B. Randall Tufts*, Richard Greenberg*, Gregory Hoppa*, Paul Geissler*

For submission to Icarus

September 14, 1998

*Lunar and Planetary Laboratory

University of Arizona

Tucson, AZ 85721-0092

Phone: 520-621-6520

Fax: 520-621-5133


Send all correspondence to:

B. Randall Tufts

Lunar and Planetary Laboratory

University of Arizona

Tucson, AZ 85721-0092


Astypalaea Linea is an 810km strike-slip fault, located near the south pole of Europa. In length, it rivals the San Andreas Fault in California, and it is the largest strike-slip fault yet known on Europa. The fault was discovered using Voyager 2 images, based upon the presence of familiar strike-slip features including linearity, pull-aparts, possible braids, and the offset of multiple piercing points. Fault displacement is 42 km, right-lateral. Crosscutting relationships suggest the fault to be of intermediate relative age.

The fault probably initiated as a crack due to tension from combined diurnal tides and nonsynchronous rotation, according to the tectonic model of Greenberg et al., 1998a. Under the influence of varying diurnal tides, strike-slip offset may have occurred through a process called "walking," which depends upon an inelastic lithospheric response to displacement. Because of the presence of possible braids, the chance that fault displacement was caused by simple shear must also be considered, including the possibility that currents in the theorized Europan ocean provided the driving force.

The discovery of Astypalaea Linea extends the range of confirmed lateral motion on Europa. Such motion requires the presence of a decoupling zone of ductile ice or liquid water, a rigid lithosphere, and mechanisms to create and consume surface area.

Key words: Europa; tectonics; satellites of Jupiter; surfaces, satellite.


I. Introduction

Strike-slip faults have been considered rare in the solar system, aside from on Earth (Golombek 1984). However, observations of the Jovian moon Europa demonstrate the lithosphere of that satellite to be cut by numerous examples of shear structures: large strike-slip faults truncate the antijovian wedge-shaped bands (Schenk and McKinnon 1989); a multitude of small strike-slip faults appear in nearly all Galileo high-resolution imaging (Hoppa et al. 1998; Hoppa 1998); and Astypalaea Linea*, near the south pole, is among the largest strike-slip faults anywhere (Tufts 1993, 1996, 1998). Astypalaea's 810 km length and characteristic strike-slip features are comparable to the San Andreas Fault in California. In this paper we describe and characterize Astypalaea Linea in detail, and interpret its tectonic significance.

The Voyager 2 image on which the fault was discovered by Tufts (1993, 1996, 1998) is shown in Figure 1. The possibility that the lineament is a strike-slip fault was suggested because it exhibits what appear to be a well-known strike-slip features (e.g. Crowell 1974; Harding et al. 1985; Sylvester 1988). These features include an elongate linear trace, plus rhomboidal "pull-aparts," and "braids." Also, Astypalaea Linea seems to truncate older lineaments.

Figure 1

Astypalaea Linea is located in the extreme southern hemisphere of Europa. Given our usual perspective, it can be thought of as going "under" the satellite.

Starting at 60°S, 191.1°W the lineament follows a generally curvilinear path passing within ten degrees of the south pole. It reaches at least as far as 78.5°S, 268.4°W where quality of the high-resolution Voyager 2 frame declines. See Figures 2 and 7. Low-resolution images of the trailing hemisphere of Europa taken by Galileo in its second orbit failed to reveal a continuation of the structure. Thus it is not known how far Astypalaea Linea may extend to the west. The northern end of the fault, visible in the Voyager image, falls where the lineament appears to abut the SW-trending gray band Libya Linea. Between these apparent endpoints Astypalaea Linea spans over 29 degrees on the Europan globe.

Prior to its identification as a strike-slip feature, Astypalaea Linea was included in various investigations. Portions of the fault have been mapped in regional studies by Lucchitta et al. (1981), Lucchitta and Soderblom (1982), Schenk and McKinnon (1989) and Pappalardo and Sullivan (1996). Tufts (1993) and Pappalardo and Sullivan (1996) noted a rhomboidal expansion contained in the fault plane (Cyclades Macula, see Fig. 2), likening it to a shear feature.

Figure 2

Lucchitta et al. (1981) and Lucchitta and Soderblom (1982) categorized Astypalaea Linea as a "gray band," a prominent lineament type found in the south polar region. (When it was mapped by Lucchitta and Soderblom (1982) as an albedo unit instead of a structural feature, the fault was categorized as a "brown band" and grouped with "brown spots," creating some ambiguity in terminology.) The gray band classification also encompassed neighboring bands Thynia Linea and Libya Linea. According to Lucchitta et al. (1981) gray bands are relatively old structures, although Pappalardo and Sullivan (1996) noted that the gray bands cut older thin lineaments. Pappalardo and Sullivan (1996) also pointed out that Thynia Linea represents dilation of the lithosphere.

Schenk and McKinnon (1989) grouped the fault with structures making up an extensive "fracture zone" near the antijovian point. This fracture zone is characterized by strike-slip, dilation, and block rotation, recorded in numerous lineaments and wedge-shaped, dilational bands. While not assessing the structural style of Astypalaea Linea specifically, they identified it as a wedge-shaped band as defined by Lucchitta and Soderblom (1982).

