Terrestrial Sea Ice Morphology: Considerations for …Two forms of ice are encountered on the seas, true sea ice, formed by the freezing of salt water at the ocean’s surface, and

ICARUS 135, 25–40 (1998)ARTICLE NO. IS985977

Terrestrial Sea Ice Morphology: Considerations for Europa

Ronald Greeley and Robert Sullivan1

Department of Geology, Arizona State University, Box 871404, Tempe, Arizona 85287-1404E-mail: [email protected]

Max D. Coon

Northwest Research Associates, Inc., 14508 NE 20th Street, Bellevue, Washington 98007-3713

Paul E. Geissler and B. Randall Tufts

Lunar and Planetary Laboratory, Space Sciences Building, University of Arizona, Tucson, Arizona 85271

James W. Head III and Robert T. Pappalardo

Department of Geological Sciences, Brown University, Box 1846, Providence, Rhode Island 02912

and

Jeffrey M. Moore

NASA–Ames Research Center, Mail Stop 245-3, Moffett Field, California 94035-1000

Received November 21, 1997; revised April 2, 1998

strength of ice sheets. Salts are also suspected in the europanice and could lead to similar differences, enhancing the creationThe Galileo mission has returned the first high-resolutionof topographic relief from density contrasts and the formation(21 m/pixel) images of the surface of Europa. These imagesof fractures from brittle failure of the ice. Differences in thereveal structures with morphologies reminiscent of those seenenvironments between Europa and terrestrial sea ice in termson terrestrial sea ice. Although it is premature to make one-of parameters such as temperature, gravity, time, and ice com-to-one analogies between sea ice and Europa’s surface, a reviewpositions suggest caution in drawing direct analogies. Futureof the types of surface features commonly formed on Earthwork by the planetary and sea-ice communities must includeand of various sea-ice processes can provide insight into theunderstanding the terrestrial processes sufficiently for extrapo-complex geology of Europa. For example, deformation of terres-lation to Europa. 1998 Academic Presstrial sea ice results from winds, tides, and currents and from

Key Words: Europa; geological processes (Europa); surfacesthermally induced stresses; the resulting features include frac-(satellites); planetary ices.tures ranging in width from millimeters to kilometers, pressure

ridges, shear ridges, and rafted ice. Potential agents of deforma-tion on Europa are more likely to be limited to tidal flexing

1. INTRODUCTIONand possibly convection, but could produce similar featuresand perhaps account for the ridges and fractures seen in many

Most of the outer planet satellites are composed of mix-areas. Subtle differences in albedo and color in terrestrial seatures of ices and other materials, including silicate minerals.ice result from differences in ice thickness and grain size, attrib-Many of these satellites show surfaces that have beenuted to factors such as the rate of ice-crystal growth, water

turbulence, age of the ice, and deformation. Similar factors shaped by impact cratering and deformed by tectonic pro-could account for differences observed in the bright icy plains cesses. Europa is particularly interesting because it appearsof Europa. Moreover, salts in both the solid form and as brine to consist of an inner core surrounded by a rocky zonevary in concentration and composition as a function of space and an outer shell of water composition (Anderson et al.and time on Earth, leading to differences in density and the 1997). Ground-based observations of Europa established

water ice as a significant component of the surface priorto the arrival of the Voyager spacecraft (Kuiper 1957, Pil-1 Currently at Cornell University, 308 Space Sciences Building, Ithaca,

NY 14853. cher et al. 1972, Fink et al. 1973). Some of the water could

250019-1035/98 $25.00

Copyright 1998 by Academic PressAll rights of reproduction in any form reserved.

26 GREELEY ET AL.

FIG. 1. Galileo image of Europa showing part of the antijovian region where the ice surface has been fractured and separated. The prominentdark band appears to have been infilled by darker material. In turn, it has been cut by two sets of ridges. Area shown is 116 by 123 km; resolutionis about 420 m/pixel. Illumination is from the left; north is to the top (Galileo frame, ASU IPF-1034).

be liquid beneath an ice crust under current conditions or 1995 and began returning image data from Europa on thefirst of 11 flybys (Belton et al. 1996). Prime mission imagesin the past, based on models of interior heating (Cassen

et al. 1982, Squyres et al. 1983, Schubert et al. 1986). range from global color observations at .6 km/pixel tohigh-resolution samples at 21 m/pixel. These images clearlyVoyager images of Europa obtained during two flybys

in 1979 revealed the ice surface to exhibit complex patterns show the tectonic complexity that characterizes the icysurface of Europa. For example, the disruption of platesof lineaments generally inferred to be fractures (Smith et al.

1979a, 1979b). Although the highest resolution Voyager associated with wedge-shaped dark bands seen on Voyagerpictures was confirmed and extended to other areas withimages were only about 2 km/pixel, general geological

terrains were identified along with a wide variety of surface 1.6 km/pixel and 420 m/pixel images (Fig. 1) (Sullivan et al.1997, 1998, Tufts et al. 1997). Higher resolution data (,35features and a few impact craters (Lucchitta and Soder-

blom 1982, Malin and Pieri 1986). Of particular interest m/pixel) reveal the surface to have been broken into icyplates on very small scales (Fig. 2), with seams betweenwere some fractures that had widened into wedge-shaped

dark bands representing as much as 25 km of relative plates commonly represented as ridge and groove terrainthat in some places show lateral offsets. Except for isolatedmovement between icy plates (Schenk and Seyfert 1980,

Pieri 1981, Schenk and McKinnon 1989, Golombek and massifs, ridges and other surface features generally exhibit,300 m relief (Lucchitta and Soderblom 1982, Malin andBanerdt 1990, Pappalardo and Sullivan 1996).

