prominent, active cracks with
variable motion behaviors, analogous to plate boundary
processes). These localized plate-like fragments of the icy
shell have been referred to as microplates (Schenk and Mc-
Kinnon, 1989; Rothery, 1992; Sullivan et al., 1998; Sarid et
al., 2002) and sometimes form triple junctions where three
microplates meet (Patterson and Head, 2007). The exist-
ence of these mostly intact portions of the icy shell that
undergo lateral shifts or rotations justifies local usage of
plate tectonic analysis techniques, similar to terrestrial ana-
logs, despite the lack of a global plate tectonic system. Ac-
cordingly, Pa tt erso n et a l. (2006) examined fault- and dila-
tional band-bounded microplates, tens to hundreds of kilo-
meters wide, in the equatorial region near Castalia Macula,
northwest of Argadnel Regio, and were able to determine
Euler poles of rotation. Nonetheless, the nonuniqueness of
derived Euler poles using several adjacent microplates and
a deduced deformation sequence implies nonrigid deforma-
tion within the microplates. In the absence of a known driv-
ing mechanism for any type of large-scale plate motion on
Europa, small rotations of inferred microplates are osten-
sibly driven by differential motions induced by local tec-
tonic activity along the microplate boundary (such as the
opening of a dilational band), or in adjacent deforming
regions. This process may be particularly common where
the icy shell is heavily dissected by multiply oriented dila-
tional bands (such as in Argadnel Regio) because dilational
bands potentially form zones of weakness within the icy
shell across which later deformation may be localized.
4. CAUSES OF TECTONIC DEFORMATION
Fractures on Europa are produced in response to stress
in the icy shell. As described in previous sections of this
chapter, a diverse range of tectonic features has been pro-
duced in response to these stresses, causing initial brittle
failure of the ice and subsequent modification of these struc-
tures over time. Thus, the tectonic record preserves a record
of the stresses experienced by the icy shell and holds the
key to identifying the processes that imparted those stresses.
We now highlight the processes that can create the stresses
responsible for fractures in the europan shell at global, re-
gional, and local scales (also see chapters by Nimmo and
Manga, Sotin et al., and Bills et al.).
4.1. Global-Scale Stress
Some global-scale ridges span more than 50% of the
circumference of Europa, so the stress conditions needed
to form the fractures along which these ridges formed must
have been global in scale. We therefore examine the range
of plausible factors that may induce global-scale stress
fields, which, at their essence, require a global change in
shape of the satellite.
4.1.1. Thin shell approximation. Using Mercury as an
example, Melosh (1977) showed that a global change in
shape caused by a change in rotation rate produces stress
that could drive the formation of tectonic features. In its
simplest form, the model approximates the surface stress
as occurring in a thin elastic shell that is decoupled from a
fluid interior as it deforms. The horizontal strain in the shell
induced by the tidally distorted interior results in stress on
the surface, given by
σθθ = – 1
3
1 + ν
5 + ν
µf (5 + 3 cos 2θ)
(1)
and
σ
φφ
= 1
3
1 + ν
5 + ν
µf (1 – 9 cos 2θ)
(2)
The quantity σθθ is the stress along the surface in the
direction radial to the axis of symmetry, while σϕϕ is the
stress along the surface in a direction orthogonal to the σθθ
stress. Also, θ is the angular distance to any point on the
surface measured with respect to the tidal distortion’s axis
of symmetry, f is the flattening, µ is the rigidity (shear mod-
ulus) of the shell, and ν is Poisson’s ratio. Compressive
stresses are defined here to be positive and tensional stresses
negative. Flattening, f, is defined as f = (rsym – rorth)/r where r
is the mean radius of the body, rsym is the radius along the
axis of deformation symmetry, and ror th is the radius along
an orthogonal axis. The flattening is positive when rsym >
rorth, otherwise it is negative.
The thin shell approximation of surface stress is also ap-
plicable to Europa, where conservative estimates of tidal
heating predict a H2O layer that might be liquid underneath
an icy shell (Pea le a nd C asse n, 1978; Moore, 2006). More-
over, magnetometer measurements from Galileo strongly
imply the existence of a liquid layer (Kivelson et al., 2000).
222 Europa
A liquid ocean would decouple the icy shell from the inte-
rior of Europa. Although the thickness of the icy shell is
not known, a range of techniques have constrained it to be
probably less than 30 km (Billings and Kattenhorn, 2005,
and references therein; see chapter by McKinnon et al.). In
the absence of a good physical constraint on the thickness
of the icy shell, the assumption that it behaves as a thin
elastic layer is assumed to be valid. If this icy shell is on the
order of 10 km thick, the stresses on its surface may still
approximate a thin elastic shell (Hurford, 2005). A thicker
icy shell is expected to have its upper portion acting elas-
tically (Willi ams and Greeley, 1998; Billings and Katten-
horn, 2005; Hurford et al., 2005) with deeper, warmer ice
behaving viscously at low strain rates (Rudolph and Manga,
2009). Even so, the entire icy shell may still behave in a
manner analogous to a thin elastic layer if deformed rap-
idly, such as at the timescale of diurnal stressing (Wahr et
al., 2009).
Although we describe the results of a thin shell approxi-
mation for the tidal stress, an alternative model that can
account for thicker icy shells has been recently developed
(Wahr et al., 2009). Models of tidal stress for arbitrary icy
shell thicknesses are based on standard techniques used to
calculate tides on Earth (Dahlen, 1976). These models de-
fine the stress in terms of the material properties of the
satellite at it surface and a description of the satellite’s tidal
deformation, which is given by tidal Love numbers h2 and
l2. The models also assume that the satellite is composed
of distinct layers, each defined by its own material proper-
ties: density ρ, viscosity η, rigidity µ, and compressibility
λ. Measurements of the moment of inertia allow some con-
straints on the distribution of mass within the body (Ander-
son et al., 1997). However, the structure of Europa’s inte-
rior is poorly constrained.
4.1.2. Despinning. After the Voyager flybys, the pat-
tern of global ridges on Europa (Fig. 1) was described as
consisting of radial and concentric fractures resembling ten-
sion cracks near the sub- and antijovian points, while else-
where on the surface, 60° intersections of fractures were
interpreted as conjugate shear fractures (Helfenstein and
Parm en ti er, 1980). This pattern suggested that cracks were
produced by a change in shape of Europa by a tidal proc-
ess and not by a change in shape controlled by the rotation
of Europa. Early analysis favored a change in shape induced
by a tidal response driven by Europa’s changing distance
from Jupiter because of its finite eccentricity (Helfenstein
and Parmentier, 1980). In Europa’s early history, it is ex-
pected that the entire satellite spun nonsynchronously; how-
ever, tidal torques would have changed its spin rate, despin-
ning it on a timescale much shorter than the age of the solar
system such that the satellite’s solid interior now rotates
synchronously (Peale, 1977; Squyres and Croft, 1986), al-
though the decoupled icy shell may rotate nonsynchro-
nously (see section 4.1.6). Nonetheless, even after reaching
a synchronously rotating state, the spin rate will continue to
change as Europa’s orbit evolves. Outward orbital migra-
tion (Yo de r, 1979) increases the semimajor axis length, forc-
ing Europa’s spin rate to decrease as tidal torques further
slow its rotation rate in order to maintain a synchronous
rotation state.
A spinning satellite in hydrostatic equilibrium deforms
into a shape in which the force of gravity balances the cen-
tripetal force produced by its rotation. This deformation pro-
duces an oblate spheroid whose radius is depressed along
an axis defined by the pole of rotation and enhanced along
any orthogonal axis, extending through the equator. If the
spin rate changes, then the oblateness of the spheroid will
change in response and produce stress on the surface. De-
spinning reduces its oblateness (∆f > 0), producing a re-
bound of the radius along the axis of symmetry (through
the rotational poles) and all-around tensile stress (i.e. σθθ <
0 and σϕϕ < 0) at latitudes poleward of ±48.2° (Fig. 12a).