Several models have been proposed to explain Astypalaea Linea and associated features. Lucchitta and Soderblom (1982) proposed that gray bands acted as a coherent structural set and are not the product of global stress. The responsible regional stresses may have come from sources that produce circular features on planets, that is, tides, impacts, or volcano-tectonic processes. Schenk and McKinnon (1989) proposed that internal heating is responsible for the wedge-shaped bands (and hence Astypalaea Linea). Carr et al. (1997) and Sullivan et al. (1998) proposed that impacts or regional melting, respectively, might have created both gray bands and wedge-shaped bands. Greenberg et al. (1998a) showed that large-scale tectonic features, including Astypalaea Linea were probably initiated by global tidal stress.

Improved understanding of Astypalaea Linea allows us to interpret the underlying physical processes that drive it. In this paper we describe fault morphology, relief, albedo, crosscutting relationships and related features, plus displacement (or "offset") and relative motion (Sect. II). The fault's structural style and the role of key substructures are characterized in Section III. Next, we describe the relative chronology implied by the fault, and assess the likelihood that Astypalaea Linea is currently active (Sect. IV). Following this descriptive and kinematic discussion, we consider and evaluate genetic models for the shear offset of Astypalaea Linea (Sect. V), including driving by tides and by shear in the underlying medium. Finally, we consider Europan tectonic conditions which are implied by the presence of the fault, including the existence of a subsurface decoupling zone, lithospheric rigidity, and the requirement for mechanisms that create and consume surface area (Sect. VI)

II. Description

A. Morphology, Relief, and Albedo

Astypalaea Linea is broadly curvilinear and generally continuous, although it steps to the right at one or two locations. See Figure 2. The fault trace consists of thin, straight segments alternating with narrow, lens-shaped or irregular expansions. The major irregularity in the fault is the prominent, rhomboidal Cyclades Macula near the fault's northern termination. The fault may include a second, narrow rhomboid (Rs) but its shape is unclear. No en echelon shears, parallel folds, or left steps can be seen associated with the fault. Except in intervals where visibility of the fault is poor, Astypalaea Linea never appears to be less than 5 km wide. Lens-shaped expansions in the fault trace at intervals opposite c' and between a and b' measure 14 km across and 18 km across, respectively. There seem to be no branches or splays, although the southern end of Thasus Linea may merge with the fault.

Cyclades Macula occupies a right step in the fault trace and trends approximately NE, intersecting Astypalaea Linea at roughly a 50 degree angle. Its boundaries appear straight or smoothly curvilinear. The rhomboid links the principal fault trace to its south with a shorter subparallel trace (the "Eastern Trace") which extends north to the intersection with Libya Linea. The principal trace and the Eastern Trace are separated by 47 km at Cyclades Macula and do not overlap. A dimly visible lineament (the "Western Trace") which may be an extension of the principal fault trace, also continues northward from the west edge of Cyclades Macula to join the Eastern Trace where both meet Libya Linea.

Possible rhomboid Rs is located at a narrow step where the fault is intersected by multiple flexii. This rhomboid may be as much as 46 km long and its bounding master faults appear separated by 19 km.

No relief can be seen associated with the trace of Astypalaea Linea despite the low solar incidence angle; the fault casts no shadow and has no discernible central ridge. Even when seen in stereo using different Voyager 2 frames, no relief can be observed.

The fault is usually darker than surrounding terrain. Unlike other lineaments on Europa it seems to have no medial bright stripe or flanking low albedo strips or spots. Its edges seem moderately sharp and fairly well-defined. In Voyager 2 images, no internal albedo patterns are visible except for vague bright regions within Cyclades Macula and some of the lens-like intervals.

B. Crosscutting and Related Structures

Various, prominent, EW-trending flexii cut Astypalaea Linea uninterrupted. These flexii include Cilicia Flexus which crosses north of Cyclades Macula, and Delphi Flexus and Sidon Flexus which cross at the possible rhomboid Rs. See Figure 2.

Various lineae abut the fault: (1) Two dark, NE-trending lineaments intersect the east side of Astypalaea Linea at acute angles. One lineament is dark Thasus Linea and the other is Lineament L; (2) Dark Lineament T', resembling Thasus Linea and trending parallel to it, intersects the west side of Astypalaea Linea not far from where Thasus Linea intersects the fault on the east side; (3) Thin, curved Lineament L' abuts the west side of the fault north of the intersection of Lineament L with the fault; (4) Multiple, NW-trending thin lineaments (a-e, a'-e') abut both sides of the fault and appear truncated by it. While abutting Astypalaea Linea, one of these thin lineae seem to cross it part of the way too. Truncations of lineaments seem confined to the southern half of Astypalaea Linea, although that pattern may be an artifact of image resolution and geometry.

Near its southern end, Astypalaea Linea seems to divide two vague, "pancake-shaped" regions (p, p'). Illumination patterns suggest both these regions are gently elevated.