The Galileo spacecraft arrived at Jupiter in December Pieri 1986, Greeley et al. 1998). The low relief of ridges

TERRESTRIAL SEA ICE MORPHOLOGY 27

FIG. 2. Conamara Chaos is the disrupted terrain on Europa bounded to the northwest (upper left) and northeast (upper right) by complexridge systems. This chaos appears to have formed by the breakup of bright terrain into segments that have rotated and translated into new positions.Arrow indicates location of Fig. 3, which shows part of the area in more detail. Resolution is 180 m/pixel; illumination is from the right. Areashown is about 215 km wide; north is to the top (Galileo mosaic; ASU IPF 1060-2).

and other features, as well as areas where plates have been Earth (Schenk and Seyfort 1980, Tufts 1996). For example,the pre-Galileo work of Pappalardo and Coon (1996) com-destroyed and replaced with lower standing icy rubble (Fig.

3), suggest that mechanically brittle surface materials are pared ice-ridge formation in polar seas on Earth with theformation of ridges on Europa. Because of the potentialisostatically compensated above ductile or liquid materials

at relatively shallow depths (Carr et al. 1997, Greenberg comparisons that might be drawn between Europa andEarth, various conferences and organizations have broughtet al. 1998).

The morphologies of europan features inspired compari- together the planetary and terrestrial sea-ice communities.Meetings include the Europa Ocean Conference held insons with structures and processes observed in sea ice on

FIG. 3. High-resolution (54 m/pixel) view of part of Conamara Chaos, showing individual blocks set in a topographically lower matrix of chaoticterrain. The flat surfaces of some of the blocks appear to be titled (JPL P-48526).

28 GREELEY ET AL.

FIG. 5. Intersecting cracks ,2 cm wide formed in sea ice from bend-ing and flexing of the ice sheet. Glove indicates scale (ASU photograph

FIG. 4. Oblique aerial photograph of grounded sea-ice rubble sur- no. 4371-H, by Max Coon).rounded by sea ice in the Bering Sea. The grounded feature is about100 m wide; general direction of sea ice movement is to the upper left,indicated by the lead (dark area) in the ‘‘wake’’ of the grounded ice(ASU photograph no. 4384-H, by Max Coon).

FIG. 6. Landsat image showing ice-sheet breakup, producing ice floes and areas of open water, or polynyas (dark areas). Area shown is 166by 148 km. Landsat image 1576-16095; ASU photograph No. 2039-A).


FIG. 7. Aerial photograph of a lead in the Beaufort Sea, in which the ice has opened partly along preexisting cracks. Older pressure ridgesare also visible; deposits of snow and ice indicate a prevailing wind direction from the upper right. Area shown is about 2 by 2.8 km; illuminationis from the lower left (ASU photograph no. 4374-H).

San Juan Capistrano, California (November 12, 1996), the Before a terrestrial analogy can be made, the fundamentalprocesses involved in the development of a feature must bePlanetary Ices Conference in Flagstaff, Arizona (June 11–

13, 1997), and planning sessions for a proposed project to understood well enough to allow extrapolation to anotherplanetary environment. Unfortunately, many terrestrialdrill through the ice crust over Lake Vostok in Antarctica,

sponsored jointly by NASA and the NSF. sea ice processes are not yet fully understood. Moreover,comparisons are complicated not only by differences inIn this paper, we illustrate surface features that result

from various terrestrial sea-ice processes. Our goal is to scale, but by fundamental physical differences betweenenvironments on Earth and Europa, especially regardingacquaint the planetary community with the morphology

and formational processes of some of the more common surface temperature, vapor pressure, gravity, potential icethickness, and perhaps ice compositions. For this reason,sea-ice features that appear to be appropriate for consider-

ations of the europan surface. For the most part, we are the differences between europan landforms and their ter-restrial counterparts are likely to be as instructive as theirnot drawing direct analogies with Europa and the features

imaged by Galileo. The history of planetary science is similarities. In the Discussion section (Section 3), we out-line some of the key differences and similarities betweenfraught with the errors of drawing simplistic analogies.

30 GREELEY ET AL.

FIG. 8. Oblique-aerial photograph of a refrozen lead in the BeaufortSea, showing evidence for multiple opening events, in which the darkerareas are the most recent. Shear deformation is indicated by changes inwidth of the most recent opening event. Area shown is about 1 km wide;illumination from the left (ASU photograph no. 4367-H).

FIG. 10. Aerial photograph in the Beaufort Sea showing an areaabout 3 km across which has undergone breakup and divergence. Thedarkest areas are open water. Pressure ridges (the bright lines) andsmooth, undeformed ice are visible. Illumination is from the left (ASUphotograph no. 4366-H).

europan and terrestrial sea ice and some of the areas forfuture research.