In the midlatitudes, meridional stresses (i.e., radial to the
poles) are tensile while stresses in the orthogonal direction
are compressive (i.e. σθθ < 0 and σϕϕ > 0). Tectonic fea-
tures produced by this type of global stress pattern would
most likely consist of radial and concentric tensile fractures
centered on the polar regions of Europa. These have not
been observed, suggesting that despinning is not responsi-
ble for the fracture patterns on Europa.
4.1.3. Orbital recession. The pattern of stress due to
tidal deformation (as opposed to despinning) predicts a
symmetry more consistent with the pattern of global frac-
tures observed. Tidal deformation of Europa produces a
prolate spheroid elongated along an axis connecting the
center of Europa and Jupiter. Along any orthogonal axis the
radius of the deformed body is depressed. The tide-raising
potential, W, and the response of the body to that potential,
denoted by the Love number h2, determines the height of
the tidal deformation given by H = –h2W/g, where g is the
acceleration of gravity. The value of h2 depends on the ma-
terial properties and configuration of mass within Europa’s
interior. The tide-raising potential W depends strongly on
the distance between Europa and Jupiter: W = –GMa–3 R2
(1.5 cos2 θ – 0.5), where M is the mass of Jupiter, a is the
semimajor axis of the orbit, R is the radius of Europa, and
θ is the angular distance between any point on Europa’s sur-
face and the axis between the centers of Jupiter and Europa.
Since Jupiter rotates faster than Europa orbits, the tidal
bulge raised on Jupiter by Europa is oriented slightly ahead
of its position, providing orbital energy to Europa, which
causes the orbit to migrate outward. As Europa’s orbit mi-
grates outward, the tide-raising potential deforming its sur-
face decreases and the height of the tide raised on its sur-
face is gradually reduced (Squyres and Croft, 1986). This
change in shape (∆f < 0) should produce an all-around com-
pressive stress on its surface (σθθ > 0 and σϕϕ > 0) in the
region within 41.8° of the axis of symmetry, defined by the
line between the centers of Europa and Jupiter (Fig. 12b).
In the region between the compressive zones, the stresses
are compressive radial to the axis of symmetry and tensile
in the orthogonal direction (σθθ > 0 and σϕϕ < 0). If frac-
Kattenhorn and Hurford: Tectonics of Europa 223
tures on Europa form mainly in tension, the stress from or-
bital recession would not produce fractures in the sub- and
antijovian regions. Moreover, the cracks that would form
between the sub- and antijovian regions would be radial to
those regions. Hence, the stress field due to orbital reces-
sion also fails to produce tensile fractures in a pattern that
matches Europa observations (Fig. 1).
4.1.4. Internal differentiation. During differentiation,
mass within a body such as Europa is redistributed, with
heavier materials moving toward the center. The redistribu-
tion of mass, along with any changes in material properties
that occur, will change Europa’s response to a tide-raising
potential (cf. Squyres and Croft, 1986). This change mani-
fests itself as a change in Europa’s Love number h2. As
Europa’s response to Jupiter’s tide-raising potential changes,
the prolateness of Europa is affected and the resulting
change in shape produces stress on its surface.
Differentiation reduces Europa’s response to a tide-rais-
ing potential, reducing the prolateness of the shape (∆f <
0). The resultant stress on the surface is similar to the stress
produced by outward orbital migration: compressive in the
region within 41.8° of the axis of symmetry, defined by the
line between the centers of Europa and Jupiter. In the re-
gion between the compressive zones centered about the axis
of symmetry, the stresses are compressive radial to the axis
of symmetry and tensile in the orthogonal direction (i.e.,
analogous to Fig. 12b). Again, this stress field would not
produce tension fractures in the sub- and antijovian regions
but would produce cracks radial to those regions, which does
not agree with the pattern of lineaments observed (Fig. 1).
Moreover, the initial differentiation of Europa occurred early
in its history, not long after its initial formation. Because
Europa’s surface is young (~40–90 m.y.), any tectonic fea-
tures that formed as a result of the stress produced by dif-
ferentiation have been erased by the formation of subse-
quent terrains.
4.1.5. Eccentricity and diurnal tidal stresses. If the
semimajor axis were to migrate inward, the stress field
would be the exact opposite of that described above for
orbital recession, with a region of tension within 41.8° of
the axis of symmetry. This stress field could produce ten-
sile fractures in the sub- and antijovian regions as observed
(Fig. 1). For this reason, Helfenstein and Parmentier (1983)
proposed that global-scale lineaments formed in response
to orbital eccentricity, e, which causes Europa’s distance
from Jupiter to change throughout the orbit, resulting in an
oscillating diurnal tidal height response. At perijove, Europa
is closer than average and the stresses near the sub- and
antijovian points are tensile, allowing fractures to form. At
apojove, Europa is farther than average and the stresses in
this region are compressive. We now understand that this
characterization of the stress field is incomplete. It neglects
other important effects that orbital eccentricity has on tidal
deformation, such as an oscillation in the longitudinal lo-
cation of the tidal bulge, resulting in a poor match to tec-
tonic patterns.
Because of Europa’s finite eccentricity, as Europa moves
from perijove to apojove, there is a small variation in the
radial tide with an amplitude of ∆H = (9eh2MR)/(4πρava3)
that affects f. In addition, Jupiter’s angular position with
respect to a fixed location above Europa’s surface oscillates
with an amplitude of 2e radians in longitude (and ε radi-
Fig. 12. (a) The stress field produced by a 5% decrease in Europa’s rotation rate due to orbital recession and subsequent despinning.
(b) The stress field produced by outward orbital migration, resulting in a 1% semimajor axis lengthening.
224 Europa
ans in latitude if there is a finite obliquity), changing slightly
the angular distance θ of any point on the surface relative
to the center of the tidal bulge. These two effects combine,
yielding a diurnally varying component of the tide (Green-
berg et al., 1998; Hurford et al., 2009).
The diurnal stress produced by Europa’s orbital eccen-
tricity thus changes throughout the orbit (Fig. 13). Zones
of tension and compression along the equator migrate east-
ward throughout the day. The orientations of the principal
stress axes rotate clockwise in the southern hemisphere and
counterclockwise in the northern hemisphere, with 180° of
rotation in principal stress orientations each orbit. This
changing stress field provided the context for the develop-
ment of the tidal walking hypothesis for strike-slip fault
motions (see section 2.3.4) as well as for unraveling the
patterns of cycloidal cracks on Europa (see section 2.1.2).
Nonetheless, there are caveats to the use of the diurnal stress
field pattern for unraveling all tectonic features. The con-
stantly changing nature of the stress field makes it difficult
to account for the extremely linear nature of most cracks,
and the stress magnitudes are extremely small (on the or-
der of a few tens of kilopascals) (Hoppa et al., 1999a) such
Fig. 13. Stresses produced by the diurnal oscillations of Europa’s tidal bulge are shown in one-quarter orbital increments beginning at
perijove. Zones of tension and compression along the equator migrate eastward throughout the day. The orientation of the principal stress
axes rotate clockwise in the southern hemisphere and counterclockwise in the northern hemisphere, changing by 180° each orbit.
Kattenhorn and Hurford: Tectonics of Europa 225
that it is has been questioned whether they are sufficient to
overcome the likely tensile strength of ice at the low sur-
face temperatures of Europa (Harada and Kurita, 2006).
Also, diurnally varying tidal stresses cannot account for
cycloids that cross the equator because of the mutually
opposite rotation sense of stresses in each hemisphere. Such
cycloids may provide evidence that there also exists a com-
ponent of stress due to a small amount (perhaps ~0.1°) of
obliquity that affected their formation patterns (Hurford et
al., 2006, 2009; Sarid-Rhoden et al., 2009; see chapter by
Bills et al.). The presence of a small forced libration (per-
haps 150 m or so) of the decoupled icy shell over the diur-
nal period may also contribute a component of global stress
(e.g., Rambaux et al., 2007; Hurford et al., 2008; Van Hoolst
et al., 2008; see chapter by Bills et al.). It is also possible
that Europa’s eccentricity has changed through time (see
chapter by Sotin et al.), resulting in different diurnal stress
field characteristics during the geological history. Nonethe-
less, pervasive fracturing of the icy shell and the existence
of global-scale lineaments that are linear or broadly curved
over great distances implies the existence of a higher-mag-
nitude, static state of stress that dominates in controlling
fracture geometries. Such a stress field may have been cre-
ated in the icy shell through the process of nonsynchronous
rotation.
