*Lunar and Planetary Laboratory

University of Arizona

Tucson, AZ 85721-0092
 
 
 
 

Phone: 520-626-3154

Fax: 520-621-5133

email: rtufts@pirl.lpl.arizona.edu
 
 

Keywords: Europa, tectonics, satellites of Jupiter, satellite, surfaces.
 
 
 
 

Send all correspondence to:

B. Randall Tufts

Lunar and Planetary Laboratory

University of Arizona

Tucson, AZ 85721-0092
 
 

Lithospheric dilation on Europa has occurred at Class 2 ridges (Greenberg taxonomy) and at bands, on a global scale over a large part of the geological age of the surface. Class 2 ridges are elevated, a few km across, and bilaterally symmetrical about a pronounced central groove. Bands are lower than Class 2 ridges, and may be wider than a few km, with neither a prominent central groove nor strong bilateral symmetry. Some lineaments combine characteristics of both Class 2 ridges and bands. The character of Class 2 ridges, bands, and the intermediate forms suggests that they all are dilational gaps in the Europan lithosphere and reflect a continuum of formation process between end members, namely: (1) outward pushing by material lodged in cracks (Class 2 ridges), and (2) pulling-apart of lithospheric plates by any of various possible forces (bands). Variation in morphology among dilational lineaments depends upon the degree to which pulling-apart counteracts tidal compression. On a global scale, dilation may be accommodated by melting-through of lithosphere at various times and places, creating chaotic terrain. Dilational lineaments represent an important agent of resurfacing on Europa.
 
 

I. INTRODUCTION

Lateral displacements of the lithosphere have been discovered at multiple locations on Europa (Schenk and McKinnon, 1989; Pappalardo and Sullivan, 1996; Tufts, 1996; Tufts et al., 1997a; Prockter et al., 1998b; Sullivan et al., 1998a; Tufts, 1998; Tufts et al., 1998b; Hoppa et al., 1999). Generally, lateral motion has consisted of translation and rotation of lithospheric blocks that range from a few kilometers to hundreds of kilometers across. The motion has been accompanied by the opening of dilational lineaments or by shear on strike-slip faults. Such horizontal displacements imply the existence of a subsurface decoupling zone to which cracks and faults that bound the mobile blocks penetrate, allowing the lateral motion (Schenk and McKinnon, 1989; Golombek and Banerdt, 1990). The Europan lithosphere may be defined as the rigid, mobile ice layer lying above the decoupling zone (Schenk and McKinnon, 1989; Golombek and Banerdt, 1990; Tufts, 1998). Lateral displacements also imply that as lithospheric blocks separate and create new surface area, pre-existing surface must be consumed elsewhere. In this paper we develop a kinematic and dynamic model for the formation of lineaments that record dilation and the consequent creation of new surface area on the satellite. As a basis for this model we document occurrences of these lineaments, describe their morphology, and confirm their structural style. We then discuss their tectonic significance in view of the model we propose.

Dilational lineaments are part of a larger array of Europan lineae which crisscross the satellite, discovered by the Voyager spacecraft (Smith et al., 1979; Lucchitta and Soderblom, 1982), and viewed at higher resolution by the Galileo spacecraft (Belton et al., 1996; Greeley et al., 1998; Greenberg et al., 1998a). Generally, Europan lineaments have been thought to initiate by tensile cracking, in many cases due to global stress (Lucchitta and Soderblom, 1982). A source of such stress may be a combination of diurnal tides and nonsynchronous rotation (Greenberg et al., 1998a). Global lineaments have now been shown to be large examples of ubiquitous ridge systems that cover the satellite at a wide range of scales.

Greenberg et al. (1998a) explain ridge morphologies by extrusion of crushed ice resulting from the repeating sequence of tidally-driven opening, water-filling, and closing of cracks between the surface and a liquid decoupling layer. Ballistic eruption of liquid water exposed in the cracks may contribute to the development of ridges (Kadel et al., 1998). Lateral force against the walls of cracks might also distort the lips, contributing to ridge growth (Sullivan et al., 1998b; Turtle et al., 1998). These models are probably incompatible with ridge formation by linear diapirism, as proposed by Head et al. (1997b) and Pappalardo et al. (1997; 1998b).

