Printer Friendly

Planetary Volcanism.

Planetary Volcanism

Planetary Volcanism

Lionel Wilson, Lancaster UniversityLancaster, United Kingdom

Volcanism is one of the major processes whereby a planet transfers heat produced in its interior outward to the surface. Volcanic activity has been directly responsible for forming at least three quarters of the surface rocks of Earth and Venus, all of the surface materials of Jupiter's satellite Io, and extensive parts of the surfaces of Mars, Earth's Moon, and probably Mercury. Investigations of the styles of volcanic activity (e.g., explosive or effusive) on a planet's surface, when viewed in the light of environmental factors such as atmospheric pressure and acceleration due to gravity, provide clues to the composition of the erupted magma and hence, indirectly, to the chemical composition of the interior of the planet and its thermal state and history. Investigations of volcanic features on other planets have been an important spur to the development of an understanding of volcanic processes on Earth.

1. Summary of Planetary Volcanic Features

1.1 Earth

Only in the middle part of the 20th century did it become entirely clear that the [?]70% of Earth's surface represented by the crust forming the floors of the oceans consists of geologically very young volcanic rocks. These erupted from long lines of volcanoes, generally located along ridges near the centers of ocean basins, within the last 300 Ma (million years). Along with this realization came the development of the theory of plate tectonics, which explained the location and distribution of volcanoes over Earth's surface. Volcanoes erupting relatively metal-rich, silica- and volatile-poor magmas (called basalts) tend to concentrate along the midocean ridges, which mark the constructional margins of Earth's rigid crust plates. The semagmas represent the products of the partial melting of the mantle at the tops of convection cells in which temperature variations cause the solid mantle to deform and flow on very long timescales. Magma compositions are very closely related to the bulk composition of the mantle, which makes up most of Earth's volume outside of the iron-dominated core. The volcanic edifices produced by ocean-floor volcanism consist mainly of relatively fluid (low-viscosity) lava flows with lengths from a few kilometers to a few tens of kilometers. Lava flows erupted along the midocean ridges simply add to the topography of the edges of the growing plates as they move slowly ([?]10 mm/year) away from the ridge crest. [See EARTH AS A PLANET: ATMOSPHERE AND OCEANS; EARTH AS A PLANET: SURFACE AND INTERIOR.]

Lavas erupted from vents located some distance away from the ridge crest build up roughly symmetrical edifices that generally have convex-upward shapes and are described, depending on their height-to-width ratio, as shields (having relatively shallow flank slopes)or domes (having relatively steeper flanks). Some of these vent systems are not related to the spreading ridges at all, but instead mark the

FIGURE 1 A Hawaiian-style lava fountain feeding a lava flow and building a cinder cone (Pu'u 'O'o on the flank of Kilauea volcano in Hawai'i). (Photograph by P. J. Mouginis-Mark.)

locations of "hot spots" in the underlying mantle, vigorously rising plumes of mantle material from which magmas migrate through the overlying plate. Because the plate moves over the hot spot, a chain of shield volcanoes can be built up in this way, marking the trace of the relative motion. The largest shield volcanoes on Earth form such a line of volcanoes, the Hawaiian Islands, and the two largest of these edifices, Mauna Loa and Mauna Kea, rise [?]10 km above the ocean floor and have basal diameters of about 200 km.

Eruptive activity on shield volcanoes tends to be concentrated either at the summit or along linear or arcuate zones radiating away from the summit, called rift zones. The low viscosity of the basaltic magmas released in Hawaiian-style eruptions on these volcanoes (Fig. 1) allows the lava flows produced to travel relatively great distances (a few tens of kilometers), and is what gives shield volcanoes their characteristic wide, low profiles. It is very common for a long-lived reservoir of magma, a magma chamber, to exist at a depth of a few to several kilometers below the summit. This reservoir, which is roughly equant in shape and may be up to 1 to 3 km in diameter, intermittently feeds surface eruptions, either when magma ascends vertically from it in the volcano summit region or when magma flows laterally in a subsurface fracture called a dike, which most commonly follows an established rift zone, to erupt at some distance from the summit. In many cases, magma fails to reach the surface and instead freezes within the fracture it was following, thus forming an intrusion. The summit reservoir is fed, probably episodically, from partial melt zones in the mantle beneath. Rare but important events in which a large volume of magma leaves such a reservoir lead to the collapse of the rocks overlying it, and a characteristically steep-sided crater called a caldera is formed, with a width similar to that of the underlying reservoir.

Volcanoes erupting silica- and volatile-rich magma (andesite or, less commonly, rhyolite) mark the destructive margins of plates, where the plates bend downward to be subducted into the interior and at least partly remelted. These volcanoes tend to form an arcuate pattern (called an island arc when the volcanoes rise from the sea floor), marking the trace on the surface of the zone where the melting is taking place, at depths on the order of 100-150 km. The andesitic magmas thus produced represent the products of the melting of a mixture of subducted ocean floor basalt, sedimentary material that had been washed onto the ocean floor from the continents (which are themselves an older, silica-rich product of the chemical differentiation of Earth), seawater trapped in the sediments, and the primary mantle materials into which the plates are subducted. Thus, andesites are much less representative of the current composition of the mantle. Andesite magmas are rich in volatiles (mainly water, carbon dioxide, and sulfur compounds), and their high silica contents give them high viscosities, making it hard for gas bubbles to escape. As a result, andesitic volcanoes often erupt explosively in Vulcanian-style eruptions, producing localized pyroclastic deposits with a range of grain sizes; alternatively, they produce relatively viscous lava flows that travel only short distances (a few kilometers) from the vent. The combination of short flows and localized ash deposits tends to produce steep-sided, roughly conical volcanic edifices.

When large bodies of very silica- and volatile-rich magma (rhyolite) accumulate--in subduction zones or, in some cases, where hot spots exist under continental areas, leading

FIGURE 2 The upper three layers of gray, dark, and bright material are air-fall pyroclastic deposits from the 1875 Plinian eruption of Askja volcano in Iceland. They clearly mantle earlier, dark, more nearly horizontal pyroclastic deposits. (Photograph by L. Wilson.)

to extensive melting of the continental crustal rocks--the potential exists for the occurrence of very large scale explosive eruptions in which finely fragmented magma is blasted at high speed from the vent to form a convecting eruption cloud, called a Plinian cloud, in the atmosphere. These clouds may reach heights up to 50 km, from which pyroclastic fragments fall to create a characteristic deposit spreading downwind from the vent area (Fig. 2). Under certain circumstances, the cloud cannot convect in a stable fashion and collapses to form a fountain-like structure over the vent, which feeds a series of pyroclastic flows--mixtures of incandescent pyroclastic fragments, volcanic gas, and entrained air--that can travel for at least tens of kilometers from the vent at speeds in excess of 100 m/s, eventually coming to rest to form a rock body called ignimbrite. These fall and flow deposits may be so widespread around the vent that no appreciable volcanic edifice is recognizable; however, there may be a caldera, or at least a depression, at the vent site due to the collapse of the surface rocks to replace the large volume of material erupted from depth.

It should be clear from the foregoing descriptions that the distribution of the various types of volcano and characteristic volcanic activity seen on Earth are intimately linked with the processes of plate tectonics. A major finding to emerge from the exploration of the solar system over the last 30 years is that this type of large-scale tectonism is currently confined to the Earth and may never have been active on any of the other bodies. Virtually all of the major volcanic features that we see elsewhere can be related to the eruption of mantle melts similar to those associated with the midocean ridges and oceanic hot spots on Earth. However, differences between the physical environments (acceleration due to gravity, atmospheric conditions) of the other planets and Earth lead to significant differences in the details of the eruption processes and the deposits and volcanic edifices formed.