4.1.6. Nonsynchronous rotation. Even though tidal
torques work to force Europa to rotate synchronously, that
rotation state can only be maintained if the orbit is circular
or there exists a permanent and significant mass asymme-
try within Europa (Greenberg and Weidenschilling, 1984).
However, the Laplace resonance between Io, Europa, and
Ganymede prevents tidal torques from circularizing Eu-
ropa’s orbit, forcing a small but finite orbital eccentricity.
In a noncircular orbit, torques on Europa will tend to force
it to rotate slightly faster than synchronous. This process
would be accentuated by the potential development of thick-
ness changes in an icy shell floating on a liquid ocean as
tidal dissipation moves it toward thermal equilibrium (Oja-
kangas and Stevenson, 1989a). Gravitational torques from
Jupiter acting on a variable thickness icy shell could induce
nonsynchronous rotation on the thermal diffusion timescale
(<10 m.y. per rotation), resulting in the shell attaining a state
of dynamic equilibrium. Hence, as long as tidal heating
within Europa prevented any permanent mass asymmetries
from forming within the icy shell, the tidal torques may
have forced Europa to rotate nonsynchronously (Greenberg
and Weidenschilling, 1984), producing a significant contri-
bution to the global stress state (Helfenstein and Parmentier,
1985).
Stresses from nonsynchronous rotation are caused by the
reshaping of Europa as the location of the tidal bulge moves
with respect to a fixed location on the surface. The actual
amount of tidal deformation does not change, only the geo-
graphic location relative to the tidal bulges. The stress pro-
duced by 1° of nonsynchronous rotation produces an equa-
torial zone of tension to the west of the sub- and antijovian
points and a zone of compression to the east of the sub-
and antijovian points (Fig. 14). At high latitudes, the prin-
cipal stresses (σθθ and σφφ) have a range of magnitudes and
signs. This stress field is broadly consistent with the pattern
of ridges observed on the surface (Figs. 1 and 15). There-
fore, although other sources of stress may contribute to
global-scale crack formation, the stress from nonsynchron-
ous rotation may be an important contributor to the global
tectonics. Moreover, the amount of nonsynchronous rota-
tion needed to produce stress of sufficient magnitude to
allow tensile failure can be obtained within a few degrees
of reorientation of the icy shell, assuming the rate at which
the shell reorients is sufficiently rapid that stresses can ac-
cumulate elastically (Harada and Kurita, 2007; Wahr et al.,
2009). However, a far greater amount of cumulative nonsyn-
chronous rotation is needed to account for the current longi-
tudes of many fractures relative to their probable formation
locations, if they indeed formed in response to nonsynchron-
ous rotation (McEwen, 1986) (Fig. 15; see section 5.1).
Geissler et al. (1998a,b) identified what appeared to be
a systematic change in the azimuth of cracks over time, sug-
gesting that Europa’s surface has migrated eastward relative
to the direction of Jupiter. Hence, at different times in the
past, regions where cracks formed in a range of orientations
may have been located within a zone of tension within the
nonsynchronous rotation stress field, allowing these features
to have formed as unrelated, superposed episodes of tension
cracks that were subsequently translated eastward with re-
spect to the zone of tension. As the icy shell migrated east-
ward, the changing orientation of the stress field at these
locations resulted in new episodes of tension fracturing
being superposed on older episodes, sometimes resulting
in a seemingly conjugate pattern. Similar studies, utiliz-
Fig. 14. The stress field produced by 1° of nonsynchronous rota-
tion produces zones of tension just west of 0° and 180° longitude.
226 Europa
Fig. 15. See Plate 13. (a) The tidal stress-field produced by 1º of nonsynchronous rotation (Greenberg et al., 1998) shows a good fit
to the locations and orientations of several lineaments. The lineaments are numbered (1) Astypalaea, (2) Thynia, (3) Libya, (4) Agenor,
(5) Udaeus, and (6) Minos Lineae. A better fit is produced if fractures are back-rotated westward in longitude relative to the stress
field [by an amount given by their color coding, as defined in (b)] such that their orientations are perpendicular to the direction of
maximum tension, providing a possible original longitude for the creation of the crack. (b) The global pattern of lineaments based on
Galileo observations illustrates a range of orientations that cannot all be fitted to the same stress template, such as that shown in (a),
because they formed at different times. The lineaments are color coded to indicate how far westward they must be back-rotated in
order to form perpendicular to the maximum tension produced by the stress of nonsynchronous rotation [map courtesy of Z. Selvans
(Selvans et al., in preparation)]. Some of the numbered lineae in (a) are numbered accordingly in (b).
Kattenhorn and Hurford: Tectonics of Europa 227
ing higher-resolution data, have confirmed that a correla-
tion between age and orientation exists (see section 5.1),
and have been related to various interpretations of the ef-
fects of nonsynchronous rotation of the icy shell (Figueredo
and Greeley, 2000; Kattenhorn, 2002; Sarid et al., 2004,
2005, 2006; Groenleer and Kattenhorn, 2008).
4.1.7. Polar wander. Another potential contributor to
the global stress state is provided by long-term latitudinal
reorientation of the icy shell, or polar wander, in response
to latitudinal changes in ice thickness as the icy shell ther-
mally evolves (Ojakangas and Stevenson, 1989a,b). Leith
and McKinnon (1996) explored the theoretical stress pro-
duced by this phenomenon and found only limited evidence
for polar wander within observable crack patterns. Sarid et
al. (2002) noted that the slip sense along some near-equa-
torial strike-slip faults did not agree with the predictions of
tidal walking (see section 2.3.4) and that a small amount
of polar wander since fault motions accrued may explain
the distribution of the observed slip sense. However, finite
obliquity is another potential mechanism for such aberra-
tions (Hurford et al., 2009) as well as the effects of con-
vergence across ridges, which may create an apparent sense
of offset in the absence of any actual strike-slip motions
(see section 2.2.3). Schenk et al. (2008) interpret the loca-
tions and geometries of inferred extensional zones called
small-circle depressions (SCDs) as being due to true polar
wander. These curved depressions are up to 1.5 km deep
and trace out segments of the arcs of almost perfect circles
centered ~25° from the equator, forming an antipodal pat-
tern in the leading and trailing hemispheres. The shapes and
locations of the troughs are suggested to be consistent with
the stress field that would result from 80° of polar wander.
It is unclear what the timing of this hypothesized shell re-
orientation event may have been because other surface fea-
tures are seemingly unaffected by the locations of the SCDs.
The associated polar wander stress field (e.g., Matsuyama
and Nimmo, 2008) has also not been compared to the ori-
entations of the vast number of other tectonic features on
Europa. Therefore, polar wander has yet to be shown to be
a likely source of global stresses on Europa in terms of its
potential contribution to the pattern of global tectonics.
4.1.8. Ice shell thickening. If Europa’s icy shell in-
creased in thickness relatively recently, the related stress
may have contributed to tectonic activity (see chapter by
Nimmo and Manga). For example, Nimmo (2004a) showed
that for a nonconvecting icy shell that thickens in excess
of 20 km, isotropic tensile stresses are produced within the
icy shell, attaining magnitudes of perhaps 10 MPa at the
surface, which are sufficient to induce tension fracturing and
perhaps even shell-penetrating cracks (see also Manga and
Wang , 2007). Because these stresses are isotropic, the prin-
cipal stress orientations would still be controlled by the
nonisotropic (e.g., tidal) components, which would thus
control fracture orientations. This effect provides a poten-
tial solution to the mystery of tensile fracturing in equato-
rial compressive zones within the nonsynchronous rotation
stress field, as described by Spaun et al. (2003) (see sec-
tion 2.3.2). Another effect of an increasing icy shell thick-
ness is a decreasing tidal response (e.g., a ~1% reduction
if the shell thickens from 1 km to 10 km) (Moore and Schu-
bert, 2000; Hurford, 2005), resulting in slightly reduced
tidal stress magnitudes. Also, a change in ice thickness re-
sults in a change in the Love number, h2, creating an addi-
tional component to the stress field similar to that induced
by internal differentiation (see section 4.1.4).