Dilation may accompany ridge formation when ice or slush accumulates in the cracks, forcing the opposing blocks apart when daily compressive tidal stress is applied, according to Greenberg et al. (1998a). Greenberg et al. (1998a) proposed that such a dilational process may have formed the class of ridges they defined as Class 2. Such ridges are defined morphologically in the Greenberg et al. (1998a) taxonomy, independent of a kinematic and genetic model: they are elevated, frequently in a platform-shape, internally striated, and usually strongly bilaterally symmetrical with a pronounced medial groove. Through a palinspastic reconstruction (Sec. II) we demonstrate that an archetypal example of a Class 2 ridge is, in fact, dilational. We discuss its mode of formation in light of the Greenberg et al. (1998a) model and discuss how that model results in various morphological details of Class 2 ridges.

Several bands have also been shown to be dilational (Schenk and McKinnon, 1989; Pappalardo and Sullivan, 1996; Tufts, 1996; Tufts et al., 1997a; Sullivan et al., 1998a; Tufts, 1998; Tufts et al., 1998c), and it is likely that most are. Bands are morphologically distinct from Class 2 ridges. They are linear or curvilinear strips, generally wider than Class 2 ridges, sometimes up to a few tens of kilometers across. They may be wedge-shaped or of variable width. Bands seem to be topographically lower than Class 2 ridges. When seen at sufficient resolution, they exhibit internal subparallel striations. This array of internal linear features is roughly bilaterally symmetrical but not as regular as ridges. It is unclear whether bands have central grooves; if so, the central grooves are not as pronounced as the grooves associated with ridges. In Section III we describe the morphology and the displacement of nine Europan bands, and discuss their kinematic character. We then present a formational model which accounts for their morphology and kinematics, and which distinguishes them from Class 2 ridges.

Some lineaments exhibit morphological characteristics of both ridges and bands, as described in Section IV.A. These lineaments include: (1) bands that are bordered by pairs of single ridges; (2) bands that are elevated; and (3) lineaments that are wide, low but in some ways resemble Class 2 ridges. Based upon the existence of these mixed cases and on the characteristics of bands and Class 2 ridges, we propose (Sec. IV.B) a process of forming the range of dilational lineament morphologies, with Class 2 ridges and bands as end-members. In this model, Class 2 ridges, bands, and intermediate structures are variations on the tidal tectonic theme (Greenberg et al., 1998a), representing a continuum of products resulting from a generalized formation mechanism which includes tensile and compressive tidal stress.

Last, we examine the global tectonic significance of dilational lineaments (Sec. V). We consider the contribution of dilation toward resurfacing the satellite, and discuss implications of accommodating dilation by development of chaos regions.

In this paper we use images returned by Galileo through the E17 orbit.

II. CLASS 2 RIDGES

A. Distribution of Class 2 ridges

Class 2 ridges are distributed widely on Europa. They can be seen in most high-resolution images from the Galileo spacecraft. The Class 2 ridge we investigate in the next section is located on the trailing side of the satellite. Various other examples (Fig. 1) include a Class 2 ridge located in the antijovian region (also in Fig. 4a,b; for a stereo view see Figure 15), which strongly resembles the ridge we examine below. Other Class 2 ridges are found near the "E11 band" on the leading side of Europa (Fig. 9a), and on the trailing side, near Androgeos Linea. An older lineament that is cut by the ridge we examine in the next section is also a Class 2 ridge.

Figure 1

B. "Ridge C2a"

Figure 2a

Beginning at the edge of the main ridge and extending as far out as one kilometer, there is a series of cracks that are roughly parallel to the main ridge despite some local irregularity. These cracks cut older ridges that are cut by the main ridge, suggesting they are the same relative age as the main ridge. Also adjacent to the main ridge on both sides are isolated regions of flat, smooth terrain. This terrain embays older lineaments as well as most of the flanking cracks associated with the main ridge.

Figure 2b

However, the reconstruction does have some apparent imperfections. First, it leaves a residual elevated ridge-like strip. Because this strip has a medial crack it may be part of an older, more sinuous ridge or it may be a remnant of Class 3-type flanking ridges. Second, some older lineaments in the eastern part of the image do not realign completely. This residual misalignment may be due to mechanical errors in the reconstruction or to the presence of unrelated deformation.

The geometry of the reconstruction shows that Ridge C2a represents a gap created by opening of the lithosphere. The crack that opened was probably vertical given its straight path across the landscape. (This configuration would also be consistent with formation by horizontal tensional stress.) Realignment of older lineaments was only achievable by fully closing the ridge. Errors in the reconstruction would involve not closing it enough. Thus, the ridge cannot represent a graben, because fault blocks would prevent full closure (see also Schenk and McKinnon(1989)). With a graben, only a small closure would realign the offset lineae. The resulting gap was filled with fresh surface material that protrudes upward to a level higher than the ambient terrain. The demonstrable character of this ridge provides a type example of Class 2 ridges in the taxonomy of Greenberg et al. (1998a) (which is why we call it "Ridge C2a").