1.2 The Moon

During the 1970s, analyses of the samples collected from the Moon by the Apollo missions showed that there were two major rock types on the lunar surface. The relatively bright rocks forming the old, heavily cratered highlands of the Moon were recognized as being a primitive crust that formed about 4.5 Ga (billion years) ago by the accumulation of solid minerals at the cooling top of an at least 300 km thick melted layer referred to as a magma ocean. This early crust was extensively modified prior to about 3.9 Ga ago by the impacts of meteoroids and asteroids with a wide range of sizes to form impact craters and basins. Some of the larger craters and basins (the mare basins) were later flooded episodically by extensive lava flows, many more than 100 km long, to form the darker rocks visible on the lunar surface. [See THE MOON; PLANETARY IMPACTS.]

Radiometric dating of samples from lava flow units showed that these mare lavas were mostly erupted between 3 and 4 Ga ago, forming extensive, relatively flat deposits inside large basins. Individual flow units, or at least groups of flows, can commonly be distinguished using multispectral remote sensing imagery on the basis of their differing chemical compositions, which give them differing reflectivities in the visible and near-infrared parts of the spectrum. In composition, these lavas are basaltic, and their detailed mineralogy shows that they are the products of partial melting of the lunar mantle at depths between 150 and more than 400 km, the depth of origin increasing with time as the lunar interior cooled. Melting experiments on samples, supported by theoretical calculations based on their mineralogies, show that these lavas were extremely fluid (i.e., had very low viscosities, at least a factor of 3 to 10 less than those of typical basalts on Earth) when they were erupted. This allowed them to travel for great distances, often more than 100 km (Fig. 3) from their vents; it also meant that they had a tendency to flow back into, and cover up, their vents at the ends of the eruptions. Even so, it is clear from the flow directions that the vents were mainly near the edges of the interiors of the basins that the flows occupy. Many vents were probably associated with the arcuate rilles found in similar positions. These are curved grabens, trench-like depressions parallel to the edges of the basins formed as parts of the crust sink between pairs of parallel faults caused by tension. This tension, due to the weight of the lava ponded in the middle of the basin, makes it easier for cracks filled with magma to reach the surface in these places.

FIGURE 3 Lava flows in southwest Mare Imbrium on the Moon. The source vents are off the image to the lower left and the [?]300 km long flows extend down a gentle slope toward the center of the mare basin beyond the upper right edge of the frame. (NASA Apollo photograph.)

A second class of lunar volcanic features associated with the edges of large basins is the sinuous rilles. These are meandering depressions, commonly hundreds of meters wide, tens of meters deep, and tens of kilometers long, which occur almost entirely within the mare basalts. Some are discontinuous, giving the impression of an underground tube that has been partly revealed by partial collapse of its roof, and these are almost certainly the equivalent of lava tube systems (lava flows whose top surface has completely solidified) on Earth. Other sinuous rilles are continuous open channels all along their length; these generally have origins in source depressions two or three times wider than the rille itself, and become narrower and shallower with increasing downslope distance from the source. At least some of these sinuous rilles appear to have been caused by long-duration lava flows that were very turbulent (i.e., the hot interior was being constantly mixed with the cooler top and bottom of the flow). As a result the flows were able to heat up the preexisting surface until some of its minerals melted, allowing material to be carried away and an eroded channel to form.

In contrast to the lava flows and lava channels, two types of pyroclastic deposit are recognized on the Moon. There are numerous regions called dark mantles, often roughly circular and up to at least 200 km in diameter, where the fragmental lunar surface regolith is less reflective than usual, and spectroscopic evidence shows that it contains a component of small volcanic particles in addition to the locally derived rock fragments. The centers of these regions are commonly near the edges of mare basins, suggesting that the dark mantle deposits are produced by the same (or similar) source vents as the lava flows. Chemical analyses of the Apollo lava samples show that the Moon's mantle is totally devoid of common volatiles like water and carbon dioxide due to its hot origin [see THE MOON] and suggest that the main gas released from mare lava vents was carbon monoxide, produced in amounts up to a few hundred parts per million by weight as a result of a chemical reaction between free carbon and metal oxides, mainly iron oxide, in the magma as it neared the surface.

Several smaller, dark, fragmental deposits occur on the floor of the old, 90 km diameter impact crater Alphonsus. These patches, called dark haloes, extend for a few kilometers from the rims of subdued craters that are centered on, and elongated along, linear fault-bounded depressions (called linear rilles) on the crater floor. It is inferred that these are the sites of less energetic volcanic explosions.

Localized volcanic constructs such as shield volcanoes and domes are generally rare on the Moon, though more than 200 low, shieldlike features with diameters mainly in the range 3-10 km are found in the Marius region within Oceanus Procellarum, in northeast Mare Tranquillitatis, and in the region between the craters Kepler and Copernicus. Conspicuously absent are edifices with substantial summit calderas. This implies that large, shallow magma reservoirs are very rare, almost certainly a consequence of the difficulty with which very dense magmas rising from the mantle penetrate the low-density lunar crust. However, a few collapse pits with diameters up to 3 km do occur, located near the tops of domes or aligned along linear rilles.

1.3 Mars

About 60% of the surface of Mars consists of an ancient crust containing impact craters and basins. Spectroscopic evidence from orbiting spacecraft suggests that it is composed mainly of volcanic rocks. The other 40% of the planet consists of relatively young, flat, lower lying, plains-forming units that are a mixture of wind-blown sediments, lava flows, and rock debris washed into the lowlands by episodes of water release from beneath the surface. Combining orbital observations with analyses made by the five probes that have so far landed successfully on the surface

FIGURE 4 The Olympus Mons shield volcano on Mars with the Hawaiian Islands superimposed for scale. (NASA image with overlay by P. J. Mouginis-Mark. Reproduced by permission of the Lunar and Planetary Institute.)

suggests that most of the magmas erupted on Mars are basalts or basaltic andesites. [See MARS: SURFACE AND INTERIOR; MARS: LANDING SITE GEOLOGY, MINERALOGY, AND GEOCHEMISTRY.]

The most obvious volcanic features on Mars are four extremely large ([?]600 km diameter, heights up to >20 km) shield volcanoes (Olympus Mons, Ascraeus Mons, Pavonis Mons, and Arsia Mons) with the same general morphology as basaltic shield volcanoes found on Earth (Fig. 4). There are also about 20 smaller shields on Mars in various stages of preservation. Counts of small impact craters seen in high-resolution ([?]10 m/pixel) spacecraft images show that the ages of the lava flow units on the volcanoes range from more than 3 Ga to less than [?]50 Ma. Complex systems of nested and intersecting calderas are found on the larger shields, implying protracted evolution of the internal plumbing of each volcano, typified by cycles of activity in which a volcano is sporadically active for [?]1 Ma and then dormant for [?]100 Ma. Individual caldera depressions are up to at least 30 km in diameter, much larger in absolute size than any found on Earth, and imply the presence of very large shallow magma reservoirs during the active parts of the volcanic cycles. The large size of these reservoirs, like that of the volcanoes themselves, is partly a consequence of the low acceleration due to gravity on Mars and partly due to the absence of plate tectonics, which means that a mantle hot spot builds a single large volcano, rather than a chain of small volcanoes as on Earth. The availability of large volumes of melt in the mantle beneath some of the largest shield volcanoes has led to the production of giant swarms of dikes, propagating radially away from the volcanic centers for more than 2000 km in some cases.

Most shields appear to have flanks dominated by lava flows, many more than 100 km long. The flanks of Elysium Mons contain some sinuous channels like the sinuous rilles on the Moon that we think are caused by hot, turbulent, high-speed lavas melting the ground over which they flow. Some of the older and more eroded edifices, like Tyrrhena Patera and Hadriaca Patera, appear to contain high proportions of relatively weak, presumably pyroclastic, rocks. There is a hint, from the relative ages of the volcanoes and the stratigraphic positions of the mechanically weaker layers within them, that pyroclastic eruptions were commoner in the early part of Mars' history. More contentious is the suggestion that some of the plains-forming units, generally interpreted as weathered lava flows, in fact consist of pyroclastic fall or flow deposits.