2. TERRESTRIAL SEA ICE

Nearly two percent of the total amount of water onEarth is in its solid state, ice. Ice appears in a variety offorms, each with characteristic modes of formation, evolu-tionary cycles, and descriptive terminology (Steffen 1986).Two forms of ice are encountered on the seas, true sea ice,formed by the freezing of salt water at the ocean’s surface,and glacier, or land ice, formed by snow metamorphismon land, or on extensions of freshwater land ice into thesea (ice shelves). Glaciers and ice formed on fresh waterFIG. 9. Landsat image of new leads (dark, thin stripes) in the Beau-are not considered here, other than the breakup of floatingfort Sea that have crosscut older, refrozen leads (grayer bands). Area

shown is 145 by 185 km (ASU photograph no. 2042-A). glacier ice to form ice islands (icebergs) in the sea.


FIG. 11. This figure shows an aerial view of networks of pressure ridges in the Beaufort Sea, and a reconstruction of the area. (a) The brightlines are newly formed pressure ridges, and the dark area in the upper left-hand corner is open water. Area shown is about 5 km wide; illuminationis from the left (ASU photograph no. 2037-H). (b) Sketch map of the photograph, identifying key ridges. (c) Reconstruction of the ice sheet of (a)so that ridges 1, 2, and 3 are realigned. This suggests that ridge 4 formed by the closure of the lead shown in (c).

2.1. Overview of Sea Ice ally drains downward out of the ice through brine drain-age channels 2 to 10 mm in diameter. Near the bottom

Sea ice begins to freeze into continuous sheets in the of the ice sheet, brine drainage channels are spaced 15to 20 cm apart and are about 1 cm in diameter. Theautumn or early winter when temperatures lower to 21.8

to 08C. The size and orientation of the first crystals that type and amount of brine in sea ice is governed by thewater salinity and ambient temperature. As either theform depend on the sea state at the time of formation

(Michel 1978). Turbulence favors the formation of small (a salinity or temperature increases, the brine volume in-creases to maintain phase equilibrium with the ice. Infew millimeters) equigranular crystals randomly oriented,

while large (10–20 mm) plate-like crystals grow in calm addition to sodium chloride, there are other solid saltsthat form in sea ice, such as KCl and MgCl2 · 8H2Owater.

During freezing, part of the salt in the sea water is ex- (Assur 1958). Studies show that the strength, elasticmodulus, and other properties of sea ice are stronglypelled into the ocean and the remaining salt is either re-

tained as solids or trapped in brine pockets as liquid inclu- influenced by the ice brine volume and the number andspacing of brine pockets and drainage channels (Weekssions. The spacing among brine pockets is a few

millimeters, but varies inversely with the speed of ice and Assur 1969, Schwarz and Weeks 1977).Once an initial ice skin covers the sea, the crystals growgrowth. Very rapid ice growth incorporates most of the

salt into the ice and very slow growth results in near- downward from the interface of the ice and water andincrease in size. Mattes of fine spicules and plates of seatotal salt rejection. Brine has a higher density and gradu-

32 GREELEY ET AL.

FIG. 12. Oblique aerial photograph of a newly formed pressure ridge showing blocks of ice forming the ‘‘sail.’’ About 300 m of the ridge lengthis shown. In general, ridge sails (and keels) vary in height and width. Illumination is from the left (ASU photograph no. 4386-H, by Max Coon).

ice, called frazil, first develop and can evolve into thin, thicker and less dense than first-year ice and, thus, floatshigher in the water. Such multiyear ice has a smoothercontinuous elastic crusts, or nilas. A combination of broken

nilas and aggregates of frazil ice form circular pieces 0.3 surface than first-year ice due to weathering and oftenappears greenish-blue.to 3 m in diameter called pancake ice. Pancakes are charac-

terized by raised rims that result from collisions with neigh- In shallow seas, features such as ice ridges or rubblefields can become fixed to the sea floor or grounded. Whenbors. Pancake ice can also form at depth, where water

zones of different temperature and density interface. This ice grounds, it builds a foundation that can pile ice toheights of 30 m or more. Major pile-ups of ice often occursubmarine pancake ice can later rise to the surface and

cover large areas. on beaches. At times, ice can completely override lowislands and the surrounding sea ice can move past theSea ice formed in one winter’s growth is termed first-

year or white ice. Ice surviving one summer’s melt becomes grounded features, leaving swaths in the ice (Fig. 4). Sea

FIG. 13. Calculated profile for 20 cm thick sheet of ice deformed into a pressure ridge, showing the sail and the keel, as derived from a computermodel (from Parmerter and Coon 1972).