C. Displacement

The presence of possible strike-slip features and the apparent truncation of various lineaments by Astypalaea Linea suggested that fault displacement might have occurred. To test this hypothesis, we made a palinspastic reconstruction of the fault using an orthographic reprojection of Voyager image 1372J2, shown in Figure 3. Using digital techniques, we translated the territory lying east of the fault 42 km to the north and rotated it 2.5 degrees counterclockwise, resulting in the reconstructed view in Figure 4.

Figure 3
Figure 4

This reconstruction realigned multiple features. Cyclades Macula closed up neatly, measured across its center. The small rhomboid Rs also may have closed, although the presence of crosscutting flexii hinder an interpretation. The reconstruction also matched five pairs of the truncated, mutually similar, thin lineaments that abutted opposite sides of the fault (ae, a'e'), and aligned both Thasus Linea and Lineament L with Lineaments T' and L', respectively. A map depicting the reconstruction is provided in Figure 5. Two notable realignments matched, first, a three-way intersection of Astypalaea Linea, Lineament L, and lineament bb', and second, two thin lineaments that form a small angle of intersection near the southern end of Astypalaea Linea. See Figure 6. With the exception of Lineaments L and L', neither the appearance nor the direction of the various realigned lineaments change significantly as they approach Astypalaea Linea. The two "pancake-shaped" regions near the southern "end" of the fault trace (p,p') may also join although image quality is poor. The reconstruction disrupted Cilicia, Sidon and Delphi flexii which presently cross the fault.

Figure 6

Displacement on Astypalaea Linea at Cyclades Macula and Thasus Linea, and near the southern end of the fault, was 42 km, 41.5 km, and 42.5 km, respectively. These values give an average 42 km displacement for the fault.

The 42 km relative motion vector of the fault is roughly the same as that required to close Libya Linea according to a preliminary reconstruction using Galileo images (E14 orbit). Boundaries of Libya Linea appear to reconstruct well, although the apparent lack of piercing points has hindered past efforts using Voyager 2 images (see also Pappalardo and Sullivan 1996).

(The reconstruction of Astypalaea Linea takes at face value the apparent truncation of the thin lineaments (a-e and a'-e') by the fault. The apparent crossing of the fault by a dim extension of one of these lineae (Sect. II-B) may be attributable to reactivation or may be the result of "pixel mixing.")

D. Relative Motion

The uniformity of displacement along such a lengthy structure suggested that large-scale and coherent relative motion had taken place between the terrain on

opposite sides of it. As a test of this relative motion, we identified an Euler pole of rotation describing the relative motion vector, based on three straight segments of the fault. See Figure 5. We did not include the Eastern Trace, because local effects near the intersection with Libya Linea might have made it unrepresentative of overall displacement on Astypalaea Linea.

Using the approach of Morgan (1967) and Cox and Hart (1986), and confirmed with a least squares technique (G. Hoppa 1997, pers. comm.), we calculated an Euler pole of rotation of 48°S, 247.25° W. The plausibility of this pole can be inferred from the reprojected view centered above it, in Figure 7. This Euler pole is subject to Voyager 2 navigation data, which is accurate to 0.5 degree (P. Geissler 1996, pers. comm.).

Figure 7

This Euler pole differs considerably from the Euler poles describing the relative motions in the antijovian fracture zone of Schenk and McKinnon (1989), and from the center of curvature of gray bands proposed by Lucchitta and Soderblom (1982).

III. Characterization of the Fault

A. Strike-Slip

Astypalaea Linea is a right-lateral strike-slip fault. Reconstruction of the lineament demonstrates a 42 km right-lateral separation that is consistent along its length. Offset is established by seven piercing points, including lineaments of two types which are spaced at irregular intervals and which impinge on the fault at different angles, as well as by the conformable closure of Cyclades Macula. When realigned, the lineaments display a consistent nature and orientation where they cross the fault. Realignment of a three-way and two-way lineament intersection makes the reconstruction particularly convincing.

However, while this evidence of separation strongly suggests that strike slip has occurred on Astypalaea Linea, it is not sufficient to characterize the structure strictly. On Earth, the identification of a major strike-slip fault would be expected to include direct measurements of slip instead of just the apparent offset, that is, separation (Crowell 1962). But on Europa, such measurements are not possible. So, the identification of Astypalaea Linea as a strike-slip fault depends also upon its linearity, the lack of adjacent uplift, and the existence of an Euler pole. Also, lineaments which intersect Astypalaea Linea at acute angles reconstruct coherently, implying that the component of net motion normal to the fault was negligible.

B. The Nature of Cyclades Macula

The rhomboidal Cyclades Macula occupies a dilational gap between en echelon segments of the fault , and can thus be considered a "pull-apart." A step between segments ("overstep") that is in the same direction as the motion of the fault forces the terrain within the step to stretch, hence pulling-apart. Motion opposite to the direction of a step forces the intervening terrain to shorten, forming a "push-up" (e.g. Crowell 1974; Aydin and Nur 1982; Mann et al. 1983).