1.4 Venus

Because of its dense, optically opaque atmosphere, the only detailed synoptic imaging of the Venus surface comes from orbiting satellite-based radar systems. Despite the differences between optical and radar images (radar is sensitive to both the dielectric constant and the roughness of the surface on a scale similar to the radar wavelength), numerous kinds of volcanic features have been unambiguously detected on Venus. Large parts of the planet are covered with plains-forming units consisting of lava flows, having well-defined lobate edges and showing the clear control of topography on their direction of movement (Fig. 5). The lengths (which can be up to several hundred kilometers) and thicknesses (generally significantly less than 30 m, since they are not resolvable in the radar altimetry data) of these flows suggest that they are basaltic in composition. This interpretation is supported by the (admittedly small) amounts of major-element chemical data obtained from six of the Soviet probes that soft-landed on the Venus surface. Some areas show concentrations of particularly long flows called fluctus (Latin for floods). Most of the lava plains, judging

FIGURE 5 A variety of radar-bright lava flows radiate from the summit area down the flanks of a shield volcano on Venus. (NASA Magellan image.)

by the numbers of superimposed impact craters, were emplaced within the last [?]700 Ma. [See VENUS: SURFACE AND INTERIOR].

Many areas within the plains and within other geological units contain groupings (dozens to hundreds) of small volcanic edifices, from less than one to several kilometers in diameter, with profiles that lead to their being classified as shields or domes. These groupings are called shield fields, and at least 500 have been identified. Some of the individual volcanoes have small summit depressions, apparently due to magma withdrawal and collapse, and others are seen to feed lava flows. Quite distinct from these presumably basaltic shields and domes is a class of larger, steep-sided domes (Fig. 6) with diameters of a few tens of kilometers and heights up to [?]1 km. The surface morphologies of these domes suggest that most were emplaced in a single episode, and current theoretical modeling shows that their height-to-width ratio is similar to that expected for highly viscous silicic (perhaps rhyolitic) lavas on Earth.

Many much larger volcanic constructs occur on Venus. About 300 of these are classed as intermediate volcanoes and have a variety of morphologies, not all including extensive lava flows. A further 150, with diameters between 100 and about 600 km, are classed as large volcanoes. These are generally broad shield volcanoes that have extensive systems of lava flows and heights above the surrounding plains of up to about 3 km.

Summit calderas are quite common on the volcanoes, ranging in size from a few kilometers to a few tens of kilometers. There are two particularly large volcano-related depressions, called Sacajawea and Colette, located on the

FIGURE 6 A cluster of [?]25 km diameter "pancake" domes on Venus. These domes are evidence of the eruption of lava, which is more viscous than that forming the majority of flows on Venus. (NASA Magellan image.)

upland plateau Lakshmi Planum. With diameters on the order of 200 km and depths of [?]2 km, these features appear to represent the downward sagging of the crust over some unusually deep-seated site of magma withdrawal.

Finally, there are a series of large, roughly circular features on Venus, which, though intimately linked with the large-scale tectonic stresses acting on the crust (they range from a few hundred to a few thousand kilometers in diameter), also have very strong volcanic associations. These are the coronae, novae, and arachnoids. Though defined in terms of the morphology of circumferential, moatlike depressions, radial fracture systems, and so on, these features commonly contain small volcanic edifices (fields of small shields or domes), small calderas, or lava flows, the latter often apparently fed from elongate vents coincident with the distal parts of radial fractures. In such cases, it seems extremely likely that the main feature is underlain by some kind of magma reservoir that feeds the more distant eruption sites via lateral dike systems.

1.5 Mercury

Much of the surface of Mercury is a heavily cratered ancient terrain like that of the Moon. There are some relatively flat plains-forming units dispersed among the craters, and it is tempting to speculate that these contain lava flows. Half of the surface of the planet was imaged by the flyby probe Mariner 10, but the resolution of the images is too poor to allow the lobate edges of any flow units to be identified unambiguously. Earth-based spectroscopic measurements suggest that many of the surface rocks are similar to basalts in composition. In places, patches of materials with these kinds of compositions have shapes consistent with explosive volcanic processes like those that we know occurred early in

FIGURE 7 The upper part of the figure shows the chain of calderas called Tvashtar Catena on Io, showing a fissure eruption in progress. The high temperature of the lava overloaded the spacecraft imaging system causing "bleeding" of data values down vertical lines of the image. Using later images, the appearance of the eruption as it would have been seen by human eyes was reconstructed as shown in the lower part of the figure. (NASA Galileo image.)

the history of the Moon, but this does not in itself guarantee that these materials on Mercury were emplaced volcanically after the era of early intense bombardment that created the craters. [See MERCURY.]

1.6 Io

The bulk density of Io is about the same as that of Earth's Moon, suggesting that it has a silicate composition, similar to that of the inner, Earth-like planets. Io and the Moon also have similar sizes and masses, and it might therefore be expected by analogy with the Moon's thermal history that any volcanic activity on Io would have been confined to the first one or two billion years of its life. However, as the innermost satellite of the gas-giant Jupiter, Io is subjected to strong tidal forces. An orbital period resonance driven by the mutual gravitational interactions of Io, Europa, and Ganymede causes the orbit of Io to be slightly elliptical. This, coupled with the fact that it rotates synchronously (i.e., the orbital period is the same as that of the axial rotation), means that the interior of Io is subjected to a periodic tidal flexing. The inelastic part of this deformation generates heat in the interior on a scale that far outweighs any remaining heat source due to the decay of naturally radioactive elements. As a result, Io is currently the most volcanically active body in the solar system. At any one time, there are likely to be up to a dozen erupting vents. Roughly half of these produce lava flows, generally erupted from fissure vents (Fig. 7) associated with calderas located at the centers of very low shieldlike features, and half produce umbrella-shaped eruption clouds into which gases and small pyroclasts are ejected at speeds of up to 1000 m/s to reach heights up to 300 km (Fig. 8). [See IO: THE VOLCANIC MOON.]

The main gases detected in the eruption clouds are sulfur and sulfur dioxide, and much of the surface is coated with highly colored deposits of sulfur and sulfur compounds that have been degassed from the interior over solar system history and are now concentrated in the near-surface layers. However, it seems very likely, based on the fluid dynamic and thermodynamic analysis of the eruption clouds, that the underlying cause of the activity is the ascent of very hot basic magmas from the interior of Io. Temperatures up to[?]1700-1900 K were initially derived from Galileo spacecraft data, suggesting that the magmas might be ultra-basic, similar to the komatiites that erupted on Earth earlier in its history. However, recent reappraisals of the early analyses suggest somewhat lower temperatures, and models of magma ascent on Io show that basalts, made unusually hot by friction effects as they rise through the crust, are more likely candidates. When these magmas, which may themselves have very low volatile contents, reach the surface in places with few volatile deposits, they produce lava flows. However, when they encounter copious deposits of sulfur compounds, they melt and then vaporize the deposits, providing the very high volatile contents needed to drive the violently explosive eruptions. Most of these volatiles condense as they expand

FIGURE 8 An explosive eruption plume on Io. The great height of the plume, more than 100 km, implies that magma is mixing with and evaporating volatile materials (sulfur or sulfur dioxide) on the surface as it erupts. (NASA Voyager image.)

and cool, and eventually fall back to the surface, providing the materials to drive future explosive eruptions.

1.7 The Icy Satellites

Many of the satellites of the gas-giant planets have bulk densities indicating that their interiors are mixtures of silicate rocks and the ices of the common volatiles (mainly water, with varying amounts of ammonia and methane). On some of these bodies (e.g., Jupiter's satellites Ganymede and Europa, Uranus' satellite Ariel, Neptune's satellite Triton, and Saturn's large satellite Titan), flowlike features that have many of the morphological attributes of very viscous lava flows are seen. However, there is no spectroscopic evidence for silicate magmas having been erupted onto the surfaces of these bodies, and the flowlike features have forced us to recognize that there is a more general definition of volcanism than that employed so far. [See PLANETARY SATELLITES.]