4.2. Regional and Local Stress
Although global-scale stresses are responsible for the
formation of the majority of the cracks on Europa’s sur-
face, regional and local scale stresses have also played an
important role in the tectonic history. These stresses essen-
tially represent regional and local perturbations to the glo-
bal stress state and are significant for the tectonics if their
magnitudes are on the order of the global stresses or higher.
Although regional and local have somewhat arbitrary defi-
nitions, the important distinction is that they do not result
from the global effects of tidal deformation or reorienta-
tion of the icy shell. Regional effects may encompass thou-
sands of kilometers across the surface, depending on the
causal factor. For example, endogenic processes such as
thermal or compositional diapirism (see chapter by Barr and
Showman) create upwelling that can impart regional stress
conditions on an overlying stagnant lid. The formation of
chaotic terrain, which represents the destruction of an earlier
surface, is likely related to this process (Collins et al., 2000;
Sotin et al., 2002; Mitri and Showman, 2008; see chapter
by Collins and Nimmo). Chaos areas can only become dis-
aggregated through the initial development of fractures that
create loose blocks of brittle ice that are subsequently rafted
around by the underlying motion of warmer ice and melt.
Indeed, one of the types of tension fractures described in
section 2.1.3, endogenic fractures, is specifically related to
this phenomenon. The characteristics of the stress field
caused by a diapir impinging on the brittle ice lid from
below are not easily quantifiable, as they would likely de-
pend on the shape and size of the diapir. Although the stress
magnitudes are likely to be small (tens of kilopascals), they
may be sufficient to drive deformation in the brittle lid
(Showman and Han, 2005). A common observation is that
endogenic fractures form concentric patterns about regions
of chaos, suggesting that the causal stress field may be
somewhat similar to that of an ascending bulge, analogous
to upwarps above laccolith intrustions on Earth (Jackson
and Pollard, 1988). Regional perturbations to the global
stress state may also be imparted by large meteorite impacts,
which can result in both radial and concentric fracture pat-
terns for the case of complete shell penetration (Melosh,
1989), although radial fractures are not observed around the
large impact site Tyre (Kadel et al., 2000).
Local stress effects occur at scales on the order of hun-
dreds of meters to perhaps hundreds of kilometers. The
causal mechanism is typically apparent, usually related to
an existing ridge, strike-slip fault, or fold, or to regions of
228 Europa
lenticulae (small disruptions related to underlying diapiric
activity). For example, ridge constructs are up to several
hundreds of meters high and their weight is expected to
push down on the icy shell. As a result, the elastic portion
of the shell flexes downward beneath the ridge, apparently
producing secondary uplifts, or forebulges, to either side of
the ridge, evident in subtle changes in shading and shadow
patterns (Tufts, 1998; Hurford et al., 2005). The resultant
stresses induced by bending of the icy shell are sufficient to
fracture the icy shell to either side of the ridge (Fig. 2b) and
can be quantified using the equations describing the flex-
ure of an elastic plate beneath a line-load (Turcotte and
Schubert, 2002). Billings and Kattenhorn (2005) used these
equations to formulate a relationship between elastic ice
thickness and the distance from a ridge load to its flanking
secondary cracks, allowing the thickness of the elastic por-
tion of the icy shell to be constrained to between ~200 m
and ~2400 m. Bending stresses are also responsible for the
local development of surface cracks along the hinge lines
of anticlinal folds, such as occurs within Astypalaea Linea
(Prockter and Pappalardo, 2000).
Strike-slip faults produce local perturbations to the glo-
bal stress field in response to lateral motions that create
stress concentrations at the fault tips (Kattenhorn, 2004a).
These stress fields are quantifiable (see section 2.3.3) and
have been shown to be responsible for the development of
secondary fractures at fault tips called tailcracks (see sec-
tion 2.1.4) (Prockter et al., 2000; Schulson, 2002; Katten-
horn, 2004a; Kattenhorn and Marshall, 2006). Tailcracks
form in the locally perturbed stress field in the vicinity of
a tip of a shearing fault or fracture. The remote (far-field)
stresses govern the sense of slip on the fault itself while
shear motion along the fault locally perturbs the stress field,
creating near-tip quadrants of extension and compression,
with an antisymmetric pattern at opposing tips. Tailcracks
form in extensional quadrants and propagate away from the
fault tip perpendicular to the direction of maximum local
tension. A similar phenomenon is responsible for the devel-
opment of cusps along cycloidal chains (Groenleer and Kat-
tenhorn, 2008) in response to lateral shearing effects during
the diurnal cycle. Compressive quadrants may also experi-
ence local deformation in the form of folds or anticracks,
the latter having been suggested to exist in the “Wedges”
region, Argadnel Regio (Kattenhorn and Marshall, 2006).
5. TECTONIC EVOLUTION OF EUROPA
5.1. Tectonic Evidence for
Nonsynchronous Rotation
Theoretical considerations of nonsynchronous rotation
stress fields (see section 4.1.6) and their match to the orienta-
tions of geologically recent large-scale lineaments prompts
the search for direct evidence for extensive nonsynchro-
nous rotation to determine its relative importance throughout
Europa’s visible tectonic history. Comparisons between lin-
eament orientations on Europa and theoretical global stress
fields suggest that many geologically recent lineaments did
not form in their current longitudes (McEwen, 1986; Green-
berg et al., 1998; Hoppa et al., 1999b). Instead, these linea-
ments are inferred to have translated longitudinally due to
nonsynchronous rotation by a few tens of degrees (Fig. 15).
Many locations on Europa show a complex sequence of
multiply superposed lineaments with differing orientations,
reflecting the dominant tensile stress orientation at that loca-
tion at the time of lineament development. This is plausi-
bly because the regional stress orientation changed as the
icy shell migrated steadily eastward through time, rotating
through a full 360° per nonsynchronous rotation cycle with
a clockwise sense in the northern hemisphere and a counter-
clockwise sense in the southern hemisphere. Accordingly,
lineament orientations have been suggested to display this
rotation sense through time (McEwen, 1986; Geissler et al.,
1998a; Figuered o and Greeley, 2000; Kadel et al., 2000;
Kattenhorn, 2002). The amount of reorientation of the icy
shell inferred from the tectonic history based on these studies
is as little as 50° and up to 1000°, but may be much more
depending on the amount of time between individual linea-
ment formation events, which is unknown. A problem with
these inferences is that changing fracture orientations do not
directly indicate that nonsynchronous rotation actually oc-
curred. Considering that nonsynchronous stress orientations
sweep through all angles, it is not surprising that interpreta-
tions are equivocal. In fact, if one arbitrarily assumes that
the sense of rotation of lineaments through time is the op-
posite of what would be expected in response to nonsyn-
chronous rotation, the number of complete shell rotations
needed to account for the full range of lineament orienta-
tions through time does not differ greatly to the nonsyn-
chronous rotation result (Sarid et al., 2004, 2005, 2006).
Hence, although changing lineament rotations through time
are certainly consistent with nonsynchronous rotation, this
phenomenon does not provide direct proof that nonsynchro-
nous rotation occurs.
Evidence for nonsynchronous rotation was suggested by
Hoppa et al. (1999c) to be evident in the slip sense of strike-
slip faults in equatorial areas, where the sense of cumula-
tive slip in the context of the tidal walking model (see sec-
tion 2.3.4) is completely dependent on the fault orientation.
Some faults do not have the correct sense of slip for their
orientation and location; however, back-rotation of a non-
synchronously rotated shell by up to 90° can place each
fault at a longitude where the slip sense is compatible with
expectations.