Cyclades Macula plays an important role in determining the displacement of Astypalaea Linea. This role is based upon the discovery by Schenk and McKinnon (1989) that Europan wedge-shaped bands could be closed by distances equal to their widths. Therefore, the bands must be gaps that opened along vertical tensile cracks, rather than being a graben. If that were true for Cyclades Macula, its length would equal the amount of offset on Astypalaea Linea, and reconstructing it would determine the offset that might realign other piercing points on the master fault. In fact, seven other piercing points literally "fell into line" when Cyclades Macula was closed, suggesting that the pull-apart is bounded by vertical cracks. The observation that the oblique boundaries of Cyclades Macula are smoothly curved brings independent support for this tensile cracking hypothesis.

Pre-existing structures probably had an important effect on Cyclades Macula, for two reasons. First, Cyclades Macula is disproportionately wide by terrestrial standards, which suggest a typical length for pull-aparts of approximately three times their width (Aydin and Nur 1982). The width of a pull-apart is influenced by the initial fault geometry, especially the size of the overstep or bend (Mann et al. 1983). Second, the main fault trace splits at the southern tip of Cyclades Macula and rejoins north of the pull-apart (X, Fig. 2), suggesting that Astypalaea Linea formed by reactivation of pre-existing parallel cracks, similar to terrestrial shear joints (Pollard and Aydin 1988).

Accordingly, based on the consistent magnitude of offset along the fault and at Cyclades Macula, it is likely that (1) Cyclades Macula and Astypalaea Linea are contemporaneous structures and represent equal, coordinated displacement; and that, (2), the Eastern Trace, bounding Cyclades Macula, accommodated the principal offset on the fault at that latitude.

C. Other Expansions of the Fault Trace

The possible rhomboid Rs may also be a narrow pull-apart. This interpretation is supported by the general shape of the feature, the fact that it seems to act as an overstep as shown in Figure 7, and because it is approximately as long as the fault displacement. Like Cyclades Macula, it may have opened by a distance equal to the fault offset, but along more closely spaced master faults.

The lens-shaped expansions in the southwestern part of the principal fault trace may have genetic implications since, in terrestrial strike-slip settings, such features are the product of "simple shear," as described by Sylvester (1988). We discuss this prospect in Section V-B.

IV. Relative Chronology and the Question of Current Activity

Astypalaea Linea establishes the basis for a regional relative chronology because its crosscutting relationships over a wide area divide the timescale in two. Because they are cut by the fault, Thasus Linea, Lineament L-L', and the thin lineaments all predate it. Multiple flexii cross Astypalaea Linea and so postdate it. Features cut by the fault were not included in the chronology developed by Lucchitta and Soderblom (1982). In contrast to their interpretation, the fault (as a gray band) is not among the oldest structures in the region.

Because there seem to be no intersecting lineae that mark an offset less than 42 km, none of them formed during the active faulting stage. This lack suggests either that Astypalaea Linea developed quickly relative to the timetable for creation of lineae or that there was a significant time gap between the formation of the older lineae and the younger flexii. However, if Cyclades Macula was linked to the motion of Astypalaea Linea, parallel linear features may have formed in its interior resembling the striations now seen in Galileo images of wedge-shaped bands (Sullivan et al. 1998, Tufts et al. 1997a). Such striations could provide a "paleoseismological" record of offsets on Astypalaea Linea.

Astypalaea Linea is probably not active because it appears that various flexii cut the fault. However, the last displacement may have been recent, if, as suggested by Greenberg et al. (1998a), ridges can be formed quickly. Also, if Europan tectonic regimes can overlap in time, the fault might still be active. If so, small offsets of the prominent flexii might be apparent in high-resolution images. A dark lineament extending from the north end of the fault into Libya Linea may record the most recent activity of Astypalaea Linea (Tufts 1998).

V. Genetic models

A. Tidal Stress

Greenberg et al. (1998a) has proposed that tidal stress can account for the existence of Astypalaea Linea, based upon first forming the fault as a tension crack, then reactivating it in shear. Given the low tensile strength of Europan ice (estimated by Tufts 1998) versus its shear strength, the fault is more likely to have initiated as a crack than as a shear fracture.

To determine if the magnitude and configuration of combined (diurnal and nonsynchronous rotation) stresses favor initiation of Astypalaea Linea as a crack, Greenberg et al. (1998a) compared the alignment of the fault to these stresses during the period in the daily orbit of maximum tension. At that point (5/8 orbit past pericenter) the amount of tension is probably sufficient to crack the ice, and the principal tensile axis is roughly perpendicular to the fault trace. See also Hoppa (1998).

Therefore, this stress configuration probably is responsible for initial cracking of Astypalaea Linea. However, the tensile is not perfectly perpendicular to the fault trace. One explanation arises from the possibility that prograde nonsynchronous rotation may be taking place (Greenberg et al. 1998a). The orientation of the stresses becomes more favorable if the fault is located farther to the east, that is, in the future. So, Astypalaea Linea may have formed on an earlier rotation (cf. Greenberg et al. 1998a). This history would be consistent with the presence of crosscutting flexii. In addition, the fault is circular (as demonstrated by the existence of an Euler pole) and tidal stress trajectories are not (Schenk 1996). Another explanation may be that propagation of the crack was slow, allowing the diurnal rotation of the principal tidal stress axes to alter the radius of curvature of the crack as it formed.