FIG. 14. Ground photograph of a newly formed pressure ridge with a block of thick ice pushed up from below. This suggests that there is alarge keel associated with the ridge. The ridge ‘‘sail’’ behind the block is about 3 m high (ASU photograph no. 4386-H, by Max Coon).

ice that forms along coastlines usually remains anchored extend tens of kilometers or more (Bazant 1992), but donot necessarily evolve into leads; they can simply separateto the land during the winter and early spring and is re-

ferred to as fast ice. a few centimeters and then refreeze.Shifting winds and ocean currents result in repeatedFractured sea ice (Fig. 5) forms flat segments called floes

(Fig. 6). Fractures range in width from millimeters to .500 opening and closing of sea ice (Fig. 8). The orientationsof leads can change within a few hours; one set of leadsm; those wider than a few tens of meters are called leads

(Fig. 7). Under loading, ice can deform into rubble fields, can become inactive and replaced by another set with adifferent orientation (Fig. 9). Sea ice can diverge and ex-pressure ridges, and rafted ice, the latter composed of newly

formed ice from refrozen leads jammed over older ice. pose the ocean through a network of openings (Fig. 10).Where leads are continually forming, they result in a wideBecause adjacent floes can move in different directions

and speeds, irregular lake-like spaces of open water called variety of ice thicknesses, with the thickest undeformedice being the oldest. Shear can occur along leads, resultingpolynyas can also develop. In winter, the open water in

cracks, leads, and polynyas freezes almost as soon as the in additional deformation.water is exposed.

2.3. Sea Ice Ridges

2.2. Sea Ice Fractures The most common type of surface features on sea iceare linear arrays of broken ice called ice ridges (WeeksFracturing of sea ice can result from loads imposed byand Kovacs 1970). From 13 to 39% of Arctic sea ice iscurrents, winds, or changes in temperature (Coon 1974,occupied by ice ridges. Ice ridges include two forms, (1)1980, Coon et al. 1998). One of the simplest fracture mecha-pressure ridges, which are created by the relative motionnisms is thermal cracking (Evans and Untersteiner 1971),of two ice floes toward one another and (2) shear ridges,caused as the top layer of an ice sheet cools, contracts, andwhich develop at the boundary between stationary, or fast,attempts to curl up from the surface of the sea. However,ice and ice moving as floes. They are less common thanthe sheet is folded back to the water by gravity, causingpressure ridges.bending stresses that fracture the ice. It is common in the

Arctic to hear thermal cracking during most of the freezing Pressure ridges. Pressure ridges (Figs. 11 and 12) formby the collision of interacting ice sheets. Forces requiredseason. The cracks form very rapidly; propagation is inde-

pendent of the ice thickness, in that they cut through thick to push ice sheets into ridges range in magnitude from 105

to 107 dynes per cm along a lead. In terrestrial polar re-and thin areas without noticeable deflections. Cracks can

34 GREELEY ET AL.

FIG. 15. Aerial photograph showing shear deformation indicated by offset ridges. Bright streaks are deposits of windblown snow and ice. Areashown is about 1 by 1 km; illumination is from the top (ASU photograph no. 4362-H).

gions, winds blowing continuously over a 20-km fetch can Although the ridging process is complex, the mechanicalinteractions between ice sheets to form pressure ridges areproduce such a force. The high-standing part of a ridge is

called the sail, while the part extending into the water well understood. The process begins as two sheets of icemove toward each other and close a gap, such as a lead,below the bottom is called the keel. Pressure ridges are

composed primarily of ice less than a meter thick derived between them. The gap can be filled with ice rubble formedfrom thin ice in a refrozen lead subsequently ruptured byfrom lead ice. When the ice pack is compressed, the weak-

est (i.e., thinnest) material is deformed first. However, relative motion, or from blocks broken from the parentbody. As the sheet advances, the rubble is thrust eitherbecause leads rarely close exactly as they open, areas of

thin lead ice commonly remain even after rather severe over or under the sheet. A model for this process wasdeveloped by Parmerter and Coon (1972), shown in Fig.deformation. Ridges of larger dimensions tend to be com-

posed of thicker ice. 13, in which the deforming ice sheet that produces the


ridge is indicated by a dashed line. This sheet breaks froma downward load imposed by the weight of the sail andan upward load from the bouyancy of the keel. These twoforces cause a moment, and the ice fails in the bending.Material is swept toward the center of the former lead anda ridge-like structure is created. Gravity forces define anangle of repose that limits the steepness of slopes of thisstructure. Floating rubble is free to seek isostatic equilib-rium, but rubble above and below the ice sheet can adjustonly indirectly by bending the ice sheet toward equilibrium(Fig. 14).

As the ridging process continues, the central rubble pilereaches a critical height and then grows laterally. Addi-tional ice breaks from the sheet, adds to the edge of therubble pile, and accumulates to the limiting height of thecentral core (Fig. 13). The limiting height of the ridgeis assumed to be that height producing bending stressessufficient to fracture the ice sheet and the thickness ofthe supporting keel (Parmerter and Coon 1972). Voidsbetween large ice blocks, common in pressure ridges, canproduce average densities less than that of solid ice. Theheight of ridges averages about 2 to 4 m, although sailheights range up to 12.8 m for free-floating ridges (Kovacset al. 1972) and more than 30 m for grounded pressureridges (Taylor 1978).

Shear ridges. Deformation also occurs when ice floesFIG. 16. Rafting of thin ice is seen as the gray areas in this lead(dark zone). Finger rafting is shown by the zig-zag pattern in the top slide past one another (Fig. 15), forming shear ridges. Shearmiddle of the photograph. Area shown is about 1 km across (NASA- ridges are common at the boundary between stationary orJSC photograph 5-98; ASU no. 2032-H). fast ice and moving ice sheets. Simple, single frictional

scarps become more complex when both normal and tan-

FIG. 17. (a) Oblique aerial photograph (illumination from the upper left) and (b) diagram showing finger rafting.