Additional evidence for icy shell reorientation, probably
through nonsynchronous rotation, has emerged from stud-
ies of cycloids. Hoppa et al. (2001) show examples of
southern hemisphere cycloids that could not have formed
in their current locations but rather at locations up to 80°
further to the west, after which nonsynchronous reorienta-
tion of the icy shell translated them eastward. Hurford et al.
(2007a) reduced this estimate to 24° for these particular
cycloids by accounting for the presence of a small amount
of nonsynchronous rotation stress during cycloid growth,
Kattenhorn and Hurford: Tectonics of Europa 229
indicating that reorientation of the icy shell is still necessary
to fit the cycloid shapes. However, the inclusion of a small
amount of obliquity may remove the need for a component
of nonsynchronous rotation stress to account for the shapes
of some cycloid examples (Sarid-Rhoden et al., 2009).
Groenleer and Kattenhorn (2008) document a history of
numerous cycloidal episodes in the northern trailing hemi-
sphere that necessitate at least 600° of nonsynchronous re-
orientation of the icy shell during the period of cycloid de-
velopment alone (i.e., excluding all other lineament types).
Crosscutting relationships among these cycloids imply a
definitive age sequence that can best be reconciled with
cycloid chains having formed during different nonsynchro-
nous rotation cycles, lending credence to the idea that only
a few cycloid chains form during each nonsynchronous ro-
tation cycle.
The wide range of estimates of nonsynchronous rotation
amounts may seem contradictory; however, disparate results
likely reflect variances in both the types of features used
to make the estimates and the differing resolutions of im-
ages studied in different regions. The combined studies sug-
gest that the decoupled icy shell of Europa may have un-
dergone almost three complete rotations relative to the rocky
interior during the visible tectonic history, although this
number may be much higher in reality. In support of this
notion, given the 40–90-m.y. surface age of Europa (see
chapter by Bierhaus et al.), nonsynchronous rotation must
have occurred rapidly enough to accrue elastic stresses
before they could be relieved viscously beyond the Max-
well relaxation time (Harada and Kurita, 2007; Wahr et al.,
2009). Given this rate of rotation, there have likely been
numerous rotations of the icy shell relative to the interior
during Europa’s visible history. Nonetheless, if nonsynchro-
nous rotation effects were relatively constant over the en-
tire visible geological history, one might expect there to be
a somewhat unchanging tectonic response throughout this
time. On the contrary, Europa exhibits a diverse geological
history with distinct temporal variability in the tectonic
features that developed, implying that other factors need to
be considered when examining the tectonic evolution of
Europa, such as variations in the rate of nonsynchronous
rotation (Nimmo et al., 2005; Hayne et al., 2006).
5.2. Temporal Changes in Tectonic Style
Previous sections of this chapter considered the range
of tectonic features that pervasively deform the icy shell of
Europa and the possible sources of stress responsible for
those features. We now turn to the geological development
of Europa inferred from the many episodes of tectonic
events evident on the surface. Over the visible surface his-
tory of the icy shell, repeated fracturing events have left a
remarkably coherent story regarding the tectonic evolution
in the form of clearly identifiable crosscutting relationships.
In so doing, tectonic features reveal that there has been a
gradual change in the nature of deformation that points
toward temporal variations in the thickness of the icy shell,
processes occurring within the shell, and the response of
the shell to tidal deformation.
Numerous studies aimed at unraveling the geological
history of Europa have revealed significant changes through
time (Lucchitta and Soderblom, 1982; Prockter et al., 1999,
2002; Figueredo and Greeley, 2000; 2004; Greeley et al.,
2000; Kadel et al., 2000; Kattenhorn, 2002; Spaun et al.,
2003; Riley et al., 2006; Doggett et al., 2007). A simple
summary of these changes has early development of ridged
plains (a multitude of closely spaced ridges) being followed
by periods of ridge and band development, endogenic cryo-
magmatic disruption and cryovolcanism (chaos, lenticulae,
and smooth plains) with some contemporaneous and sub-
sequent ridge formation, and a late stage of tectonic frac-
turing (see chapter by Doggett et al.). Many lineaments
were reactivated as strike-slip faults during the periods of
tectonic activity. The following synthesis considers the in-
dividual evolution of the most prominent types of tectonic
features that have been placed into a stratigraphic context.
5.2.1. Ridges. The oldest portions of the surface ap-
pear to consist of finely crenulated ridged plains or sub-
dued plains that may represent the earliest visible ridges,
with spacings on the order of hundreds of meters. Most of
the ridged plains were subsequently destroyed by the re-
peated development of successively younger double ridges
(i.e., a prominent central trough flanked by two raised rims)
and dilational bands in a range of orientations. Older ridges
appear to be more numerous, narrower, lower, and more
closely spaced (tens to hundreds of kilometers) than rela-
tively younger ridges (Kadel et al., 2000; Kattenhorn, 2002;
Figueredo and Gre eley, 2004). Many of the prominent (and
typically geologically young) global-scale ridges that cover
broad portions of the europan surface have evolved into
ridge complexes, forming the highest and widest ridge struc-
tures. Ridge development thus spans most of the geological
history of the surface, although the youngest ridges could
have formed many tens of degrees of nonsynchronous rota-
tion in the past. Hence, it is unclear if ridges continue to
form or if the processes responsible for their development
(see section 2.1.1) still occur in the icy shell.
5.2.2. Cycloids. There has been some uncertainty as
to whether cycloid development represents a geologically
recent phenomenon on Europa. Groenleer and Kattenhorn
(2008) show that there has been an extended period of cy-
cloid development in the trailing hemisphere encompass-
ing at least 600° of nonsynchronous rotation, with the most
recent event occurring at least 30° of nonsynchronous ro-
tation in the past. These cycloid-forming events were punc-
tuated by periods of intervening linear ridge development,
suggesting that there was an oscillation between the con-
ditions conducive to cycloid and linear fracture formation,
respectively. Although these oscillations are observed in
specific areas, it is unknown whether or not a similar global
temporal pattern of fracturing occurs or whether cycloid de-
velopment in one region of Europa could have occurred
contemporaneously with linear ridge development else-
where. Regardless, whatever processes were responsible for
230 Europa
ridge construction continued regardless of whether cracks
were linear or cycloidal, indicated by ridge development
along the flanks of cycloids as well as linear fractures.
Unlike linear ridges, however, cycloids have not been expli-
citly described from the earliest part of the geological his-
tory, raising the possibility that the initiation of the condi-
tions needed to form cycloids occurred at some critical
juncture during the tectonic history.
5.2.3. Dilational bands. There is also a wide range in
the ages of dilational bands, some of which formed around
the time of the earliest ridges (Greeley et al., 2000; Katten-
horn, 2002); however, it is possible that the driving proc-
ess behind dilational band development ceased on Europa
even while ridge development continued. Prockter et al.
(2002) indicate that there are always features younger than
dilational bands, including troughs, ridges, and lenticulae.
Nonetheless, dilational bands change in albedo from dark
to light through time and the youngest dilational bands still
have dark albedos. It is not known how quickly this light-
ening process occurs, so dilational bands could be many
millions of years old. Dilational band widths are inferred
to have decreased through time (Figueredo and Greeley,
2004), perhaps indicating a thinner icy shell and higher ther-
mal gradients earlier in Europa’s visible geological history.
Nimmo (2004b) modeled the extension of icy shells and
demonstrated a trade-off between strain rate and shell thick-
ness in terms of whether wide or narrow rifts will develop.
Narrow rifts are analogous to europan dilational bands and
imply high strain rates or relatively thick icy shells. It is
possible that changing shell thickness and strain rate con-
ditions on Europa resulted in a transition in the mode of
extension to no longer favor dilational band development.
Hence, dilational bands rarely postdate chaos (see chapter
by Prockter and Patterson). Those young dilational bands
that appear to postdate chaos formation have dilated pre-
existing cycloids (Figueredo and Greeley, 2000); therefore,
the initiation of the process responsible for cycloid devel-
opment did not correspond to a termination of the process
responsible for dilational band development, with both pro-
cesses possibly overlapping with the period of chaos for-
mation.