C. Formational Model

In the model of Greenberg et al. (1998a), Class 2 ridges develop if some of the ice that forms during the tensional tidal phase gets stuck in the central crack. This accumulation prevents the crack from closing during the subsequent compressional phase and, in effect, pushes the adjacent lithospheric blocks apart. Note that material is not injected forcibly into the crack, prying the plates apart, but rather, the crack wall in Class 2 ridges experiences an outward pushing force in the compressional phase due to the presence of new material lodged within it. The active crack dilates in a ratcheting motion with repeated openings and closings. In the process, successive paired ridges accumulate in the widening gap between initial ridge-pairs, if any. Tidal compression associated with this forced separation of opposing plates shoves the new interior ridges upward. This shoving results in an elevated surface with a parallel lineated fabric that symmetrically flanks the active crack. Because the ridge is elevated, a groove develops at the active crack as some of the newly deposited ice falls back into the crack when it periodically opens. This groove may be maintained in the same relative location by compressional consolidation of the crack walls. The hummocky surface of Class 2 ridges may result from crushing in the compressional phase.

The dilational character of the ridge and the features derived from that separation fit the formation model of Greenberg et al. (1998a). The central groove appears to be the location of the most recent cracking and the axis of spreading. Located outboard of the midline is each half of the original doublet ridge. The inner slopes of these original half ridges form the twin lines used to bound the reconstruction that was performed. The reconstruction suggests that those inner slopes once bordered the midline crack before spreading began. Lineated terrain between those two grooves is formed by the new material emplaced from below (Greenberg et al., 1998a). All these characteristics fit the definition of Class 2 ridges by Greenberg et al. (1998a) and their model for their formation.

Ridge C2a also exhibits features consistent with downward, ridge-induced flexing (Pappalardo and Coon, 1996; Greenberg et al., 1998a). Outboard of the ridge are crack sets that might result from such flexure. Lateral smooth deposits may have been formed by flooding as the weight of the ridge forced the plate down below a shallow water line (Greenberg et al., 1998a). The alternative or accompanying process of ballistic spray (Kadel et al., 1998) could only have happened early because it did not darken the ridge structure.

In most settings, extensional tectonics produces crustal thinning. However, in this case, dilation is associated with apparent thickening. This dilemma is resolved by the Greenberg et al. (1998a) ridge-building model. The ice slurry which has filled the gap from below is shoved upward by repeated daily compression, forming an elevated platform. At the same time, the opposing blocks move farther apart, pushed outward by the increasing volume of material in the gap. Lateral motion is accommodated by the subsurface decoupling layer.

III. BANDS

A. Distribution of Bands

Bands were noted in Voyager images and first examined by Pieri (1981) and Lucchitta and Soderblom (1982). Schenk and McKinnon (1989), Tufts (1993), Pappalardo and Sullivan (1996), and Tufts (1996) have investigated specific bands in the antijovian and south polar regions using Voyager images. Higher resolution Galileo data has enabled more detailed assessments to be made of Europan bands (Sullivan et al., 1996a; Sullivan et al., 1996b; Sullivan et al., 1997; Tufts et al., 1997a; Tufts et al., 1997b; Prockter et al., 1998a; Sullivan et al., 1998a; Tufts et al., 1998a; Tufts, 1998; Tufts et al., 1998c).

Bands are widely distributed on Europa. The range of cases we cite here extends to all regions except the subjovian hemisphere and the north polar region, both of which have not yet been imaged at better than a few kilometers per pixel resolution. Bands are apparent in most medium- to high-resolution Galileo images. In addition, they may have occurred at various times in Europan geologic history, as suggested by examples of multiple overlapping bands. Many older bands have been disrupted by tectonic activity or muted by geomorphic processes but still retain sufficient topographic or albedo signature to be identified.

In addition to the bands we examine in detail here, other examples of bands include: (1) a sickle-shaped band south of the crater Manannan, and an old band which it overlaps (Fig. 3); (2) a long band on the leading side (Fig. 3); (3) multiple band fragments near Band B (Sec. III.B.1) in the antijovian wedges region (L. Prockter , 1998, pers. comm.); (4) Agenor Linea, a prominent bright band crossing the antijovian hemisphere (Geissler et al., 1998b). In addition, an IR-bright feature near the intersection of Cadmus and Minos lineae may be an ancient band (Geissler et al., 1998a).