36 GREELEY ET AL.

FIG. 18. Oblique aerial photograph showing an ice island several hundred meters across (ASU photograph no. 4395-H).

gential forces occur at the shear zone. The ice then takes dozens of kilometers across, but too small to be consideredopen seas. Polynyas form by one of two mechanisms or athe form of chaotic, randomly dispersed rubble or a zonecombination of the two (den Hartog et al. 1983). (1) Iceof quasi-parallel shear ridges. Shearing and grinding pro-can alternately form and be removed by winds or currentscesses produce the most compact and finely ground form(Pease 1987); in this mechanism, the heat lost to the atmo-of sea ice.sphere is maintained by the latent heat of fusion of the ice

2.4. Ice Rafts as it forms. These are referred to as latent heat polynyas.(2) Convecting sea water or currents can introduce suffi-

A phenomenon similar to ridge formation is ice rafting. cient heat to prevent ice formation locally. These are re-Converging ice sheets do not always fragment upon con- ferred to as sensible heat polynyas. Polynyas also can formtact, but remain coherent with one being thrust onto the or be maintained by a combination of the two mechanismsother, forming rafted ice (Fig. 16). Rafted ice is most com- involving turbulent mixing in the ocean below the polynya.mon in new, thin ice. Blocks are sometimes broken from Ice fog or sea smoke composed of water droplets can formthe leading edge of the thrust sheet and commonly develop in large quantities over polynyas and other bodies ofa pressure ridge along the edge of the rafted ice. Rafted open water.ice thicknesses 2–4 times greater than the ice sheet are

2.6. Ice Islandscommonly found within a few hundred meters of pressureridges and are the result of repeated overthrusts. ‘‘Icebergs’’ in the Arctic Ocean are technically called

In some cases, interlocking thrusts are created between ice islands and are composed of fragments split or ‘‘calved’’two ice sheets, resulting in striking rectilinear patterns. from ice shelves. The thicknesses of ice islands in the ArcticEach flow thrusts ice segments alternately over and under are variable but can exceed 70 m; thickness decreases withits neighbor and produces finger rafting (Fig. 17). time as ice islands melt and ablate. The lateral dimensions

of ice island are also variable, ranging from a few tens of2.5. Polynyas

meters to more than 10 km across. Warren et al. (1994)noted that calving is a poorly understood process and leadsPolynyas are rectangular, elliptical, or irregular openings

in sea ice ranging in size from a few hundred meters to to a wide variety of sizes and shapes of ice islands (Fig. 18).


ment of the differences and similarities between sea iceand Europa enables a better understanding of the originof the features revealed by the Galileo spacecraft.

Earth-based spectral observations of icy satellites, re-viewed recently by Brown and Cruikshank (1997), andnear-infrared mapping spectrometer data from the Galileomission show that Europa has a variety of nonice materialson the surface. Most models suggest that various salts, suchas MgSO4 , and other compounds (Hogenboom et al. 1997)could be present. Similarly, terrestrial sea ice contains notonly NaCl, but a variety of other salts (Assur 1958) in thesolid form and in brine pockets. These not only causedensity differences within the sea ice (thus affecting its‘‘float height’’), but also influence the strength of the icesheet and its susceptibility to deformation. Similarly, thesalts and other nonice components in europan ice couldlead to density differences and weaken the ice, enhancingthe formation of fractures and other structures.

Sea ice on Earth desalinates with age because the denserbrines segregate and drain downward. Salts are also ex-cluded from ice in the freezing process. The contaminationof terrestrial sea ice with salts could be analogous to someof the red–brown nonice components on Europa. The eu-ropan material forms mottled terrain which is often foundin low-lying areas where the crust has been disrupted

FIG. 19. Aerial photograph of sea ice in the summer showing areas (Greeley et al. 1998); it appears to be excluded in the ridge-of melt water (dark areas) formed on the surface of the ice. White dots building process because geologically young ridges lackin middle of photograph right of center are field camp huts about 4 by the coloration (Geissler et al. 1997). Some ridges could be4 meters (Photograph by Max Coon; ASU photograph no. 4365-H).

composed of relatively pure H2O ice rising through higherdensity ices or liquids containing salts.

Galileo data show distinctive differences in the color2.7. Weathering of Sea Ice and albedo of bright plains of Europa, which are attributed

to differences in ice-grain sizes and perhaps the presenceFreshwater melt ponds form on the surface of sea iceof nonice components (Belton et al. 1996, Helfenstein et al.by the melting of snow. First-year sea ice subjected to1998). Similar differences in sea ice are observed as asummer melting contributes additional water to the meltfunction of age, in which first-year ice tends to be fine-ponds (Fig. 19). Many ponds form by the melting of ridges,grained and white; in subsequent years the grain size iswhich receive more radiation on their sun-facing sides thancoarser and the ice becomes bluish. However, ice-graindoes the flat ice. Melt ponds grow because they trap moresizes are also a function of the state of the water (turbulentradiation than the surrounding ice and, in some cases, meltversus quiet) from which it freezes, as well as subsequentthrough the floes. The thin ice in leads around floes meltscomminution resulting from fracturing and deformation.in the spring and the floes are surrounded by open water.Moreover, terrestrial sea ice can be colored by the presenceIce that has survived more than one summer ultimatelyof algae.becomes a layer cake of annual layers that are formed