Volcanism is the generation of partial melts from the internal materials of a body and the transport out onto the surface of some fraction of those melts. In the ice-rich bodies, it is the generation of liquid water from solid ice that mimics the partial melting of rocks, and the ability of this water to erupt at the surface is influenced by the amounts of volatiles like ammonia and methane that it contains. Because the surface temperatures of most of these satellites are very much less than the freezing temperature of water, and because they do not have appreciable atmospheres (except Titan), the fate of any liquid water erupting at the surface is complex. Cooling will produce ice crystals at all boundaries of the flow, and these crystals, being less dense than liquid water, will rise toward the flow surface. Because of the negligible external pressure, evaporation (boiling) will take place within the upper few hundred millimeters of the flow. The vapor produced will freeze as it expands, to settle out as a frost or snow on the surrounding surface. The boiling process extracts heat from the liquid and adds to the rate of ice crystal formation. If enough ice crystals collect at the surface of a flow, they will impede the boiling process, and if a stable ice raft several hundred millimeters thick forms, it will suppress further boiling. Thus, if it is thick enough, a liquid water flow may be able to travel a significant distance from its eruption site. It is even possible that solid ice may form flowlike features on a much longer timescale, in essentially the same way that glaciers are able to flow on Earth.

Thick, glacier-like flow features have been detected in flyby radar images of the surface of Titan taken by the Cassini spacecraft in orbit around Saturn. Although they probably consist mainly of water ice, the composition of the other volatile compounds that they may contain is still under debate. One candidate, present as an important addition to the mainly nitrogen atmosphere, is methane. Injection of methane into the atmosphere from cryovolcanic eruptions and its subsequent condensation as "rain" is one possible explanation for the depressions looking strikingly like river valleys imaged on Titan's surface by the Huygens lander probe.

If liquid water produced below the surface of an icy satellite contains a large enough amount of volatiles like ammonia or methane, it will erupt explosively at high speed in what, near the vent, is the equivalent of a Plinian eruption. The expanding volatiles will cause the eruption cloud to spread sideways (like the umbrella-shaped plumes on Io) and disperse the water droplets, rapidly freezing to hailstones, over a wide area. If the eruption speed is high enough and the parent body small enough, some of the smaller hailstones may be ejected with escape velocity. Recent data from the Cassini spacecraft provide graphic evidence for this process occurring near the South Pole of Saturn's small satellite Enceladus. The orbit of Enceladus is very close to the brightest of Saturn's many rings, the E ring, which appears to be composed of particles of ice. It now seems clear that these are derived directly from Enceladus, having been ejected fast enough to escape from the satellite but not from Saturn itself. [See PLANETARY RINGS.]

1.8 The Differentiated Asteroids

The meteorites that fall to the Earth's surface are fragments ejected from the surfaces of asteroids during mutual collisions. Most of these meteorites are pieces of silicate rock and, even though many have rather simple chemical compositions consistent with their never having been strongly heated, it has long been realized that the mineralogy of some others can only be explained if they are either solidified samples of what was once magma or pieces of what was once a mantle that partially melted and then cooled again after melt was removed from it. Additionally, some meteorites are pieces of a nickel-iron-sulfur alloy that was once molten but subsequently cooled slowly. Taken together, these observations imply that some asteroids went through a process of extensive chemical differentiation by melting to form a crust, mantle, and core. The trace element composition of the meteorites from these differentiated asteroids shows that they were heated by the radioactive decay of a group of short-half-life isotopes that were present at the time the solar system formed, the most important of which was [sup.26]Al, which has a half life of [?]0.75 Ma. Thus, all the heating, melting, and differentiation must have taken place within an interval of only a few million years. Yet during this brief period, quite small asteroids, only [?]100-500 km in diameter, were undergoing patterns where the mantle melts, the melt rises to the surface, and explosive and effusive eruptions occur. Such activity began on Earth, Mars, and Venus many tens of million years later.

Spectroscopic evidence very strongly suggests that the asteroid 4 Vesta is the parent body of one group of surface, crust, and mantle rocks, the Howardite-Eucrite-Diogenite group of meteorites. We have not yet identified any other parent asteroids with as much certainty, but we know from their composition that the Aubrites and the Ureilite meteorites are rocks from the mantles of two different asteroids that had violently explosive eruptions, which ejected what should have become their crustal rocks into space at escape velocity. And the Acapulcoites and Lodranites are rocks from the shallow crust or upper mantle of a body that produced rather small amounts of gas during melting in its mantle so that in these meteorites we see gas bubbles trapped in what was once magma traveling through fractures toward the surface. The importance of these meteorites is that they give us copious samples of the very deep interiors of their parent bodies as well as the surfaces; such samples will not be available for a very long time for Venus and Mars and are rare even for the Earth. [See METEORITES; MAIN-BELT ASTEROIDS.]

2. Classification of Eruptive Processes

Volcanic eruption styles on Earth were traditionally classified partly in terms of the observed composition and dispersal of the eruption products. Over the last 20 years, it has been realized that they might be more systematically classified in terms of the physics of the processes involved. This has the advantage that a similar system can be adopted for all planetary bodies, automatically taking account of the ways in which local environmental factors (especially surface gravity and atmospheric pressure) lead to differences in the morphology of the deposits of the same process occurring on different planets.

Eruptive processes are classified as either explosive or effusive. An effusive eruption is one in which lava spreads steadily away from a vent to form one or more lava flows, whereas an explosive eruption is one in which the magma emerging through the vent is torn apart, as a result of the coalescence of expanding gas bubbles, into clots of liquid that are widely dispersed. The clots cool while in flight above the ground and may be partly or completely solid by the time they land to form a layer of pyroclasts. There is some ambiguity concerning this basic distinction between effusive and explosive activity because many lava flows form from the recoalescence, near the vent, of large clots of liquid that have been partly disrupted by gas expansion but that have not been thrown high enough or far enough to cool appreciably. Thus, some eruptions have both an explosive and an effusive component.

There is also ambiguity about the use of the word "explosive" in a volcanic context. Conventionally, an explosion involves the sudden release of a quantity of material that has been confined in some way at a high pressure. Most often the expansion of trapped gas drives the explosion process. In volcanology, the term "explosive" is used not only for this kind of abrupt release of pressurized material but also for any eruption in which magma is torn apart into pyroclasts that are accelerated by gas expansion, even if the magma is being erupted in a steady stream over a long time period. Eruption styles falling into the first category include Strombolian, Vulcanian, and phreatomagmatic activity, whereas those falling into the second include Hawaiian and Plinian activity. All of these styles are discussed in detail later.

3. Effusive Eruptions and Lava Flows

Whatever the complications associated with prior gas loss, an effusive eruption is regarded as taking place after lava leaves the vicinity of a vent as a continuous flow. The morphology of a lava flow, both while it is moving and after it has come to rest as a solid rock body, is an important source of information about the rheology (the deformation properties) of the lava, which is determined largely by its chemical composition, and about the rate at which the lava is being delivered to the surface through the vent. Because lava flows basically similar to those seen on Earth are so well exposed on Mars, Venus, the Moon and Io, a great deal of effort has been made to understand lava emplacement mechanisms.