A second problem with the fit is that the orientation of the Eastern Trace of the fault is not well matched to the stress, and stress magnitudes there are slightly lower there than for the fault as a whole. But as noted earlier, the Eastern Trace may be influenced by its intersection with Libya Linea (Tufts 1998).

Tides may also have been responsible for the 42 km dextral displacement of the fault, using the mechanism of "walking" proposed conceptually by Tufts et al. (1997b), and quantified by Hoppa et al. (1998). Once cracking takes place, strain due to nonsynchronous rotation is relieved so that the subsequent strain experienced at the crack is due to varying diurnal tides. Alternating tension and compression directed normal to the fault cause it to open and close daily. These stress regimes are followed by right and left shear stress, respectively. Inelastic behavior of the lithosphere allows a net right-lateral displacement to result when the fault opens after the left shear phase. Even if incremental motion is small, walking could accumulate considerable offset if it continued for a long period. For example, a net movement of one-tenth of a millimeter each Europan day would result in 42 km of offset in 4.2 million years. Hoppa et al. (1998) find evidence for walking in the statistical over-representation of left-lateral and right-lateral strike-slip faults in the northern versus southern Europan hemispheres, respectively.

Walking may owe its apparent plausibility at Astypalaea Linea to the fact that the fault has no discernible restraining bends. To overcome such bends by forming push-ups would require higher stress than is generated by diurnal tides.

B. Fault Genesis by Simple Shear?

As an alternative to the Greenberg et al. (1998a) model for the formation of the fault, in this section we examine whether displacement on Astypalaea Linea came about due to an imposed shear from an underlying medium, forming simple shear structures. Simple shear is a well-understood kinematic process associated with the largest strike-slip faults on the Earth (Sylvester 1988). It is important in this study because it would imply that the immediate drivers of the strike-slip displacement of Astypalaea Linea are different than conceived in the tidal model described in the previous section. See also Tufts (1998).

The considerable length of Astypalaea Linea, lack of a conjugate fault, and the presence of expansions which can plausibly be interpreted as "braids" suggests an origin by simple shear (e.g. Sylvester 1988; Naylor et al. 1986; Woodcock and Fischer 1986). In Figure 5, lens-like expansions B1 and B2 may be braids. With simple shear faulting, braids are caused by the interaction of secondary Reidel and "P" shears which form in a predictable sequence and are then reoriented progressively as relative movement of the opposing sides of the fault continues.

Pure shear does not create such patterns, nor does it create most major terrestrial strike-slip faults - contrary to the popular faulting theory of Anderson (1951). Pure shear faults are rarely long, and rarely accommodate large displacements. The mechanics of finite pure and simple shear are "vastly"

different (Sylvester 1984).

If Astypalaea Linea is a simple shear fault, what could drive it? The uniform magnitude of offset along Astypalaea Linea suggests the relative motion of large blocks. For such motions to occur suggests that a large-scale force is at work. But, the familiar terrestrial plate tectonic drivers are not present, given the muted topography of the Europan globe and resultant lack of potential energy accumulations.

Assuming the fault initiated as a tensile crack, as proposed by Greenberg et al. (1998a), currents in the theorized Europan ocean (e.g., Cassen et al. 1979; see also Tufts 1998)may have brought about subsequent relative motion, rather than tidally-induced "walking." In this way, the moving ice lithosphere would be somewhat like the clay slab in laboratory demonstrations of simple shear (e.g. Wilcox et al. 1973), except that the fault pre-existed as a crack. This model would explain the length of the fault, its lack of conjugate, and its circularity as the products of tidally driven stress which creates single, long cracks in a rotating stress field. But, importantly, it could account for simple shear structures that might be present, such as braids, if there was at least some frictional coupling between opposing sides of the fault.

C. Distinctions between Fault Models

Both the tidal and simple shear models start with an initial cracking phase before reactivation and along-strike motion begins. Thus the distinguishing structures would be those associated with the generation of strike slip.

The tidal tectonic model might exhibit the following features. First, the "braids" would not be simple shear structures. Instead, they might be preexisting bands or lineaments cut by the main fault and which reactivated and opened when fault displacement occurred, perhaps forming small pull-aparts that resemble braids at low resolution. Second, the subsequent strike-slip might be recorded by a damaged zone caused by repeated opening and closing of the fault. Since walking rather than slipping is implied, a gouge zone might be absent or relatively undeveloped.

The simple shear model would include familiar simple shear features, perhaps superimposed on a partially annealed zone. Varying degrees of annealing might be reflected in different widths of influence of the simple shear features. A strong coupling might cause Reidel shears and other structures to extend into the icy blocks themselves. Weaker annealing might result in simple shear taking place only in a narrow zone within the displacement zone. An example of this morphology on Europa may exist near Conamara Chaos where J. Moreau (pers. comm. 1997) has identified strike-slip on a furrowed lineament. Another analogy might be the linear shear zones created in terrestrial sea ice when blocks re-abut after cracking apart, only to begin moving past each other (Greeley et al. 1997).