Figueredo a nd Greeley (2004) suggest that a long period
of tectonic resurfacing by ridge and band development was
followed by a rapid decrease in tectonic activity through
time, concomitant with a steady increase in “cryovolcanic”
activity that resulted in the formation of broad areas of
chaos and lenticulae. These changes were attributed to the
effects of a thickening icy shell that ultimately reached the
threshold thickness required to induce solid-state convec-
tion, diapirism, and cryovolcanic resurfacing (Papp ala rd o
et al., 1999; Barr et al., 2004; Barr and Pappalardo, 2005;
see chapter by Nimmo and Manga). Nonetheless, an even
younger period of trough development occurred (both lin-
ear and cycloidal), concomitant with recent cratering events.
Considering that the visible geological history of Europa
only spans around 2% of the total age of the satellite, it is
not known if tectonic resurfacing and cryovolcanic pro-
cesses occur in an oscillating cycle related to repeated thin-
ning and thickening of the icy shell, perhaps associated with
changes in Europa’s orbital eccentricity (see chapters by
Nimmo and Manga, Moore and Hussman, and Sotin et al.).
A further complication is provided by the possibility of
lateral variations in icy shell thickness due to latitudinal and
longitudinal variations in tidal heating (Ojakangas and
Stevenson, 1989a) and past tectonic modification of the icy
shell (Billings and Kattenhorn, 2005). Regardless, the in-
ferred thickening of the icy shell over the past 40–90 m.y.
clearly influenced the tectonics.
5.3. Active Tectonics on Europa?
Although many bodies in the solar system show evidence
of a remarkable geological history of tectonic activity (e.g.,
Mars, Venus, icy moons of Jupiter, Saturn, Uranus, and
Neptune), evidence of current tectonic activity outside of
Earth has been very elusive. With the recent discovery of
eruptive water-ice plumes emanating from cracks (called
“tiger stripes”) in the south polar area of Saturn’s icy moon
Enceladus (Porco et al., 2006; Spencer et al., 2006), a re-
lationship to active tectonics has been inferred, based on
models of crack motions within a tidal stress field (Hurford
et al., 2007b; Nimmo et al., 2007; Smith-Konter and Pappa-
lardo, 2008). Any icy satellite with a liquid ocean beneath
an icy shell that has apparently deformed due to the tidal
response of the underlying ocean therefore provides a good
candidate study for active tectonics. Hence, Europa is a
promising candidate for active tectonics. It is generally
accepted that a relatively thin icy shell (≤30 km) overlies a
liquid ocean on Europa (see section 4.1.1) that experiences
an oscillating tidal response due to its orbital eccentricity.
Despite the extensive body of work on the characteriza-
tion and interpretation of europan tectonic features dis-
cussed in this chapter, this work could not directly answer
the question of whether there is tectonic activity on Europa
today. Past analyses demonstrated a good correlation be-
tween tidal stresses and tectonic lineaments, but focused on
relatively old geological features such as ridges and dila-
tional bands. Although some ridges can be placed into the
youngest portions of the geological history (Figueredo and
Greeley, 2000, 2004), especially cycloidal ridges, an analy-
sis of ridge orientations or morphologies cannot be used to
answer the question about active tectonics for three impor-
tant reasons.
First, even geologically young cycloidal ridges currently
occur in longitudinal locations (relative to the tidal bulges)
that do not match the current tidal stress fields, suggesting
that some amount of nonsynchronous rotation has occurred,
implying at least several tens of thousands of years since
those ridges formed (Hoppa et al., 2001; Hurford et al.,
2007a; Groenleer and Kattenhorn, 2008). Although this is
not a great period of time, geologically speaking, it must
be remembered that the nonsynchronous rotation rate is still
not well constrained and that these ridges are not the young-
est tectonic features. Also, they could have formed 30° mod-
Kattenhorn and Hurford: Tectonics of Europa 231
ulo 180° in the past (based on the global symmetry of the
tidal stress field). The upshot is that cycloidal ridges do not
provide proof of current tectonic activity.
Second, ridge-producing cracks may take a long time to
construct their ramparts to either side of the crack (see sec-
tion 2.1.1). The construction of a ridge is a late stage in an
evolutionary process that begins with the development of
a trough. It is not known how long this ridge-building pro-
cess takes, mostly because many years of ridge analyses
have still not conclusively determined their mode of devel-
opment. Typical ridges may take tens of thousands of years
to form and can only do so when optimally oriented with
respect to the changing stress field during the nonsynchro-
nous rotation process. Hence, ridges cannot be used to make
inferences about current tectonic activity on Europa, even
if they appear to be geologically young.
Third, the still-present topography of up to 200 m still
observed along dilational bands and even higher elevations
along ridges that date back through the ~40–90 m.y. of
geological history indicates that topography can survive for
long time periods and is thus unlikely to be an indicator
of recent tectonics. Dombard and McKinnon (2006) sug-
gest that appreciable topography may actually survive for
>100 m.y. considering the low rate of viscous relaxation,
although their results vary as a function of heat flow and
surface temperature. Nimmo et al. (2003) show that com-
positional buoyancy due to lateral density differences in a
cold, near-surface elastic layer (perhaps related to variabil-
ity in either ice porosity or salt concentration) could explain
prolonged topography that might otherwise be eradicated
in less than 0.1 m.y. by ductile flow of ice from below an
uncompensated surface load. Luttrell and Sandwell (2006)
indicate that short-wavelength topography (~20 km), as
would characterize ridges, can in fact be supported by the
strength of the brittle portion of icy shell itself without the
need for underlying support.
At least three factors have complicated attempts to make
inferences about active tectonics on Europa:
1. There is a distinct lack of observational evidence for
tectonic activity in the time interval between the imaging
periods of the Voyager and Galileo spacecraft missions. A
comparison of Voyager and Galileo images based on an
iterative coregistration technique detected no evidence for
deposition by plume activity (such as that which accompa-
nies tectonic activity on Enceladus), nor any evidence for
new fracturing during the 20-year interval between missions
(Phillips et al., 2000). The lack of inferred plume activity
could simply be the result of negative buoyancy effects that
hamper surface eruptions (Crawford and Stevenson, 1988),
regardless of the fact that cracks might fully penetrate an
icy shell that has substantially cooled and thickened (Manga
and Wang, 2007). Hence, absent plumes do not necessar-
ily imply no active tectonics. Additionally, no evidence for
any geological (tectonic or otherwise) changes was discov-
ered, so inferences regarding active tectonics are inconclu-
sive. Part of the difficulty introduced into attempts at com-
paring these datasets is the low resolution of Voyager data
and differences in photometric angles, which can dramati-
cally alter the appearance and albedo of surface features.
Future missions could be targeted to match Galileo cover-
age at high resolutions, which may increase the chances of
measuring geological changes (Phillips and Chyba, 2003).
2. Inferences about tectonic activity based on the geo-
logical history in general have been inconclusive. Based on
the low density of craters relative to the expected impact
flux rate, Europa’s surface appears to be geologically young
(average age of 40–90 m.y.). This raises the likelihood that
tectonic resurfacing has been a geologically recent process
that may continue today. Although references have been
made to “recent” tectonic lineaments on Europa (e.g., Geiss-
ler et al., 1998a), inferences about what features are geolog-
ically recent are commonly thwarted by analyses of higher-
resolution images that reveal comparatively younger geo-
logical features (tectonic fractures and endogenically driven
chaos and lenticulae). Europa has experienced numerous
changes in tectonic style over its visible history, with a prev-
alence of endogenically driven surface disruption in the
recent geological history (see section 5.2). Nonetheless, a
younger period of postchaos fracturing has been noted in
several studies (Kadel et al., 2000; Figueredo and Greeley,
2000, 2004; Doggett et al., 2007), implying that the devel-
opment of chaos did not equate to the cessation of tectonic
activity on Europa. Although these geological studies point
to a high likelihood that there was a geologically young pe-
riod of tectonic activity, they have not unequivocally de-
termined whether such activity is ongoing.