Figure 3

B. Description of Bands

Figure 4a
Figure 4b

Bands B, C, and D contain internal striations of contrasting light and dark reflectance, which may be due to albedo or may be topographic in nature. Band B appears to have a central axis (Sullivan et al., 1998a), although its boundaries are somewhat indistinct. Its striations are roughly symmetrical and subparallel but there are occasional irregularities where some striations are truncated by others. Corners created by these striae are angular near the edge of the band but progressively rounded towards the middle (Sullivan et al., 1998a). Striations in Band C are subparallel, but there appears to be no central striation. Band D seems to have a groove centrally positioned within a flat central region, outside of which are strips of closely-spaced, subparallel or sinuous ridges. Band A also has internal subparallel striations, some of which may truncate the internal striae of Band B.

Borders of these bands are sharp, defined by albedo and truncation of previous structural trends. There is no indication of subordinate cracking in the bounding blocks, that might be associated with the formation of the bands. In stereo images, part of Band B appears to be slightly uplifted with respect to the adjacent bright plains. Blocks associated with Band A and a branch of it may have rotated, including a large block in the northeast part of the 420 m/pixel image which has rotated 20° counterclockwise (Tufts, 1998).

Figure 5
Figure 6

First, Band B was reconstructed by a northeastward translation and a 4° clockwise rotation of the block to its west. Band D provided a marked piercing point. The relative motion vector trends ENE. Net relative motion across Band B was mostly dilational with a slight dextral component, and included some block rotation.

Second, to close Band C, preparatory cuts were required where the band forks. Approximately half of the net closure of the band was made on each side of the headland dividing it. The net reconstruction of the main band was made along a NNE vector, with no rotation. Given the irregular course of the band, its net motion ranges from dilational to dextral. Band C is narrowest where its trend is most parallel to the relative motion vector, that is, nearly pure strike-slip.

Third, by reconstructing Bands B and C, Band D was restored to the configuration it had before the younger bands opened. Its shape suggested that it too could be reconstructed. Band D closed via a NNW translation. To accommodate the split at its eastern end required that this translation be divided between the two branches. The deformation is dilational with a very slight sinistral component.

This sequence of reconstructions restored Lineament L and Region O (Fig. 6). The reconstruction is unambiguous because linear piercing point structures trend at widely different angles. Lineament L is probably a Class 2 or Class 3 ridge in the classification system of Greenberg et al. (1998a). Region O is an area of discrete topographic identity dismembered by Bands B and D; it may be a headland dividing an older band.

Overall, the reconstruction was achieved with few gaps or overlaps. In closing each band, several minor lineaments also were realigned. The remaining gores are probably due to inaccuracies in the reconstruction technique or to residual amounts of unrecognized deformation. Except for only a few localized deformations, the major blocks seemed to maintain their general rigidity during the opening of these bands given the compatibility of this reconstruction (cf. Sullivan et al. (1998a)). Minor misfits of small lineaments (Sullivan et al., 1997; Sullivan et al., 1998a) may reflect the pattern of local discontinuity of small ridges seen elsewhere on Europa in very high-resolution images. The deformation represented by the opening of Bands B, C, and D caused a 15% NE-SW lengthening, a value similar to regional estimates by Schenk and McKinnon (1989).

Figure 7

The overall parallelogram shape of the feature resembles the shape of "pull-aparts," a common tensional element in terrestrial strike-slip faults (Sylvester, 1988). If that analogy is correct, the orientation of the band is consistent with the right-lateral shear of Astypalaea Linea. Assuming that Cyclades Macula and Astypalaea Linea represent equal, coordinated displacement, Tufts (1996; 1998) reconstructed the fault, closing the rhomboid by its 42 km width. This reconstruction realigned seven older lineae crosscut by Astypalaea Linea, including lineaments of two types that impinge on the fault from widely different angles. Reconstructions of Cyclades Macula based on Galileo images suggest a somewhat greater offset than 42 km, perhaps due to the presence of small lineaments which accommodate additional offset.

Cyclades Macula apparently accommodated shear on Astypalaea Linea. Because of its rhomboidal shape, relative motion at Cyclades Macula included a dilational component. Motion was probably concentrated along the narrow crack which diagonally crosses the rhomboid. The stairstepping internal striation pattern reflects the need to transfer slip between the opposite corners of the band. In reaches of this crack that are subparallel to overall fault movement, the motion was predominantly strike-slip, while in reaches subparallel to the rhomboid length, motion was more strongly dilational.