In general, lava contains some proportion of solid crystals of various minerals and also gas bubbles. Above a certain temperature called the liquidus temperature, all the crystals will have melted, and the lava will be completely liquid. Under these circumstances, lavas containing less than about 20% by volume of gas bubbles will have almost perfectly Newtonian rheologies, which means that the rate at which the lava deforms, the strain rate, is directly proportional to the stress applied to it under all conditions. This constant ratio of the stress to the strain rate is called the Newtonian viscosity of the lava. At temperatures below the liquidus but above the solidus (the temperature at which all the components of the lava are completely solid), the lava in general contains both gas bubbles and crystals and has a non-Newtonian rheology. The ratio of stress to strain rate is now a function of the stress, and is called the apparent viscosity. At high crystal or bubble contents, the lava may develop a nonzero strength, called the yield strength, which must be exceeded by the stress before any flowage of the lava can occur. The simplest kind of non-Newtonian rheology is that in which the increase in stress, after the yield strength is exceeded, is proportional to the increase in strain rate: The ratio of the two is then called the Bingham viscosity, and the lava is described as a Bingham plastic.

The earliest theoretical models of lava flows treated them as Newtonian fluids. Such a fluid released on an inclined plane will spread both downslope and sideways indefinitely (unless surface tension stops it, a negligible factor on the scale of lava flows). Some lavas are channeled by preexisting topography, and so it is understandable that they have not spread sideways. However, others clearly stop spreading sideways even when there are no topographic obstacles, and quickly establish a pattern in which lava moves downhill in a central channel between a pair of stationary banks called levees. Also, lavas do not flow downhill indefinitely after the magma supply from the vent ceases: They commonly stop moving quite soon afterward, often while the front of the flow is on ground with an appreciable slope and almost all the lava is still at least partly liquid. Also, liquid lava present in a channel at the end of an eruption does not drain completely out of the channel: A significant thickness of lava is left in the channel floor. These observations led to the suggestion that no lavas are Newtonian, and attempts were made to model flows as the simplest non-Newtonian fluids, Bingham plastics.

The basis of these models is the idea that the finite thickness of the levees or flow front can be used to determine the yield strength of the lava and that the flow speed in the central channel can be used to give its apparent, and hence Bingham, viscosity. Multiplying the central channel width by its depth and the mean lava flow speed gives the volume flux (the volume per second) being erupted from the vent. Laboratory experiments were used to develop these ideas, and they have been applied by numerous workers to field observation of moving flows on Earth and to images of ancient flows on other planets. For flows on Earth, it is possible to deduce all the parameters just listed; for ancient flow deposits, one can obtain the yield strength unambiguously, but only the product of the viscosity and volume flux can be determined.

There is a possible alternative way to estimate the volume flux if it can be assumed that the flow unit being examined has come to rest because of cooling. An empirical relationship has been established for cooling-limited flows on Earth between the effusion rate from the vent and the length of a flow unit, its thickness, and the width of its active channel. If a flow is treated as cooling-limited when in fact it was not (the alternative being that it was volume-limited, meaning that it came to rest because the magma supply from the vent ceased at the end of the eruption), the effusion rate will inevitably be an underestimate by an unknown amount. Cooling-limited flows can sometimes be recognized because they have breakouts from their sides where lava was forced to form a new flow unit when the original flow front came to rest.

Lava rheologies and effusion rates have been estimated in this way for lava flows on Mars, the Moon, and Venus. It should be born in mind, when assessing these estimates, that a major failing of simple models like the Bingham model is that they assign the same rheological properties to all the material in a flow, whereas it is very likely that lava that has resided in a stationary levee near the vent for a long period will have suffered vastly more cooling than the fresh lava emerging from the vent and will have very different properties. More elaborate models have been evolved since the earliest work, including some that apply to broadly spreading lava lobes that do not have a well-defined levee-channel structure, but no model yet accounts for all the factors controlling lava flow emplacement. With this caution, the values found suggest that essentially all the lavas studied so far on the other planets have properties similar to those of basaltic to intermediate (andesitic) lavas on Earth. Many of these lavas have lengths up to several hundred kilometers, to be compared with basaltic flow lengths up to a few tens of kilometers on Earth in geologically recent times, and this implies that they were erupted at much higher volume fluxes than is now common on Earth. There is a possibility, however, that some of these flow lengths have been overestimated. If a flow comes to rest so that its surface cools, but the eruption that fed it continues and forms other flow units alongside it, a breakout may eventually occur at the front of the original flow. A new flow unit is fed through the interior of the old flow, and the cooled top of the old flow, which has now become a lava tube, acts as an excellent insulator. As a result, the breakout flow can form a new unit almost as long as the original flow, and a large, complex compound flow field may eventually form in this way. Unless spacecraft images of the area have sufficiently high resolution for the compound nature of the flows to be clear, the total length of the group of flows will be interpreted as the length of a single flow, and the effusion rate will be greatly overestimated.

There are, however, certain volcanic features on the Moon and Mars that may be more unambiguous indicators of high effusion rates: the sinuous rilles. The geometric properties of these meandering channels--widths and depths that decrease away from the source, lengths of tens to a few hundred kilometers--are consistent with the channels being the result of the eruption of a very fluid lava at a very high volume flux for a long time. The turbulent motion of the initial flow, meandering downhill away from the vent, led to efficient heating of the ground on which it flowed, and it can be shown theoretically that both mechanical and thermal erosion of the ground surface are expected to have occurred on a timescale from weeks to months. The flow, which may have been [?]10 m deep and moving at [?]10 m/s, slowly subsided into the much deeper channel that it was excavating. Beyond a certain distance, the lava would have cooled to the point where it could no longer erode the ground, and it would have continued as an ordinary surface lava flow. The volume eruption rates deduced from the longer sinuous rille channel lengths are very similar to those found for the longest conventional lava flow units; modeling studies show that the turbulence leading to efficient thermal erosion was probably encouraged by a combination of unusually steep slope and unusually low lava viscosity. A few sinuous channels associated with lava plains are visible on Venus, but the lengths of some of the Venus channels are several to ten times as great as those seen on the Moon and Mars. It is not yet clear if the thermal erosion process is capable of explaining these channels by the eruption of low-viscosity basalts, or whether some more exotic volcanic fluid (or some other process) must be assumed.

There are numerous uncertainties in using the foregoing relationships to estimate lava eruption conditions. Thus, there have been many studies of the way heat is transported out of lava flows, taking account of the porosity of the lava generated by gas bubbles, the effects of deep cracks extending inward from the lava surface, and the external environmental conditions--the ability of the planetary atmosphere to remove heat lost by the flow by conduction, convection, and radiation. However, none of these has yet dealt in sufficient detail with turbulent flows, or with the fact that cooling must make the rheological properties of a lava flow a function of distance inward from its outer surface, so that any bulk properties estimated in the ways described earlier can only be approximations to the detailed behavior of the interior of the lava flow. There is clearly some feedback between the way a flow advances and its internal pattern of shear stresses. For example, lava flows on Earth have two basic surface textures. Basaltic flows that have erupted at low effusion rates or while still hot near their vents have smooth, folded surfaces with a texture called pahoehoe (a Hawaiian word), the result of plastic stretching of the outer skin as the lava advances; at higher effusion rates, or at lower temperatures farther from the vent, the surface fractures in a more brittle fashion to produce a very rough texture called 'a'a. A similar but coarser, rough, blocky texture is seen on the surfaces of more andesitic flows. Because there is a possibility of relating effusion rate and composition to the surface roughness of a flow in this way, there is a growing interest in obtaining relatively high resolution radar images of planetary surfaces (and Earth's surface) in which, as in the Magellan images of Venus, the returned signal intensity is a function of the small-scale roughness.

4. Explosive Eruptions

4.1 Basic Considerations

Magmas ascending from the mantle on Earth commonly contain volatiles, mainly water and carbon dioxide together with sulfur compounds and halogens. All of these have solubilities in the melt that are both pressure- and temperature-dependent. The temperature of a melt does not change greatly if it ascends rapidly enough toward the surface, but the pressure to which it is subjected changes enormously. As a result, the magma generally becomes saturated in one or more of the volatile compounds before it reaches the surface. Only a small degree of supersaturation is needed before the magma begins to exsolve the appropriate volatile mixture into nucleating gas bubbles. As a magma ascends to shallower levels, existing bubbles grow by decompression, and new ones nucleate. It is found empirically that after the volume fraction of the magma occupied by the bubbles exceeds some value in the range 65-80%, the foam-like fluid can no longer deform fast enough in response to the shear stresses applied to it and as a result disintegrates into a mixture of released gas and entrained clots and droplets that form the pyroclasts. The eruption is then, by definition, explosive. The pyroclasts have a range of sizes dictated by the viscosity of the magmatic liquid, in turn a function of its composition and temperature, and the rate at which the decompression is taking place, essentially proportional to the rise speed of the magma.