Fracturing and deformation of sea ice typically decreasesduring successive winter periods of ice growth. Ultimately,its bulk density. For example, ice rafting and pressure ridgemultiyear ice reaches an equilibrium of about 3 to 4 mformation involve stacking of sheets of ice, fracturing ofthickness, such that the thickness of ice ablated during thesheets into blocks, and development of sails and keels, allsummer equals the thickness grown during the winter.of which can have substantial void space and enable higherstanding features. Images of Europa clearly show surfaces3. DISCUSSION AND CONCLUSIONSthat have been severely deformed (Fig. 20), enhancing theprobability of lower bulk densities and the creation ofThe morphology of terrestrial sea ice and the processes

involved in the formation and evolution of sea ice provide topographic relief independent of compositional differ-ences.insight into the geology of Europa. Although we are not

making one-to-one comparisons of specific features, assess- Deformation in terrestrial sea ice occurs primarily from

38 GREELEY ET AL.

FIG. 20. High-resolution (21 m/pixel) Galileo image showing complex ridges, grooves, and lateral shear in the area north of the ConamaraChaos region of Europa and centered at 158 N, 2738 W. Area shown is 10 by 12 km; illumination is from the right (Galileo frame S 0383718600).

drag exerted by winds, tides, and ocean currents, or from Medial valleys separating pairs of bilaterally symmetricridges, so common on Europa (Fig. 20), are notably absentthermal expansion and contraction. On Europa, the most

likely agents of deformation are tidal stresses induced by in terrestrial sea ice. Terrestrial pressure ridges can reachheights that are many times greater than the thicknessits eccentric orbit around Jupiter (Greenberg et al. 1997,

Geissler et al. 1997), solid-state convection (Pappalardo of the ice sheet because they are supported by buoyantsubsurface keels much larger than the visible relief at theet al. 1997), or possible currents if subsurface water exists

(or existed in the past) beneath an ice crust. It should be surface (Fig. 13). The offset between the upward buoyancyof the keels and the downward weight of the ridge cannoted, however, that terrestrial sea-ice sheets are seldom

thicker than a few meters, far thinner than the hundreds create cracks parallel to the ridges, perhaps analogous tothose seen at much larger sizes on Europa; but, this doesof meters to many kilometer thicknesses estimated in some

regions of Europa (Carr et al. 1997). Moreover, fractures not account for the medial fracture.Fractures in both sea ice and on Europa appear to takein terrestrial sea ice are comparatively short and tend to

meander (Fig. 7), perhaps reflecting the local stresses that advantage of preexisting cracks (Pappalardo and Sullivan1996, Lucchitta and Soderblom 1982), and features formedformed them. In contrast, Europa’s straight lineaments

that are 1000 km or more in length could indicate global- by shear can be reconstructed (Fig. 11; Schenk and McKin-non 1989, Tufts 1996). Wedge-shaped leads and arcuatescale processes, such as deformation from tidal stresses.

The ridging model of Parmerter and Coon (1972) for ter- floes in sea ice form in regions of divergence and in plan-form superficially resemble the much larger ‘‘pull-apart’’restrial sea ice was modified for the Europa environment

by Pappalardo and Coon (1996) and Greenberg et al. features in the equatorial regions of Europa’s antijovianhemisphere (Fig. 1) thought to be caused by extension(1998), taking tidal stresses into acount, and is viable hy-

pothesis for the development of ridges of Europa. (Sullivan et al. 1997, Tufts et al. 1997, Geissler et al. 1997).


information on the morphology of floating ice in the early 1980s stimulatedEuropa could be tectonically active today, driven byby the Voyager results. Their compilation aided in the preparation ofanalogous tidal heating processes as those responsible forthis manuscript. Dan Ball and Brian Creed from Arizona State University

active volcanism on Io (Cassen et al. 1982). Plans for the are thanked for photographic support and Byrnece Erwin and Andreaextended phase of the Galileo mission include searching Gorman (also from Arizona State University) are thanked for word

processing. We also thank the Galileo project team for bringing thefor potential activity or changes on the surface since Voy-mission to success and for the return of the high quality images used inager or since the earlier mapping phases of Galileo. Giventhis report. This study was supported by the National Aeronautics andthe poor resolution of the Voyager data (2 km/pixel) andSpace Administration through contracts by the Jet Propulsion Labo-

the limited coverage by Galileo, the probability of observ- ratory.ing activity is low. However, there are some aspects of theeuropan surface seen in present Galileo images that could REFERENCESindicate present or recent activity.

Anderson, J. D., E. L. Lau, W. L. Sjorgren, G. Schubert, and W. B.The review of terrestrial ice and the discussion outlinedMoore 1997. Europa’s differentiated internal structure: Inferences fromabove suggest several tentative conclusions and avenuestwo Galileo encounters. Science 276, 1236–1239for further research.