As viewed in Galileo images, Libya Linea does not appear elevated (Fig. 8). Its interior surface is composed of subparallel striations, giving it a raked appearance, like parts of Cyclades Macula. It also contains some flat-appearing regions. Libya Linea is more consistently smooth and its internal striations more parallel than Cyclades Macula, and unlike that band, it does not exhibit a principal structure that might accommodate slip or define an axis of symmetry. No relief is apparent beyond locally elevated lineaments composing the striation pattern of the band. Dark regions interrupt the linea near its junction with Astypalaea Linea and may be associated with changes in the trend or texture of striations.

Figure 8

A possible reconstruction of Libya Linea is suggested by the shape of the band and by the presence of two possible piercing points in recent low-resolution Galileo images (E14). If it is a valid reconstruction, the band would be dilational, with perhaps a small component of strike-slip motion. This reconstruction may be on a vector similar to the displacement vector for Astypalaea Linea. However, low resolution makes this interpretation uncertain. Even in higher resolution images, no other candidate piercing points are yet apparent.

Libya Linea may be kinematically linked to Astypalaea Linea (Tufts, 1998), despite these difficulties in reconstruction. This relationship is suggested by possible similarity of their two displacement vectors and by the fact that the two lineae may intersect so that strike-slip on Astypalaea Linea could be accommodated by dilation on Libya Linea. The Astypalaea Linea-Libya Linea system would mark a boundary between regional plates. Recent Galileo results may be consistent with this hypothesis. However, the intersection of the two structures is complex, as we described above. It is possible that Libya Linea represents more than one dilational period, each of which was accommodated by a different strike-slip fault.

A portion of the band that is crossed by Sidon Flexus was imaged by Galileo (E17 orbit). There it presents a smooth, finely striated surface that does not appear to be elevated. Two trough-shaped grooves might appear to be "central" grooves, however both cut across the striation pattern, so their nature is uncertain. Otherwise, there is no clear central groove or crack. The striation pattern appears regular though preliminary analysis does not reveal clear bilateral symmetry, however, lack of a clear central groove hinders this assessment. Low-resolution illumination patterns farther to the north suggest that there the band is slightly elevated above the surrounding landscape (Malin and Pieri, 1986). In the same northern region of the band, Thynia Linea contains vague striations parallel to its boundaries.

Thynia Linea can be reconstructed across its width, matching multiple piercing points, and fully closing the band. Accordingly, the band is predominantly a dilational gap which developed from a near-vertical crack. As the band curves eastward, its recorded relative motion involves a greater proportion of strike-slip. Pappalardo and Sullivan (1996) proposed that bands such as Thynia Linea may constitute an important resurfacing mechanism.

Figure 9a
Figure 9b

Misalignment of older ridges crossed by the band suggest that it might reconstruct. A reconstruction can be accomplished by connecting these features by a direct translation. The reconstruction is unambiguous because realigned lineaments are orthogonal. See Figure 9b. The band is dilational with a moderate component of right-lateral motion.

The band can be reconstructed, realigning three older bands, as shown in Figures 10a and 10b. Relative motion included dilation and left-lateral shear. The reconstruction vectors vary slightly in length and direction.

Figure 10a
Figure 10b

  Figure 11a
  Figure 11b

C. Characterization of Bands

It is uncertain where in the band this repeated cracking occurs. Sullivan et al. (1998a) and Prockter et al. (1998a) suggest that it occurs along the midline, perhaps leaving a central striation or striation pair with a distinctive albedo or topography. Cracking may be off-axis, as we discuss in the next section. While the bands may be wide, they represent the accumulated effect of repeated opening and filling events in cracks that are relatively narrow.

D. Formation of Bands by External Forces

Bands appear to form when forces on the blocks, external to the initial crack, pull them apart, consistent with Sullivan et al. (1998a). They are distinct from Class 2 ridges which, according to the model by Greenberg et al. (1998a), develop due to a form of internal forcing (Sec. II.C). While the band experiences variable daily tidal stresses, the external pulling forces dominate, counteracting effects of tidal compression. The active crack becomes filled with ice as water from below enters the widening gap and freezes. As a result, such dilation is irreversible.

The lesser compression in the formation of bands accounts for the differences between bands and Class 2 ridges. First, compression elevates Class 2 ridges but does it to a smaller extent for bands. For this reason, there is also less tendency to form a pronounced central groove. (External pulling may enable bands to achieve greater widths than Class 2 ridges.) Second, with less compressional consolidation of its walls, the central crack may tend to lose its identity. Accordingly, the locus of cracking may shift, reducing the regularity of the band's bilateral symmetry. For example, Band B (Sec. III.B.1) has an indistinct medial groove (Sullivan et al., 1998a), and its internal striation pattern is sometimes irregular. It is unclear if Thynia Linea exhibits a central groove, and the internal striations, while perhaps sequentially symmetrical, may not be geometrically symmetrical. Libya Linea and Agenor Linea both show a striated interior but do not seem to have central cracks.