3. Theoretical stresses in a thickening icy shell allow
ongoing fracturing but imply a reduction or cessation of
tidally driven tectonic activity. Geological inferences of a
thickening icy shell based on convincing evidence that the
threshold thickness for solid-state convection was attained
(Barr et al., 2004; Barr and Pappalardo, 2005; see chap-
ter by Barr and Showman) has led to the development of
theoretical models to address the impact of a thickening icy
shell on the tidal stress field. As a result of the decrease in
tidal deformation as an icy shell thickens, and the fact that
tidal stresses are relatively small even in a thin icy shell,
there may be a decreasing likelihood of tidally driven tec-
tonic activity through time on Europa (Hurford, 2005), re-
flected in the apparent transition from tectonic resurfacing
to cryovolcanic resurfacing (see section 5.2). Hence, pro-
cesses such as cycloid development and tidal walking along
strike-slip faults, which are driven primarily by diurnal tidal
stresses, may decline. Nonetheless, tensile stresses produced
by expansion of a thickening icy shell (in the 10 MPa range)
can far exceed those induced by diurnal or nonsynchronous
rotation effects and could promote surface fracturing (see
section 4.1.8). Because the stresses are isotropic, orienta-
tions of tectonic features will still be controlled by the non-
isotropic components of the total stress field (e.g., tidal
stresses), even if they are small. Thus, theoretical consid-
erations show that the potential for sufficiently high stresses
exists to promote active tectonics that can be reconciled
with the current tidal stress state. Nonetheless, it cannot be
232 Europa
proven whether or not tectonic activity actually exists. Al-
though thickening of the icy shell may have indeed induced
changes in the tectonic style on Europa through time, as
well as plausibly initiating convective overturn and cryo-
volcanism, it did not necessarily herald the death knell for
tectonics.
The question of active tectonics on Europa is open ended.
Its importance for understanding the interior dynamics of
this icy world, as well as providing potential access path-
ways to the subsurface ocean, will undoubtedly motivate
future studies in this direction, particularly in the context
of future mission goals and design (e.g., Sandwell et al.,
2004; Panning et al., 2006; see chapter by Greeley et al.).
6. CONCLUDING REMARKS
The icy shell of Europa has experienced pervasive de-
formation that has resulted in a wide range of tectonic fea-
tures over the 40–90-m.y. visible history. This deformation
was driven by a number of factors, whether global, regional,
or local. The two most important global contributors to
stresses were probably diurnal tidal deformation induced
by a tidally responding subsurface ocean and a longer time-
scale longitudinal reorientation of the icy shell relative to
its tidal bulges by nonsynchronous rotation.
Such stress fields commonly produced tensile compo-
nents, resulting in ubiquitous extensional deformation. Ten-
sion fractures at the surface were reworked over many tens
of thousands of years to form ridges. These ridges, com-
posed of raised rims to either side of a central fracture,
likely formed in response to a combination of extensional,
compressive, and shear-related processes. Complete sepa-
ration of the icy shell is evidenced by the development of
dilational bands, which formed in response to tidally driven
extension and endogenic processes such as local upwelling
of ductile portions of the shell. Cycloidal fractures provide
a strong indication that the diurnal stress field dominated
in the creation of tectonic features at times, although this
may have been more common later in the geological his-
tory, potentially as a result of a decrease in the rate of non-
synchronous rotation stress build-up as the icy shell gradu-
ally thickened. Long, linear, or curvilinear fractures likely
formed in response to dominant nonsynchronous rotation
stresses. Extension through normal faulting is seemingly
rare within the ridged plains but commonplace within dila-
tional bands, which resemble terrestrial mid-ocean ridges
in many respects. Long-term changes in the orientations of
multiply superposed extensional lineaments strongly sug-
gest a temporally variable stress field, ostensibly in response
to nonsynchronous rotation of the icy shell.
Contractional deformation is required to balance the
surface area budget in light of the new surface area created
at dilational bands, which exceeds the amount that would
be expected purely as a result of icy shell thickening and
expansion. Initially elusive in Galileo images, mounting
evidence suggests that contraction may be common on Eu-
ropa. Although folding is not likely to be globally signifi-
cant, convergence bands and ubiquitous ridges are plausi-
bly prominent accommodators of contraction. Unanswered
questions remain regarding where convergence bands pref-
erentially develop and whether contraction across ridges
contributes to ridge construction in tandem with shearing
and frictional heating.
Lateral shearing is common as a result of the constantly
changing orientations of principal stresses predicted on the
diurnal timescale as well as over longer time periods due
to nonsynchronous rotation. As long as a fracture has not
healed, it can potentially be reactivated by shear stresses
that resolve onto the fracture plane from the tidal stress
field. These shear stresses may induce strike-slip motions,
which may or may not be frictionally resisted along the
crack depending on whether the concomitant normal stresses
are compressive or tensile (i.e., whether the crack is closed
or dilated during the shearing). Active fractures should un-
dergo cycles of tension and compression, as well as back-
ward and forward motion, over the course of the europan
day. These motions may result in a slow ratcheting with a
preferred motion sense depending on location and crack
orientation, ultimately accumulating visible strike-slip off-
sets; this process is referred to as tidal walking. Conver-
gence across ridges may also produce apparent lateral off-
sets, resembling those created by strike-slip motions.
Europa is clearly one of the most tectonically diverse
bodies in the solar system. Its crack-riddled surface has left
a trail of clues for unraveling its deformation and stress
history. The moon likely reached the pinnacle of its tectonic
activity a long time ago (perhaps millions of years). Gradual
thickening of the icy shell may have ultimately resulted in
the attainment of a threshold thickness for convectional
overturn. This thickening may have resulted in a reduced
tidal response, a slower rate of nonsynchronous rotation,
and thus smaller global stresses, causing a decrease in the
amount of tectonic activity. Nonetheless, tectonic features
have been noted to postdate endogenically driven disrup-
tions of the surface that formed regions of chaos. It is prob-
able that tectonic activity has continued to the current day
on Europa, with driving stresses plausibly being composed
partly of small magnitude tidal components in addition to
a large isotropic tension created by the expansion of the icy
shell. As with Cassini’s encounter with an obviously active
Enceladus, it is seems likely that future Europa missions
may very well be met with a few surprises.
Acknowledgments. The authors thank G. W. Patterson and R.
Pappalardo for their insightful and constructive comments on the
original manuscript and acknowledge NASA for its financial sup-
port of the authors’ research efforts. Useful discussions with L.
Prockter, W. Patterson, P. Schenk, and C. Coulter helped improve
this work.
REFERENCES
Anderson J. D., Lau E. L., Sjogren W. L., Schubert G., and Moore
W. B. (1997) Europa’s differentiated internal structure: Infer-
ence from two Galileo encounters. Science, 276, 1236–1239.
Aydin A. (2006) Failure modes of the lineaments on Jupiter’s
moon, Europa: Implications for the evolution of its icy crust.
Kattenhorn and Hurford: Tectonics of Europa 233
J. Struct. Geol., 28, 2222–2236.
Bader C. E. and Kattenhorn S. A. (2007) Formation of ridge-type
strike-slip faults on Europa. Eos Trans. AGU, 88, Fall Meet.
Suppl., Abstract P53B–1243.
Bader C. E. and Kattenhorn S. A. (2008) Formation mechanisms
of europan ridges with apparent lateral offsets. In Lunar and
Planetary Science XXIX, Abst rac t # 203 6. Lun ar and Pl ane tar y
Institute, Houston (CD-ROM).
Barr A. C. and Pappalardo R. T. (2005) Onset of convection in
the icy Galilean satellites: Influence of rheology. J. Geophys.
Res., 110, E12005, DOI: 10.1029/2004JE002371.
Barr A. C., Pappalardo R. T., and Zhong S. (2004) Convective
instability in ice I with non-Newtonian rheology: Application
to the icy Galilean satellites. J. Geophys . Res. , 109, E12008,
DOI: 10.1029/2004JE002296.
Bart G. D., Greenberg R., and Hoppa G. V. (2003) Cycloids and
wedges: Global patterns from tidal stress on Europa. In Lunar
and Planetary Science XXIV, A bstrac t # 1396. Lunar and Pl ane-
tary Institute, Houston (CD-ROM).