Assur, A. 1958. Composition of sea ice and its tensile strength. In ArcticSea Ice. National Academy of Sciences—National Research Council1. There are no one-to-one analogs of terrestrial sea icePubl. No. 598, pp. 106–138.to the uniform double ridges of Europa that exceed 1000

Bazant, Z. P. 1992. Large-scale fractures of sea ice plates. Proc. of the 11thkm in length. The longest and straightest features in terres-Intern. Assoc. Hydraulic Res. Ice Symp. Banff, Alberta, pp. 991–1005.trial sea ice are simple extensional fractures. However, the

Belton, M. J. S., and 33 colleagues 1996. Galielo’s first images of Jupiterfactors that govern the size and sinuosity of features in and the Galilean satellites. Science 274, 377–385.terrestrial ice are poorly known and merit further study. Brown, R. H., and D. P. Cruikshank 1997. Determination of the composi-

2. The height of topography can exceed the thickness tion and state of icy surfaces in the outer Solar System. In Annualof the floating ice shell if it is isostatically compensated by Review of Earth and Planetary Sciences, Palo Alto, No. 25, (R. Jeanloz,

A. L. Albee, and K. C. Burke, Eds.), pp. 243–277.a buoyant keel. Hence, the maximum elevation of Europa’sCarr, M. H., and 18 colleagues 1998. Evidence for a subsurface oceansurface may be a less significant constraint on the crustal

on Europa. Nature 391, 362–365.thickness of the satellite than the depth of its largest cavity.Cassen, P. M., S. J. Peale, and R. T. Reynolds 1982. Structure and thermalFuture imaging might allow the question to be addressed

evolution of the Galilean satellites. In Satellites of Jupiter (D. Morrison,directly using craters 1 to 10 km in diameter as probes of Ed.), pp. 93–128. Univ. Arizona Press, Tucson.Europa’s subsurface. Studies of the morphology of the Coon, M. D. 1974. Mechanical behavior of compacted arctic ice floes.three large (.20 km) craters seen on Europa are beginning J. Petroleum Tech. Apr. 1994, 466–470.to yield clues to the structure of the subsurface (Moore Coon, M. D. 1980. A review of AIDJEX modeling. In Sea Ice Processeset al. 1997). It would be useful to augment these theoretical and Models, pp. 12–27. Univ. Washington Press, Seattle.efforts with experimental data on cratering in sea ice. Indi- Coon, M. D., G. S. Knoke, D. C. Echert, and R. S. Pritchard 1998. The

architecture of an anisotropic elastic–plastic sea ice mechanics model.rect indications of ice thickness could include flexure ofJ. Geophys. Res., submitted.the surface and the height and width of icy plates.

den Hartog, G., S. D. Smith, R. J. Anderson, D. R. Topham, and R. G.3. The record of past deformation is preserved in seaPerkin 1983. An investigation of a polynya in the Canadian Archipel-ice from season to season. Terrestrial examples show that ago. J. Geophys. Res. 88, 2911–2916.

morphologically distinct ridges can result from a single set Evans, R. J., and N. Untersteiner 1971. Thermal cracks in floating iceof processes at different stages of evolution and do not sheets. J. Geophys. Res. 76, 694–703.necessarily require gradual changes in ice thickness or Fink, U., N. H. Dekkers, and H. P. Larson 1973. Infrared spectra of the

Galilean satellites of Jupiter. Astrophys. J. 179, L155–L159.other properties of the sea ice.Geissler, P. E., and 14 colleagues 1997. Evidence for non-synchronous

We note that by focusing on the morphology and surface rotation of Europa. Nature 391, 368–370.appearance of sea ice we have neglected many potentially Golombek, M. P., and W. B. Banerdt 1990. Constraints on the subsurfaceimportant aspects of sea ice such as the thermal properties structure of Europa. Icarus 83, 441–452.

Greeley, R., and 20 colleagues 1998. Europa: Initial Galileo geologicaland the spatial distribution of contaminants, such as salts.observations. Icarus 135, 4–24.Results from radar sounding and passive seismic surveys

Greenberg, R., P. Geissler, G. Hoppa, B. R. Tufts, D. D. Durda, R.of floating ice could be studied to provide insight into dataPappalardo, J. W. Head, R. Greeley, R. Sullivan, and M. H. Carr 1998.future missions to Europa that might carry these instru-Tectonic processes on Europa: Tidal stresses, mechanical response,

ments. and visible features. Icarus 135, 64–78.

Helfenstein, P., and 21 colleagues 1998. Galileo observations of Europa’sopposition effect. Icarus 135, 41–63.ACKNOWLEDGMENTS

Hogenboom, D. L., J. S. Kargel, G. J. Consolmagno, T. C. Holden, L.We thank D. B. Snyder, R. A. Schultz (Arizona State University), and Lee, and M. Buyyounouski 1997. The ammonia–water system and the

chemical differentiation of icy satellites. Icarus 128, 171–180.N. Christy (Brown University) for compiling images and background