E. Sources of External Forcing

While Class 2 ridges result from plates being pushed apart at a crack, the external pulling-apart that is needed to create bands may come from a variety of sources. The pulling-open of some bands may be caused by walking (Tufts et al., 1997b; Hoppa et al., 1999) of strike-slip faults to which the bands are linked. In the walking process, which is due to tidal effects, action at the fault is the driver, in contrast to terrestrial plate tectonics. Walking has been shown to be a plausible cause for the large strike-slip fault Astypalaea Linea (Tufts, 1998; Tufts et al., 1998c) , and for numerous smaller strike-slip faults on the satellite (Hoppa et al., 1999). Examples of bands linked to strike-slip faults include Cyclades Macula and possibly Libya Linea, in which the band directly accommodates strike-slip motion, as we discuss in Section II.B.2. Apparent dilational lineaments neighboring those bands may be other examples of such linkages (Figure 12).

Figure 12

Walking may also cause dilation in situations where a crack bends or curves considerably. Portions of the crack might shear due to walking while differently oriented segments dilate along the same vector. The dilation of "The Sickle," a large band near Manannan, for example, might be created by walking on its north-trending portion (see Fig. 3).

The block rotation responsible for various wedge-shaped bands may also be driven by walking. For example, block BR or the block bordered by Band B may have rotated due to walking on Band A (Sec. III.B.1). Similarly, small wedge-shaped bands associated with rotated, ovoid blocks near The Sickle (Sullivan et al., 1996b) may have formed by walking on the faults surrounding the blocks.

If there are local or regional currents in a subsurface ocean, they too may cause blocks to move, thereby pulling dilational bands open. (Such currents may also activate shear (Tufts, 1998)). Currents might be generated by tidal heating. For currents to create divergence of blocks, the blocks on one side of a crack must be constrained from moving relative to their neighbors, or current velocity must be gradational. For straight bands which do not seem to be connected to any strike-slip faults, currents may be the only plausible driver.

IV. LINEAMENTS WITH CHARACTERISTICS OF RIDGES AND BANDS

Some dilational lineaments combine morphological elements of both bands and ridges. In this section we describe various examples and discuss the continuity between the processes of formation of bands and ridges.

A. Description and Characterization

Figure 13

(b) Band B, identified in Section (III.B.1) as a band based on its morphology, is bordered by single ridges which appear as distinct dark and light albedo lines (Figs. 4a,b). These lines are indicated in the graph of reflectance values for the band prepared by Sullivan et al. (1998a).

(c) Several band-like arcuate lineaments in the northern hemisphere are bounded by ridges in the same fashion as the lineaments (a,b) above (Fig. 14). These arcuate lineaments are wider than Class 1 ridges alone. They show a vague internal striation pattern and no central groove. While it is unclear if their interiors are elevated, these lineaments can be considered bands because of their morphology. These arcuate bands may be incipient versions of large bands which are also arcuate in shape, such as Band B (Sec. III.B.1). Opposite sides of these lineaments match one another except for local irregularities. Thus, the lineaments probably reconstruct and are thus dilational.

Figure 14

Figure 15

The band-like character of this elevated lineament is underscored by the lack of internal bilateral symmetry. The lineament intersects a second, NNE-trending, narrow lineament which is also slightly elevated. Where it is met by this smaller lineament, the internal striations of the main band change orientation. Some striations become tangential to the sides of the main lineament and some of them align with the smaller, intersecting lineament. Opening may have taken place along the sides of the structure where striations are tangential.

In fact, some of the bands we have already considered are slightly elevated. As we noted in Section III.B.1, these include Band B which is slightly elevated despite the lack of both strong symmetry and a pronounced groove (Fig. 15c). Likewise, Thynia Linea is elevated in its northern portions, as evidenced from sun illumination patterns, although it seems to have no clear central groove in Galileo images (Sec. III.B.2). (Also, Agenor Linea, a band-like structure with no internal symmetry or central groove, appears in stereo images to be slightly elevated.)