A useful comparison may be provided by the "Great Ice Chasm," a large crevasse in the Filchner Ice Shelf discovered by the Commonwealth Trans-Antarctic Expedition of 1955-1958 (Wilson 1960). This temporary feature, thought to represent simple shear (Burke et al. 1980), exhibited sigmoidal cracks and rotated blocks in its depressed central band (Sylvester 1988). The Filchner Ice Shelf is a few hundred meters thick, a value closer to that expected for the Europan ice shell (Tufts 1998).

VI. Tectonic Implications of Lateral Motion at Astypalaea Linea

The discovery of Astypalaea Linea extends the range of confirmed lateral motion to near the south pole, at the edge of the fracture zone identified by Schenk and McKinnon (1989), consistent with the proposal by Pappalardo and Sullivan (1996) that lateral motion is a global process. In addition, the horizontal displacement that the fault represents covers a broad area.

Lithospheric mobility implied by the fault requires these boundary conditions, based on tectonic theory (Engelder 1993): (1) the presence of a basal shear or decoupling zone which accommodates horizontal displacement of overlying material (the fault presumably penetrates to the decoupling layer);

(2) rigidity great enough to transmit stress and to localize failure at terrane margins; (3) a means of creating surface area; and (4) a means of consuming surface area.

The decoupling layer below the lithosphere probably corresponds to a liquid layer (e.g. Cassen et al. 1979), if any, or to a weak zone within ductile ice (Lucchitta and Soderblom 1982; Schenk and McKinnon 1989). Decoupling occurs where the yield strength has diminished sufficiently from the higher values in the lithosphere (Golombek and Banerdt 1990). This decoupling layer may have allowed the overall inelastic lithospheric response needed for walking. Data from Astypalaea Linea alone is not sufficient to distinguish between a liquid or ductile decoupling zone. Tufts (1998) discusses the problem of making a distinction between a liquid or ductile decoupling layer, using additional evidence.

In forming Astypalaea Linea, the lithosphere there was sufficiently rigid to localize deformation over a long distance. The two sides of Astypalaea Linea behaved as brittle blocks, moving past one another along a relatively narrow and well-defined zone. Consistent offset along the length of the fault implies a minimum amount of inelastic strain within the ice when the fault was active. The simple geometric pattern of Astypalaea Linea suggests that the mechanical properties of the lithosphere were laterally uniform on the scale of the fault. It is notable that the south polar surface temperature would be relatively low and therefore the lithosphere (overlying a decoupling zone) there might be thick compared to other regions of the satellite (Ojakangas and Stevenson 1989).

However, because of the cracking that apparently preceded development of the fault, the coherence of plates and blocks during the fault's displacement implies that an annealing or strengthening process had healed the earlier breaks. Astypalaea Linea did not seem to reactivate two prominent lineae transecting it - Lineament L and Thasus Linea. Also, tensile failure at Cyclades Macula would have been unlikely if the lithosphere was cut by cracks that were still active.

A likely mechanism for annealing is the freezing of water that has filled a crack, perhaps according to the Greenberg et al. (1998a) ridge-building model. Such freezing might occur if the intensity of tidal "working" of a crack diminished due to nonsynchronous rotation of the satellite. Strengthening might also occur if lithospheric blocks crowd together by walking. If the blocks were irregularly curved, a small rotation around a vertical axis might lock them in place. A third possible mechanism may be sintering, however little work on this process seems to have been done for ice at Europan temperatures, and the effect of contaminants is unknown.

To the extent that Astypalaea Linea is linked to dilational bands it may be involved in producing new surface area. For example, a small amount of new terrain has appeared at Cyclades Macula. Likewise, the motion of Astypalaea Linea may be accommodated by opening of Libya Linea (Tufts 1998). Such dilation would be driven by the progressive displacement of the fault and would probably involve sequential cracking and spreading, followed by an influx of mobile material from below.

An increase in surface area has to be accommodated by a process which consumes surface area. But subduction zones do not appear to be present on the parts of Europa so far seen (cf. Schenk and McKinnon 1989). However, approaching the matter as an issue in surface area balance, rather than a search for familiar terrestrial tectonic features, makes it possible to identify constraints on possible consumption mechanisms. Consumption processes must: (1) remove sufficient old surface area to account for the production of new surface area; (2) operate in the same time interval as the dilation to be accommodated; (3) be structurally compatible with the dilation which creates the new surface area; and (4) affect the whole thickness of the lithosphere.

Considering these constraints, a likely accommodation mechanism is the formation of lenticulae and chaos regions, as proposed by Greenberg et al. (1998b). See also Sullivan et al. (1998) and Tufts (1998). If lenticulae and chaos regions form by melting, then volume may be lost as ice liquefies and ice blocks founder. The newly created lateral free surface might allow an inward movement of surrounding ice.

Because chaos regions and lenticulae occur in discrete events, their suitability as a consumption mechanism depends on the timing of their occurrence. Lenticulae and chaotic terranes may represent late-stage events (Pappalardo et al. 1998), in which case they could not be associated with old dilation. However, Greenberg et al. (1998b) propose that they have occurred through geologic time. Given their widespread occurrence, their development may be structurally compatible with widely distributed bands and dilational ridges (Greenberg et al. 1998a; Tufts 1998). Furthermore, assuming that they represent melting, formation of chaos regions and lenticulae would affect the whole thickness of the lithosphere. The extensive area that they seem to cover suggests that they could easily account for the surface generated by dilation at Cyclades Macula and elsewhere.