Beeman M., Durham W. B., and Kirby S. H. (1988) Friction of
ice. J. Geophys. Res., 93, 7625–7633.
Billings S. E. and Kattenhorn S. A. (2005) The great thickness
debate: Ice shell thickness models for Europa and compari-
sons with estimates based on flexure at ridges. Icarus, 177,
397–412.
Collins G. C., Head J. W. III, Pappalardo R. T., and Spaun N. A.
(2000) Evaluation of models for the formation of chaotic ter-
rain on Europa. J. Geophys. Res., 105(E1) , 1709–1716.
Cordero G. and Mendoza B. (2004) Evidence for the origin of
ridges on Europa by means of photoclinometric data from the
E4 Galileo orbit. Geofís. Intl., 43, 301–306.
Coulter C. E., Kattenhorn S. A., and Schenk P. M. (2009) Topo-
graphic profile analysis and morphologic characterization of
Europa’s double ridges. In Lunar and Planetary Science XL,
Abstract #1960. Lunar and Planetary Institute, Houston (CD-
ROM).
Crawford G. D. and Stevenson D. J. (1988) Gas-driven water vol-
canism and the resurfacing of Europa. Icarus, 73, 66–79.
Dahlen F. (1976) The passive influence of the oceans upon the
rotation of the earth. Geophys. J. R. Astron. Soc., 46, 363–406.
Doggett T., Figueredo P., Greeley R., Hare T., Kolb E., Mullins
K., Senske D., Tanaka K., and Weiser S. (2007) Global geo-
logic map of Europa. In Lunar and Planetary Science XXXVIII,
Abstract #2296. Lunar and Planetary Institute, Houston (CD-
ROM).
Dombard A. J. and McKinnon W. B. (2006) Folding of Europa’s
icy lithosphere: An analysis of viscous-plastic buckling and
subsequent topographic relaxation. J. Struct. Geol., 28, 2259–
2269.
Durham W. B., Heard H. C., and Kirby S. H. (1983) Experimen-
tal deformation of polycrystalline H2O ice at high pressure and
low temperature: Preliminary results. Proc. Lunar Planet. Sci.
Conf. 14th, in J. Geophys. Res., 88, B377–B392.
Figueredo P. H. and Greeley R. (2000) Geologic mapping of the
northern leading hemisphere of Europa from Galileo solid-state
imaging data. J. Geophys. Res., 105, 22629–22646.
Figueredo P. H. and Greeley R. (2004) Resurfacing history of Eu-
ropa from pole-to-pole geological mapping. Icarus, 167, 287–
312.
Finnerty A. A., Ransford G. A., Pieri D. C., and Collerson K. D.
(1981) Is Europa surface cracking due to thermal evolution?
Nature, 289, 24–27.
Fortt A. L. and Schulson E. M. (2004) Post-terminal compressive
deformation of ice: Friction along Coulombic shear faults. Eos
Tra ns . AG U, 8 5, T13A–05.
Gaidos E. J. and Nimmo F. (2000) Tectonics and water on Europa.
Nature, 405, 637.
Geissler P. E. and 16 colleagues (1998a) Evolution of lineaments
on Europa: Clues from Galileo multispectral imaging observa-
tions. Icarus, 135, 107–126.
Geissler P. E. and 15 colleagues (1998b) Evidence for non-syn-
chronous rotation of Europa. Nature, 391, 368–370.
Giese B., Wagner R., Neukum G., Sullivan R., and the Galileo
SSI Team (1999) Doublet ridge formation on Europa: Evi-
dence from topographic data. 31st DPS Meeting Abstracts,
Bull. Am. Astron. Soc., 31(4), 62.08.
Golombek M. P. and Banerdt W. B. (1990) Constraints on the
subsurface structure of Europa. Icarus, 83, 441–452.
Greeley R. and 17 colleagues (2000) Geologic mapping of Europa.
J. Geophy s. Res., 105, 22559–22578.
Greenberg R. (2004) The evil twin of Agenor: Tectonic conver-
gence on Europa. Icarus, 167, 313–319.
Greenberg R. and Weidenschilling S. J. (1984) How fast do Gali-
lean satellites spin? Icarus, 58, 186–196.
Greenberg R., Geissler P., Hoppa G., Tufts B. R., Durda D. D.,
Pappalardo R., Head J. W., Greeley R., Sullivan R., and Carr
M. H. (1998) Tectonic processes on Europa: Tidal stresses,
mechanical response, and visible features. Icarus, 135, 64–78.
Greenberg R., Hoppa G. V., Tufts B. R., Geissler P., and Riley J.
(1999) Chaos on Europa. Icarus, 141, 263–286.
Groenleer J. M. and Kattenhorn S. A. (2008) Cycloid crack se-
quences on Europa: Relationship to stress history and con-
straints on growth mechanics based on cusp angles. Icarus,
193, 158–181.
Han L. and Showman A. P. (2008) Implications of shear heating
and fracture zones for ridge formation on Europa. Geophys.
Res. Lett., 35, L03202, DOI: 10.1029/2007GL031957.
Haeussler P. J., Schwartz D. P., Dawson T. E., Stenner H. D.,
Lienkaemper J. J., Sherrod B., Cinti F. R., Montone P., Craw
P. A., Crone A. J., and Personius S. F. (2004) Surface rupture
and slip distribution of the Denali and Totschunda faults in the
3 November 2002 M 7.9 earthquake, Alaska. Bull. Seismol.
Soc. Am., 94(6B), S23–S52.
Harada Y. and Kurita K. (2006) The dependence of surface tidal
stress on the internal structure of Europa: The possibility of
cracking of the icy shell. Planet. Space Sci., 54, 170–180.
Harada Y. and Kurita K. (2007) Effect of non-synchronous rotation
on surface stress upon Europa: Constraints on surface rheology.
Geophys. Res. Lett., 34, L1 120 4, DOI : 10 .1029 /20 07 GL0 295 54.
Hayne P. O., Sleep N. H., and Lissauer J. J. (2006) Thickness
variations in Europa’s icy shell: Stress and rotation effects. In
Europa Focus Group Workshop Abstracts, Vol. 5 (R. Greeley,
ed.), pp. 53–54. Arizona State Univ., Tempe.
Head J. W., Pappalardo R. T., and Sullivan R. (1999) Europa:
Morphological characteristics of ridges and triple bands from
Galileo data (E4 and E6) and assessment of a linear diapirism
model. J. Geophys . Res. , 104, 24223–24236.
Helfenstein P. and Parmentier E. M. (1980) Fractures on Europa:
Possible response of an ice crust to tidal deformation. Proc.
Lunar Planet. Sci. Conf. 11th, pp. 1987–1998.
Helfenstein P. and Parmentier E. M. (1983) Patterns of fracture
and tidal stresses on Europa. Icarus, 53, 415–430.
Helfenstein P. and Parmentier E. M. (1985) Patterns of fracture
and tidal stresses due to nonsynchronous rotation: Implications
for fracturing on Europa. Icarus, 61, 175–184.
Hoppa G. V. and Tufts B. R. (1999) Formation of cycloidal features
234 Europa
on Europa. In Lunar and Planetary Science XXX, Abstract
#1599. Lunar and Planetary Institute, Houston (CD-ROM).
Hoppa G. V., Tufts B. R., Greenberg R., and Geissler P. E. (1999a)
Formation of cycloidal features on Europa. Science, 285, 1899–
1902.
Hoppa G., Greenberg R., Geissler P., Tufts B. R., Plassmann J.,
and Durda D. D. (1999b) Rotation of Europa: Constraints
from terminator and limb positions. Icarus, 137, 341–347.
Hoppa G., Tufts B. R., Greenberg R., and Geissler P. (1999c)
Strike-slip faults on Europa: Global shear patterns driven by
tidal stress. Icarus, 141, 287–298.
Hoppa G., Greenberg R., Tufts B. R., Geissler P., Phillips C., and
Milazzo M. (2000) Distribution of strike-slip faults on Europa.
J. Geophys. Res., 105, 22617–22627.