Figure 16

(a) A wide, slightly curved lineament in the northern part of the trailing hemisphere (Galileo, E15) has ridges located on the midline, straddling the center, and at the borders (Fig. 16a). Except where the ridges are located, the lineament appears to be at roughly the ambient elevation of the region. The ridges that lie between the midline and the borders are arranged in bilaterally symmetric pairs. There may be a central groove within the midline ridge but image resolution hinders an interpretation. The lineament offsets older structures which can be rejoined by palinspastic reconstruction, so it is dilational. (b) A curved lineament of variable width, located in the northern part of the trailing hemisphere (Fig. 16b), exhibits prominent bordering ridges and a midline ridge system. Except where ridged, the elevation of the lineament seems equal to the ambient terrain. The central ridge is most pronounced where the lineament in narrowest, and there it has a strong central groove. Where the lineament widens, this central ridge pair becomes more subdued, although the groove is still present. (c) A lineament near Conamara Chaos (Fig. 16c) provides another example. Here, a wide strip contains a ridge pair that is flanked by a symmetric array of parallel ridges. The strip is low, except where it is ridged, and it has been cut by strike-slip faults. The edges of this strip were destroyed by formation of chaotic terrain, so it is unknown whether it had bordering ridges; reconstruction to determine the sense of motion is not possible.

B. Formational Models: A Continuum of Properties and Formation Styles

The features of Class 2 ridges, bands, and the combined forms noted in the previous sections suggest that Class 2 ridges and bands are end members in a process continuum. See Figure 17. Lineaments with characteristics of both bands and Class 2 ridges represent intermediates created by such a continuum. Morphologic elements that may vary in response to this process continuum are: (1) elevation of the lineament surface; (2) width, although width is also a function of longevity of growth; (3) bilateral symmetry; and (4) prominence of a central groove. The key to the variations along this continuum is the degree to which tidal compression is canceled out by external pulling forces. So, mixtures of the internal forcing which produces dilational ridges and the external forcing which produces bands can create these intermediates by causing variations in these elements. In this way, the relative proportion and timing of these two driving processes substantially determines the lineament morphology. So, it may be possible to infer the stress history of a dilational lineament from its morphology.

Figure 17

Bands and ridges originally derive from cracks. Cracks may develop into doublet Class 1 ridges, dilational Class 2, or complex Class 3 ridges according to Greenberg et al. (1998a). Cracks and any of the classes of ridges may widen into bands. For example, the E11 band (Sec. III.B.3) seems to have no bordering ridges; it grew from a crack that had widened without ridge-building. Furthermore, external forces on the plates, rather than outward pushing by material lodged in the active crack, was dominant. Had widening occurred after a period of ridge development, the band would be bordered by two, oppositely-facing half-ridges. The cases described in Section IV.A.1 are examples.

The dilational process is controlled by the balance between internal and external forcing. While they may not elevate as much as Class 2 ridges do, bands can also become somewhat elevated if ice becomes partially lodged within the active crack. The elevation of the dark band in the wedges region (Sec. IV.A.2) may indicate that this effect played a role along with external forces pulling plates apart. Even though we initially labeled them as bands, the slight raised elevation of Band B, the northern part of Thynia Linea, and Agenor Linea suggests these bands or parts of bands formed in a stress regime in which compression from within the crack contributed somewhat to separating the plates. Wide, low symmetrical lineaments with central grooves (Sec. IV.A.3) may be dilational lineaments which were predominately formed by internal forcing, like end-member Class 2 ridges, but for which external pulling occasionally increased somewhat. Internal forcing would lead to the strong bilateral symmetry and a central groove; episodes of increased external forcing would create lower-elevation strips within the symmetrical lineament. In this context, it is notable that the central groove in the lineament shown in Figure 15b (Sec. IV.A.3.b) is most prominent where the lineament is narrow, that is, where external forcing may have been least. In the case of Band D (Sec. III.A.1), its platform-like, grooved central portion may record a late-stage period of internal forcing. In this context, the critical identifier of the "pure" end-member band would be a total lack of topographic elevation, combined with irregularly symmetrical internal lineaments, and a muted central groove. This configuration would come about when external pulling forces consistently dominate internal pushing forces. Among the bands examined here, perhaps Thynia Linea or Libya Linea are the best examples of an end-member band.

External pulling on plates may cease from time to time while building a Class 1 ridge at the active crack. See Figure 17. If external forcing resumes or increases, the halves of the new ridge pair might begin to spread apart. Densely ridged bands may reflect intermittent cessation and resumption of external forcing during the band's active life. Examples may include the outer portions of Band D (Sec. III.B.1), the band on the leading side (Sec. III.A, #2), or the band near Tyre (Sec. III.B.3.b). Patches of densely-ridged terrain may be remnants of such bands. A succession of half Class 1 ridges that have been spread apart in this way might be mistaken for normal fault blocks, especially if the outer slopes of double ridges (Class 1 in the Greenberg taxonomy) are consistently steeper than the inner slopes, as suggested by Kadel et al. (1998).