VII. Conclusions and Future Studies

The following conclusions can be drawn from the findings reported here:

(1) Astypalaea Linea is a strike-slip fault near the Europan south pole. It is over 810 km in length, comparable to the San Andreas Fault in California, with 42 km of right-lateral displacement. Fault motion is described by an Euler pole located at 48°S, 247.5°W;

(2) The fault contains a large pull-apart, Cyclades Macula, which spans 47 km of separation between master faults, and may contain another smaller pull-apart;

(3) The trace of Astypalaea Linea also contains lensoid expansions of unknown origin. These expansions may be braids developed by simple shear;

(4) Crosscutting relationships associated with the fault establish a relative chronology in which the fault is of intermediate age;

(5) Astypalaea Linea most likely initiated as a tensile crack caused by a combination of diurnal and nonsynchronous rotation stress, according to the model of Greenberg et al. (1998a);

(6) Subsequent strike-slip displacement probably occurred due to "walking" - a stepwise motion induced by changing diurnal tides and dependent upon inelastic lithospheric behavior.

(7) Due to the possible presence of braids, the hypothesis that displacement of Astypalaea Linea was caused by simple shear must also be considered. Such motion may have been driven by subsurface ocean currents which created shear along a pre-existing tidally-induced crack;

(8) Lateral motion evidenced by the fault implies the presence of a subsurface decoupling zone, a rigid lithosphere, and mechanisms for creating and consuming surface area. Data presented here is not sufficient to determine whether the decoupling zone is liquid and ductile. Such a zone may have allowed the inelastic behavior necessary for walking to take place. Lithospheric rigidity might be maintained by freezing of water in cracks, locking-up by walking, or by sintering. Dilation at Cyclades Macula produces new surface area. Formation of chaos regions and lenticulae may provide mechanisms for removing surface area.

Astypalaea Linea is important from the standpoint of Europan geology but also because it may provide a useful analog to terrestrial faults. Future studies of the fault should include geologic mapping and structural analyses aimed at confirming the structural style, and assessing age and slip history. Arcuate ridges which apparently cut the fault can be examined for offsets suggesting relatively recent fault activity. Detailed mapping of the displacement zone and adjacent terrain may help determine the stress orientations near the fault and the strength of the fault plane, and may provide a test of "walking." Examining the fine-scale morphology of the fault plane may determine if it exhibits the deformation we predict if walking were taking place, or if simple shear had occurred.

*Astypalaea Linea is named for the sister of Europa, according to Greek myth.


Anderson, E. M., 1951, The dynamics of faulting and dyke formation with applications to Britain: Edinburgh, Oliver and Boyd, 206 p.

Aydin, A., and Nur, A., 1982, Evolution of pull-apart basins and their scale independence: Tectonics 1, no. 1, p. 91-105.

Burke, K., Grippi, J., and Sengor, A. M. C., 1980, Neogene structures in Jamaica and the tectonic style of the Northern Caribbean boundary zone: Journal of Geology 88, p. 375-386.

Carr, M., Chuang, F., Belton, M., Bender, K., Chapman, C., Greeley, R., McEwen, A., Sullivan, R., and the Galileo Imaging Team, 1997, A large multi-ringed structure on Europa: a possible remnant of an impact basin buried beneath the ice crust, in Lunar and Planetary Science Conference XXVIII, Houston, p. 207-208.

Cassen, P., Reynolds, R. T., and Peale, S. J., 1979, Is there liquid water on Europa?: Geophysical Research Letters 6, no. 9, p. 731-733.

Cox, A., and Hart, R. B., 1986, Plate tectonics: how it works: Boston, Blackwell Scientific Publications, Inc., 392 p.

Crowell, J. C., 1962, Displacement along the San Andreas fault, California: Geological Society of America Special Paper 71.

Crowell, J. C., 1974, Origin of late Cenozoic basins in southern California, in Dickinson, W. R., ed., Tectonics and sedimentation: Tulsa, Oklahoma, Society of Economic Paleontologists and Mineralogists, p. 195-209.

Engelder, T., 1993, Stress regimes in the lithosphere: Princeton, Princeton University Press, 457 p.

Golombek, M. P., 1984, Why are there no strike-slip faults on the planets and satellites?, in Lunar and Planetary Science Conference XV, Houston, p. 306-307.

Golombek, M. P., and Banerdt, W. B., 1990, Constraints on the subsurface structure of Europa: Icarus 83, p. 441-452.

Greeley, R., Sullivan, R., Coon, M., Geissler, P., Tufts, B. R., Head III, J. W., Pappalardo, R. T., and Moore, J., 1997, Terrestrial sea ice morphology: considerations for Europa: IcarusSubmitted.

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., 1998a, Tectonic processes on Europa: tidal stresses, mechanical response, and visible features: Icarus In press.