It is not a trivial matter for the volume fraction of gas in a magma to become large enough to cause disruption into pyroclasts. The lowest pressure to which a magma is ever exposed is the planetary surface atmospheric pressure. On Venus, this ranges from about 10 MPa in lowland plains to about 4 MPa at the tops of the highest volcanoes; on Earth, it is about 0.1 MPa at sea level (and 30% less on high volcanoes) but much higher, up to 60 MPa, on the deep ocean floor; on Mars it ranges from about 500 Pa at the mean planetary radius to about 50 Pa at the tops of the highest volcanoes; and it is essentially zero on the Moon and Io. If the magma volatile content is small enough, then even at atmospheric pressure no gas will be exsolved--or at least too little will be exsolved to cause magma fragmentation. Using the solubilities of common volatiles in magmas, calculations show that explosive eruptions can occur on Earth as long as the water content exceeds 0.07 weight percent in basalt. On Mars, the critical level is 0.01 weight percent. On Venus, however, a basalt would have to contain about 2 weight percent water before explosive activity could occur, even at highland sites; this is greater than is common in basalts on Earth and leads to the suggestion that explosive activity may never happen on Venus, at least at lowland sites, or may happen only when some process leads to the local concentration of volatiles within a magma. Examples of this are discussed later. Finally, the negligible atmospheric pressures on the Moon and Io mean that miniscule amounts of magmatic volatiles can in principle cause some kind of explosive activity there.

The preceding discussion assumes that released magmatic volatiles are the only source of explosive activity. However, many Vulcanian and all phreatomagmatic explosive eruptions involve interaction of erupting magma with solid or liquid volatiles already present at the surface (always water or ice on Earth and probably on Mars; mainly sulfur compounds on Io). The total weight fraction of gas in the eruption products in such cases will depend on the detailed nature of the interaction as well as the composition and inherent volatile content of the magma; this is a critical factor in understanding explosive activity on Io.

4.2 Strombolian Activity

Strombolian eruptions, named for the style of activity common on the Italian volcanic island Stromboli, are an excellent example of how the rise speed, gas content, and viscosity of a magma are critical in determining the style of explosive activity that occurs. While the magma as a whole is ascending through a fracture in the planetary crust, bubbles of exsolved gas are rising through the liquid at a finite speed determined by the liquid viscosity and the bubble sizes. If the magma rise speed is negligible, for example, when magma is trapped in a shallow reservoir or a shallow intrusion, and if its viscosity is low, as in the case of a basalt, there may be enough time for gas bubbles to rise completely through the magma and escape into overlying fractures that convey the gas to the surface, where it escapes or is added to the atmosphere if there is one. Subsequent eruption of the residual liquid will be essentially perfectly effusive. If a low-viscosity magma is rising to the surface at a slow enough speed, most of the gas will still escape as bubbles rise to the liquid surface and burst. Because relatively large bubbles (those that nucleated first and have

FIGURE 9 Jets of hot gas and entrained incandescent basaltic pyroclasts ejected from a transient Strombolian explosion on the volcano Stromboli in Italy. (Photograph by L. Wilson.)

decompressed most) will rise faster through the liquid than very small bubbles, it is common in some magmas, especially basalts, for large bubbles to overtake and coalesce with small ones. The even larger bubbles produced in this way rise even faster and overtake additional smaller bubbles. In many cases, a runaway situation develops in which a single large bubble completely fills the diameter of the vent system apart from a thin film of magma lining the walls of the fracture. In extreme cases the bubble may have a much greater vertical extent than its width, in which case it is called a slug of gas. As this body of gas emerges at the surface of the slowly rising liquid magma column, it bursts, and a discrete layer of magma forming the upper "skin" of the bubble or slug disintegrates into clots and droplets up to tens of centimeters in size. These are blown outward by the expanding gas (Fig. 9). The pyroclasts produced accumulate around the vent to form a cinder cone that can be up to several tens of meters in size. The time interval between the emergence of successive bubbles or slugs from a vent may range from seconds to at least minutes, making this a distinctly intermittent type of explosive activity. If the largest rising gas bubble does not completely fill the vent, continuous overflow of a lava lake in the vent may take place to form one or more lava flows at the same time that intermittent explosive activity is occurring, resulting in a simultaneously effusive and explosive eruption.

A second method of producing gas slugs has been suggested for some Strombolian eruptions on Earth, in which gas bubbles form during convection in an otherwise stagnant body of magma beneath the surface and drift upward to accumulate into a layer of foam at the top of the magma body. When the vertical extent of the foam layer exceeds a critical value, it begins to collapse. Liquid magma drains from between the bubbles, and these coalesce into a large gas pocket that can now rise through any available fracture to the surface. The argument is that if a fracture was already present, the high effective viscosity of the foam would have inhibited its rise into the fracture, whereas the viscosity of the pure gas is low enough to allow this to occur. If a fracture was not already present, the changing stresses due to the foam collapse may be able to create one.

As long as any volatiles are exsolved from a low-viscosity magma rising sufficiently slowly to the surface, some kind of Strombolian explosive activity, however feeble, should occur at the vent on any planet, even at the high pressures on Venus or on Earth's ocean floors. Strombolian eruptions commonly involve excess pressures in the bursting bubbles of only a few tenths of a megapascal, so that the amount of gas expansion that drives the dispersal of pyroclasts is small. Pyroclast ranges in air on Earth can be several tens to at most a few hundred meters, and ranges would be much smaller in submarine Strombolian events on the ocean floor or on Venus because of the higher ambient pressure. Subaerial Strombolian eruptions on Mars would eject pyroclasts to distances about three times greater than on Earth because of the lower gravity; as a result, the deposits formed would have a tenfold lower relief than on Earth, and so far few examples have been unambiguously identified in spacecraft images.

4.3 Vulcanian Activity

At the other extreme of a slowly rising viscous magma, it is relatively difficult for gas bubbles to escape from the melt. Particularly if the magma stalls as a shallow intrusion, slow diffusion of gas through the liquid and rise of bubbles in the liquid concentrate gas in the upper part of the intrusion, and the gas pressure in this region rises. The pressure rise is greatly enhanced if any volatiles existing near the surface (groundwater on Earth; ground ice on Mars; sulfur or sulfur dioxide on Io) are evaporated. Eventually the rocks overlying the zone of high pressure break under the stress, and the rapid expansion of the trapped gas drives a sudden, discrete

FIGURE 10 A dense cloud of large and small pyroclasts and gas ejected to a height of a few hundred meters in a transient Vulcanian explosion by the volcano Ngauruhoe in New Zealand. (Image courtesy of the University of Colorado in Boulder, Colorado, and the National Oceanic and Atmospheric Administration, National Geophysical Data Center.)

explosion in which fragments of the overlying rock and of the disrupted magma are scattered around the explosion source: This is called Vulcanian activity (Fig. 10), named for the Italian volcanic island Vulcano. Again, as long as any volatiles are released from magmaor are present in the near-surface layers of the planet, activity of this kind can occur. Several Vulcanian events on Earth involving fairly viscous magmas have been analyzed in enough detail to provide estimates of typical pressures and gas concentrations. Bombs approaching a meter in size ejected to ranges up to 5 km imply pressures as high as a few megapascals in regions that are tens of meters in size and that have gas mass fractions in the explosion products up to 10%.