The details of the creation of the striations within dilating lineaments are uncertain. See also Prockter et al. (1998a). Individual striations are too few and too wide to represent daily opening and closing due to diurnal tides. A longer term periodic process which may bring about discernible striations is variable dilation due to the periodic changes in resolved diurnal tidal stress within a single rotation relative to the direction of Jupiter (Geissler et al., 1998a; Greenberg et al., 1998a; Hoppa, 1998). Also, variations in the efficiency of driving processes may create striations. Furthermore, the location of the active crack may change with recurrent opening and closing, as suggested by the rounding of internal corners in Band B (Sullivan et al., 1998a). Finally, some striations formed at other than the central crack may be younger ridges which exploited flaws developed in the band interior by dilational processes.

V. DISCUSSION AND CONCLUSIONS

Ridges and bands constitute a major tectonic agent of Europan resurfacing, consistent with Pappalardo and Sullivan (1996). Ridges are ubiquitous on Europa. Ridges and bands are evident on a global scale and have occurred over a long time span. Ridges resurface Europa by burying the landscape under the ice slush that extrudes from their central cracks, according to Greenberg et al. (1998a). Class 2 ridges and bands resurface by creation of new area between plates of the previous surface, and by separation and dismemberment which disorganizes the earlier landscape. Strike-slip displacement associated with ridges and bands further disorganizes the landscape, and may drive additional dilation.

The other major resurfacing process is formation of chaos regions, possibly by melting of the lithosphere (Greenberg et al., 1998b). But, in that way, chaos regions may also play a role in dilation. The availability of lateral room would influence the stress environment in which a dilational lineament forms, and may determine how wide a band or Class 2 ridge can become or whether it develops at all. Lineaments may be able to dilate if a multiple, dispersed chaos regions and lenticulae provide space for expansion by creating "holes" in the lithosphere. Irregular spacing of nearby chaos regions and lenticulae may explain variations in the width of some dilational lineaments. The refreezing of accommodating chaos regions may cause a band to stop spreading; new melting-through may allow spreading to resume. Such a sequence might further explain those lineaments with properties between end-member Class 2 ridges and bands. In any case, there must be a means to consume surface area to allow for the area created by dilation (Schenk and McKinnon, 1989; Pappalardo and Sullivan, 1996; Tufts, 1998; Tufts et al., 1998c). If the process involves melting-through (with attendant loss of lateral confinement) the development of chaos regions may provide this consumption mechanism while helping to resurface the satellite (Greenberg et al., 1998b; Tufts, 1998); see also Sullivan et al. (1998a).

Dilational lineaments form a continuum of types. At one extreme are Class 2 ridges. They are elevated, usually only a few kilometers wide, have a striated fabric that is strongly parallel and bilaterally symmetrical, and they exhibit a pronounced central groove. Class 2 ridges form when ice or slush accumulates in a tidally worked crack, forcing the opposing blocks apart as daily compressive tidal stress is applied. Compression elevates the fill material as the lineament widens.

At the opposite extreme are bands, which are not elevated. They are up to a few tens of km wide with their width varying along their length. They may exhibit a subparallel lineated fabric which is only approximately bilaterally symmetrical, and they do not have a well-developed central groove. Like Class 2 ridges, bands are dilational gaps in the lithosphere, filled with material from below. They may also accommodate shear, and vertical axis rotation of the adjacent lithospheric blocks. The opening of bands occurred episodically, by repeated cracking and spreading.

Bands form when a crack is pulled apart by external forces applied to the adjacent lithospheric blocks. Drivers may include: (a) walking on strike-slip faults connected to the band; or (b) currents in a subsurface ocean. Between the extremes of Class 2 ridges and bands are the lineaments that exhibit characteristics of both. These lineaments include: (a) bands that are bordered by single ridges; (b) bands that are elevated; and (c) wide, low lineaments with bilateral symmetry and a central groove.

Such mixed lineaments arise when dilation occurs after cracks have undergone ridge development, or if a mixture of the internal forcing which produces Class 2 ridges and the external forcing which produces bands creates lineaments with the characteristics of both. Because external pulling counteracts tidal compression, the relative importance of the two end-member processes determines the morphology of Europan dilational lineaments, by controlling the degree of elevation, and the prominence of a central groove and of bilateral symmetry. This proportion may change within the lifetime of a dilational lineament and may be affected by the evolution of chaos regions.

Acknowledgments:

Images used in this paper were provided by the NASA Galileo Project and its Solid State Imaging Team. Funding was provided through the NASA Galileo Project.