Greenberg, R., Hoppa, G., Tufts, R., and Geissler, P., 1998b, Chaos regions: widespread melt-through to the surface of Europa?: Bulletin of the American Astronomical Society. Submitted.

Harding, T. P., Vierbuchen, R. C., and Christie-Blick, N., 1985, Structural styles, plate tectonic settings, and hydrocarbon traps of divergent (transtensional) wrench faults, in Biddle, K., and Christie-Blick, N., eds., Strike-slip deformation, basin formation, and sedimentation: Tulsa, Oklahoma, Society of Economic Paleontologists and Mineralogists, p. 51-77.

Hoppa, G. V., 1998, Europa: effects of rotation and tides on tectonic processes [Doctoral thesis]: University of Arizona.

Hoppa, G., Tufts, B. R., Greenberg, R., and Geissler, P., 1998, Strike-slip faults on Europa: global shear patterns driven by tidal stress: Icarus Submitted.

Lucchitta, B. K., and Soderblom, L. A., 1982, The geology of Europa, in Morrison, D., ed., Satellites of Jupiter: Tucson, University of Arizona Press, p. 521-555.

Lucchitta, B. K., Soderblom, L. A., and Ferguson, H. M., 1981, Structures on Europa, in Lunar and Planetary Science Conference, p. 1555-1567.

Mann, P., Hempton, M. R., Bradley, D. C., and Burke, K., 1983, Development of pull-apart basins: Journal of Geology 91, p. 529-554.

Morgan, W. J., 1968, Rises, trenches, great faults, and crustal blocks: Journal of Geophysical Research 73, no. 6, p. 1959-1982.

Naylor, M. A., Mandl, G., and Sijpsteijn, C. H. K., 1986, Fault geometries in basement-induced wrench faulting under different initial stress states: Journal of Structural Geology 8, no. 7, p. 737-752.

Ojakangas, G. W., and Stevenson, D. J., 1989, Polar wander of an ice shell on Europa: Icarus 81, p. 242-270.

Pappalardo, R. T., and Sullivan, R. J., 1996, Evidence for separation across a gray band on Europa: Icarus 123, p. 557-567.

Pappalardo, R. T., Head, J. W., Greeley, R., Sullivan, R. J., Pilcher, C., Schubert, G., Moore, W., Carr, M., Moore, J. M., Belton, M. J. S., and Goldsby, D. L., 1998, Geological evidence for solid-state convection in Europa's ice shell: Nature 391, p. 365-368.

Pollard, D. D., and Aydin, A., 1988, Progress in understanding jointing over the past century: Geological Society of America Bulletin 100, p. 1181-1204.

Schenk, P. M., 1996, Wedge-shaped bands and pulling-apart structures on Europa: evidence for lithospheric decoupling and a possible ocean, in Europa Ocean Conference, San Juan Capistrano, California, p. 62.

Schenk, P. M., and McKinnon, W. B., 1989, Fault offsets and lateral crustal movement on Europa: evidence for a mobile ice shell: Icarus 79, no. 1, p. 75-100.

Sullivan, R. J., Greeley, R., Homan, K., Klemaszewski, J., Belton, M. J. S., Carr, M. H., Chapman, C. R., Tufts, R., Head III, J. W., Pappalardo, R. T., Moore, J., Thomas, P., and the Galileo Imaging Team, 1998, Episodic plate separation and fracture infill on the surface of Europa: Nature 391, p. 371-373.

Sylvester, A. G., 1984, Wrench fault tectonics: Tulsa, Oklahoma, American Association of Petroleum Geologists.

Sylvester, A. G., 1988, Strike-slip faults: Geological Society of America Bulletin 100, p. 1666-1703.

Tufts, B. R., 1993, Photogeological analysis of Europan tectonic features, in Lunar and planetary science conference XXIV, Houston, p. 1445-1446.

Tufts, B. R., 1996, A San Andreas-sized strike-slip fault on Europa, in Lunar and Planetary Science Conference XXVII, Houston, p. 1343-1344.

Tufts, B. R., 1998, Lithospheric displacement features on Europa and their interpretation [Doctoral thesis]: University of Arizona, 288 p.

Tufts, B. R., Greenberg, R., Sullivan, R., Pappalardo, R., and the Galileo Imaging Team, 1997a, Reconstruction of Europan terrain in the Galileo C3 "Wedges" image and its geological implications, in Lunar and Planetary Science Conference XXVIII, Houston, p. 1455-1456.

Tufts, R., Greenberg, R., Geissler, P., Hoppa, G., Pappalardo, R., Sullivan, R., and the Galileo Imaging Team, 1997b, Crustal displacement features on Europa: Bulletin of the American Astronomical Society 29, no. 3, p. 983.

Wilcox, R. E., Harding, T. P., and Seely, D. R., 1973, Basic wrench tectonics: American Association of Petroleum Geologists Bulletin 57, no. 1, p. 74-96.

Wilson, G., 1960, The tectonics of the "Great Ice Chasm", Filchner Ice Shelf, Antarctica: Proc. Geol. Assoc. 71, p. 130-138.

Woodcock, N. H., and Fischer, M., 1986, Strike-slip duplexes: Journal of Structural Geology 8, no. 7, p. 725-735.