40 GREELEY ET AL.

Kuiper, G. P. 1957. Infrared observations of planets and satellites. Astron. compositions, and internal structures of the moons of the Solar System.In Satellites (J. A. Burns and M. S. Matthews, Eds.), pp. 224–292. Univ.J. 62, 295.Arizona Press, Tucson.Kovacs, A., W. F. Weeks, S. Ackley, and W. D. Hibler III 1972. A study

of a multi-year pressure ridge in the Beaufort Sea. Arctic Ice Dynamics Schwarz, J., and W. F. Weeks 1977. Engineering properties of sea ice.J. Glaciol. 19, 499–531.Joint Exp., Division of Marine Resources, Univ. Washington, Bulletin

No. 12, pp. 17–28. Smith, B. A., and 21 colleagues 1979a. The Jupiter system through theeyes of Voyager 1. Science 204, 951–972.Lucchitta, B. K., and L. A. Soderblom 1982. Geology of Europa. In

Satellites of Jupiter (D. Morrison, Ed.), pp. 521–555. Univ. Arizona Smith, B. A., and 21 colleagues 1979b. The Galilean satellites and Jupiter:Press, Tucson. Voyager 2 imaging results. Science 206, 927–950.

Malin, M. C., and D. C. Pieri 1986. Europa. In Satellites (J. A. Burns and Squyres, S. W., R. T. Reynolds, P. M. Cassen, and S. J. Peale 1983. LiquidM. S. Matthews, Eds.), pp. 689–717. Univ. Arizona Press, Tucson. water and active resurfacing on Europa. Nature 301, 225–226.

Michel, B. 1978. Ice Mechanics. Les Presses De L’universite Laval, Steffen, K. 1986. Atlas of the sea ice types, deformation processes andQuebec. openings in the ice (North Water Project). Zurcher Geographische

Schriften 20. Geographischer Institut Eidenossische TechnischeMoore, J. M., K. C. Bender, R. J. Sullivan, R. Greeley, A. S. McEwen,Hochschule, Zurich.B. R. Tufts, J. W. Head III, R. T. Pappalardo, M. J. S. Belton, and

the Galileo SSI Team 1997. Europan macula: Possible origins. Lunar Sullivan, R., and 10 colleagues 1997. Galileo views of crustal disruptionPlanet. Sci. 28, 973–974. on Europa. Lunar Planet Sci. 28, 1395–1396.

Sullivan, R., R. Greeley, K. Homan, J. Klemaszewski, M. J. S. Belton,Pappalardo, R. T., and M. D. Coon 1996. A sea ice analog for the surfaceM. H. Carr, C. R. Chapman, R. Tufts, J. W. Head III, R. Pappalardo,of Europa. Lunar Planet Sci. 28, 997–998.J. Moore, P. Thomas, and the Galileo Imaging Team 1998. EpisodicPappalardo, R. T., and R. J. Sullivan 1996. Evidence for crustal separationplate separation and fracture infill on the surface of Europa. Natureon Europa. Icarus 123, 557–567.391, 371–373.

Pappalardo, R. T., and 10 colleagues 1998. Geological evidence for solid-Taylor, R. B. 1978. The occurrence of grounded ice ridges and shorestate convection in Europa’s ice shell. Nature 391, 365–367.

ice piling along the north coast of Somerset Island, N.W.T. ArcticParmerter, R., and M. D. Coon 1972. A model of pressure ridge formation 31, 133–149.

in sea ice. J. Geophys. Res. 77, 6565–6575.Tufts, B. R. 1996. A San-Andreas-sized strike-slip fault on Europa. Lunar

Pease, C. H. 1987. The size of wind-driven coastal polynyas. J. Geophys. Planet. Sci. 27, 1343–1344.Res. 92, 7049–7059.

Tufts, B. R., R. Greenberg, R. Sullivan, R. Pappalardo, and the GalileoPieri, D. C. 1981. Lineament and polygon patterns on Europa. Nature Imaging Team 1997. Reconstruction of europan terrain in the Galileo

289, 17–21. C3 ‘‘wedges’’ image and its geological implications. Lunar Planet Sci.Pilcher, C. B., S. T. Ridgway, and T. B. McCord 1972. Galilean satellites: 28, 1455–1456.

Identification of water frost. Science 178, 1087–1089. Warren, C. R., N. F. Glasser, S. Harrison, V. Winchester, A. R. Rerr,Schenk, P. M., and W. B. McKinnon 1989. Fault offsets and lateral crustal and A. Rivera 1994. Characteristics of tide-water calving at Glacier

movement on Europa: Evidence for a mobile ice shell. Icarus 79, San Rafael Chile. J. Glaciol. 40, 273–289.75–100. Weeks, W. F., and A. Assur 1969. Fracture of Lake and Sea Ice. U.S.

Schenk, P. M., and C. K. Seyfert 1980. Fault offsets and proposed plate Army Cold Reg. Res. Eng. Lab. Research Report. Hanover, NH.motions for Europa. EOS Trans. Amer. Geophys. Union 61, 286. Weeks, W. F., and A. Kovacs 1970. On Pressure Ridges. Rept., U.S. Army

Cold Reg. Res. Eng. Lab. Report. Hanover, NH.Schubert, G., T. Spohn, and R. T. Reynolds 1986. Thermal histories,

Documents

Terrestrial Sea Ice Morphology: Considerations for …Two forms of ice are encountered on the seas, true sea ice, formed by the freezing of salt water at the ocean’s surface, and