On Mars, with the same initial conditions, the lower atmospheric pressure would cause much more gas expansion to accelerate the ejected fragments, and the lower atmospheric density would exert much less drag on them; also the lower gravity would allow them to travel farther for a given initial velocity. The result is that the largest clasts could travel up to 50 km. This means that the roughly circular deposit from a localized, point-source explosion would be spread over an area 100 times greater than on Earth, being on average 100 times thinner. Apart from the possibility that the pattern of small craters produced by the impact of the largest boulders on the surface might be recognized, such a deposit, with almost no vertical relief and having very little influence on the preexisting surface, would almost certainly go unnoticed in even the latest spacecraft images, and indeed no such features have yet been identified. However, if the explosion involves a larger, more complex, and especially elongate vent structure, there would not be such large differences. In the Elysium region of Mars, a large, water-carved channel, Hrad Vallis, has a complex elongate source depression that appears to have been excavated by a Vulcanian explosion when a dike injected a sill into the ice-rich permafrost of the cryosphere--the outer several kilometers of the crust, which is so cold that any H[sub.2]O must be present as ice. As heat from the sill magma melted the ice and boiled the resulting water in the cryosphere, violent expansion of the vapor forced intimate mixing of magma and lumps of cryosphere, encouraging ever more vapor production. Soon all the cryosphere above the sill was thrown out in what is called a fuel-coolant explosion (here the fuel is the magma and the coolant is the ice) to produce a deposit extending about 35 km on either side of the 150 km long depression. Residual heat from the magma melted the remaining ice in the shattered cryosphere rocks so that for a while, until it froze again, there was liquid water present to form a characteristic "muddy" appearance in the deposit (Fig. 11).

A Vulcanian explosion on Venus would also be very different from its equivalent on Earth. In this case, however, the high atmospheric pressure would tend to suppress gas expansion and lead to a low initial velocity for the ejecta, and the atmospheric drag would also be high. Pyroclasts that would have reached a range of 5 km on Earth would travel less than 200 m on Venus. On the one hand, this should concentrate the eruption products around the vent and make the deposit more obvious; however, the resolution of the best radar images from Magellan is only [?]75 m, and so such a deposit would represent only three or four adjacent pixels, which again would probably not be recognized.

On the Moon, a number of Vulcanian explosion products have been identified. The dark halo craters on the floor of Alphonsus have ejecta deposits with ranges up to 5 km. Since the Moon has a much lower atmospheric pressure than Mars (essentially zero), the preceding analysis suggests at first sight that lunar Vulcanian explosions should eject material to very great ranges. However, the Alphonsus event seems to have involved the intrusion of basaltic magma into the [?]10 m thick layer of fragmental material forming the regolith in this area, and the strength of the resulting mixture of partly welded regolith and chilled basalt was quite low. Thus, only a small amount of pressure buildup occurred before the retaining rock layer fractured. As a

FIGURE 11 Part of the Hrad Vallis depression in the Elysium Planitia area of Mars. The depression is surrounded by a "muddy" deposit and is interpreted to have formed when a volcanic explosion excavated the depression and threw out a mixture of hot rocks and overlying cryosphere--cold rocks containing ice. (NASA Mars Global Surveyor image.)

result, the initial speeds of the ejected pyroclasts were low, and their ranges were unusually small.

4.4 Hawaiian Activity

In some cases, especially where low-viscosity basaltic magma travels laterally in dikes at shallow depth, enough gas bubble coalescence and bubble rise occurs for much of the gas to be lost into cracks in the rocks above the dike. Magma then emerges from the vent as a lava flow. However, when basaltic magmas rise mainly vertically at appreciable rates (more than about 1 m/s), some gas bubble coalescence occurs but little gas is lost, and the magma is released at the vent in a nearly continuously explosive manner. A lava fountain, more commonly called a fire fountain, forms over the vent, consisting of pyroclastic clots and droplets of liquid entrained in a magmatic gas stream that fluctuates in its upward velocity on a timescale of a few seconds. The largest clots of liquid, up to tens of centimeters in size, rise some way up the fountain and fall back around the vent to coalesce into a lava pond that overflows to feed a lava flow--the effusive part of the eruption--whereas smaller clasts travel to greater heights in the fountain. Some of the intermediate-sized pyroclasts cool as they fall from the outer parts of the fountain and collect around the lava pond in the vent to build up a roughly conical edifice called an ash

FIGURE 12 A Hawaii eruption from the Pu'u 'O'o vent in Hawaii showing a convecting cloud of gas and small particles in the atmosphere above the 300 m high lava fountain (commonly termed fire fountain) of coarser basaltic pyroclasts. (Photograph by P. J. Mouginis-Mark.)

cone, cinder cone, or scoria cone, the term used depending on the sizes of the pyroclasts involved, ash being smallest. Such pyroclastic cones are commonly asymmetric owing to the influence of the prevailing wind.

Atmospheric gases are entrained into the edge of the fire fountain and heated by contact with the hot pyroclasts and mixing with the hot magmatic gas. In this way, a convecting gas cloud is formed over the upper part of the fountain, and this gas entrains the smallest pyroclasts so that they take part fully in the convective motion. The whole cloud spreads downwind and cools, and eventually the pyroclasts are released again to form a layer on the ground, the smallest particles being deposited at the greatest distances from the vent. This whole process, involving formation of lava flows and pyroclastic deposits at the same time, is called Hawaiian eruptive activity (Fig. 12). This style of activity should certainly have occurred on Mars, but may be suppressed in basaltic magmas on Venus by the high atmospheric pressure, especially in lowland areas, unless, as noted earlier, magma volatile contents are several times higher than is common on Earth.

Figure 13 shows qualitatively how the combination of erupting mass flux and magma gas content in a Hawaiian eruption on Earth determines the nature and size of the possible products: a liquid lava pond at the vent that directly feeds lava flows; a pile of slightly cooled pyroclasts accumulating fast enough to weld together and form a "rootless" lava flow; a cone in which almost all of the pyroclasts are welded together; or a cone formed from pyroclasts that have had time to cool while in flight so that none, or only a few, weld on landing. Attempts have been made to quantify the results in Fig. 13 and extend them to other planetary environments. These results confirm that hot lava ponds around vents on Earth are expected to be no more than a few tens of meters wide even at very high mass eruption rates. On the Moon, the greater gas expansion due to the lack of an atmosphere causes very thorough disruption of the magma (even at the low gas contents implied by analysis of the Apollo samples) and gives the released volcanic gas a high speed. This, together with the lower gravity, allows greater dispersal of pyroclasts of all sizes and provides an explanation of the 100-300 km wide dark mantle deposits

FIGURE 13 Schematic indication of the relative influences of the volatile content and the volume eruption rate of magma on the dispersal and thermal state of pyroclastic material produced in explosive eruptions. (Reprinted from Fig. 5 in the Journal of Volcanology and Geothermal Research, Vol. 37, J. W. Head and L. Wilson, Basaltic pyroclastic eruptions: Influence of gas-release patterns and volume fluxes on fountain structure, and the formation of cinder cones, spatter cones, rootless flows, lava ponds and lava flows, pp. 261-271, (c) 1989, with kind permission of Elsevier Science-NL, Sara Burgerhartstraat 25, 1025 KV Amsterdam, The Netherlands.)

as the products of extreme dispersal of the smallest, 30-100 micrometer-sized particles.

Nevertheless, it appears that hot lava ponds up to [?]5 km in diameter could have formed around basaltic vents on the Moon if the eruption rates were high enough--as high as those postulated to explain the long lava flows and sinuous rilles. The motion of the lava in such ponds would have been thoroughly turbulent, thus encouraging thermal erosion of the base of the pond, and this presumably explains why the circular to oval depressions seen surrounding the sources of many sinuous rilles have just these sizes. Similar calculations for the Mars environment show that, as long as eruption rates are high enough, the atmospheric pressure and gravity are low enough on Mars to allow similar hot lava source ponds to have formed there, again in agreement with the observed sizes of depressions of this type that are seen.

Some noticeable differences occur when Hawaiian eruptions take place from very elongate fissure vents. Instead of a roughly circular pyroclastic cone containing a lava pond feeding one main lava flow, a pair of roughly parallel ridges forms, one on either side of the fissure. These are generally called spatter ramparts. Along the parts of the fissure where the eruption rate is highest, pyroclasts may coalesce as they land to form lava flows so that there are gaps in the ramparts from which the flows spread out. One striking example of this has been found so far on Mars (Fig. 14).

FIGURE 14 Mosaic of two images showing fissure vent near Jovis Tholus volcano on Mars. The vent has produced multiple lava flow lobes, probably of basaltic composition. The area shown is 24 km wide. (NASA Mars Odyssey image.)

4.5 Plinian Activity

In the case of a basaltic magma that is very rich in volatiles, or (much more commonly on Earth) in the case of a volatile-rich andesitic or rhyolitic magma, fragmentation in a steadily erupting magma is very efficient, and most of the pyroclasts formed are small enough to be thoroughly entrained by the gas stream. Furthermore, the speed of the mixture emerging from the vent, which is proportional to the square root of the amount of gas exsolved from the magma, will be much higher (perhaps up to 500 m/s) than in the case of a basaltic Hawaiian eruption (where speeds are commonly less than 100 m/s). The fire fountain in the vent now entrains so much atmospheric gas that it develops into a very strongly convecting eruption cloud in which the heat content of the pyroclasts is converted into the buoyancy of the entrained gas. The resulting cloud rises to a height that is proportional to the fourth root of the magma eruption rate (and hence the heat supply rate) and that may reach several tens of kilometers on Earth. Only the very coarsest pyroclasts fall out near the vent, and almost all of the erupted material is dispersed over a wide area from the higher parts of the eruption cloud (Fig. 15). This activity is termed Plinian, after Pliny's description of the A.D. 79 eruption of Vesuvius. Not all eruptions of this type produce stable convection clouds. If the vent is too wide or the eruption speed of the magma is too low, insufficient atmospheric gas may be entrained to provide the necessary buoyancy for convection, and a collapsed fountain forms over the vent, feeding large pyroclastic flows or smaller, more episodic pyroclastic surges.

Mars is the obvious place other than Earth to look for explosive eruption products: The low atmospheric pressure encourages explosive eruptions to occur, and the atmospheric density is high enough to allow convecting eruption clouds to form, at least up to [?]20 km. However, we think that stable eruption clouds much higher than this cannot form on Mars because the atmosphere becomes too thin to provide the amount of entrained gas that is assumed in current theoretical models. In fact, only one potential fall deposit has yet been identified on Mars with any confidence. This is a region on the flank of the shield volcano Hecates Tholus, where, in contrast to the rest of the volcano, small impact craters appear to be hidden by a blanket of fine material in a region about 50 km wide and at least 70 km long. The sizes of the hidden craters suggest that the deposit is [?]100 m thick, giving it a volume of [?]65 km[sup.3]; if we allow for the likely low bulk density of the deposit, this is equivalent to a dense rock volume of 23 km[sup.3]. The volumes of the four summit depressions on Hecates Tholus range from [?]10 to [?]30 km[sup.3], suggesting that they may be calderas produced by collapse of the summit to compensate for the volume removed from a fairly shallow magma storage reservoir in each of a series of eruptions, the most recent of which produced the deposit described above.

FIGURE 15 The Plinian phase of the explosive eruption of Pinatubo volcano in 1991. A dense cloud of large and small pyroclasts and volcanic gases is ejected at high speed from the vent and entrains and heats the surrounding air. Convections then drives the resulting cloud to a height of tens of kilometers, where it drifts downwind, progressively releasing the entrained pyroclasts. (Photo credit: R. S. Culbreth, U. S. Air Force. Photo courtesy of the National Oceanic and Atmospheric Administration, National Geophysical Data Center.)

Although the high magma gas contents needed suggest that large-scale, steady (Plinian) explosive eruptions are rare on Venus, it is possible to calculate the heights to which their eruption clouds would rise. The high density and temperature of the atmosphere lead to rise heights about a factor of 2 lower than on Earth for the same eruption rate, and very large (at least a few tens of meters) clasts may be transported into near-vent deposits. At distances greater than a few kilometers from the vent, pyroclastic fall deposits will not be very different from those on Earth. A few examples of elongate markings on the Venus surface have been proposed as fall deposits, but no detailed analysis of them has yet been carried out.

The conditions that cause a steady explosive eruption to generate pyroclastic flows instead of feeding a stable, convecting eruption cloud are fairly well understood. If the eruption rate exceeds a critical value (which increases with increasing gas content of the mixture emerging through the vent and decreases with increasing vent diameter), stable convection is not possible whatever the nature of the atmosphere. Because pyroclastic flow formation is linked automatically to high eruption rate and, in general, to high eruption speed, which will encourage a great travel distance, it would not be surprising if large-scale pyroclastic flow deposits distributed radially around a vent were the products of high discharge rate eruptions of gas-rich magmas. Many of the flanking deposits of some martian volcanoes, especially Tyrrhena and Hadriaca Paterae, may have been produced in this way.

Theoretical work has shown that pyroclastic flows on Mars may be able to transport quite large blocks of rock (up to several meters in size, similar to those found on Earth) out of the vent and into nearby deposits. These pyroclast sizes are much greater than those expected in fall deposits on Mars, thus making it potentially possible to distinguish flow and fall deposits in future, high-resolution spacecraft images of martian vents. No equivalent work has yet been carried out for Venus, again mainly because of the expectation that voluminous explosive eruptions may be rare under the high atmospheric pressure conditions.

Short-lived or intermittent explosive eruptions (e.g., Vulcanian explosions, phreato-magmatic explosions, or events in which a gas-rich, high-viscosity lava flow or dome disintegrates into released gas and pyroclasts as a result of excessive gas pressure) can also produce small-scale pyroclastic flows. Because these are shorter lived and have characteristically different grain size distributions, they are called surges. The least well understood aspect of these phenomena is the way in which the magmatic material interacts with the atmosphere. As a result, it is currently almost impossible to predict in detail what the results of this kind of activity on Mars or Venus would look like. Such deposits, by the nature of the way they are generated, would not be very voluminous, however, and so would be spread very thinly, and might not be recognized if they were able to travel far from the vent.

4.6 Phreato-Magmatic Activity

Some types of eruption on Earth are controlled by the vigorous interaction of magma with surface or shallow subsurface water. If an intrusion into water-rich ground causes steam explosions, these are called phreatic events (from the Greek word for a well). If some magma also reaches the surface, the term used is phreato-magmatic, as distinct from normal, purely magmatic eruptions. When the equivalents of Strombolian or Hawaiian explosive events take place from eruption sites located in shallow water, they lead to much greater fragmentation of the magma than usual because of the stresses induced as pyroclasts are chilled by contact with the water. This activity is usually called Surtseyan, named after the eruption that formed the island of Surtsey off the south coast of Iceland. A much more vigorous and long-lived eruption under similar circumstances leads to a pyroclastic fall deposit similar to that of a Plinian event, but again involving greater fragmentation of magma: The result is called phreato-Plinian activity. Since the word "phreatic" does not specifically refer to water as the nonmagmatic volatile involved in these kinds of explosive eruption, it seems safe to apply these terms, as appropriate, to the various kinds of interactions between magma and liquid sulfur or sulfur dioxide forming the plumes currently seen on Io. These eruptions appear to involve about 30% by weight volatiles mixed with the magma; these proportions are close to the optimum for converting the heat of the magma to kinetic energy of the explosion products. Phreatic and phreato-magmatic eruptions should also have occurred on Mars in the distant past if, as many suspect, the atmospheric pressure was high enough to allow liquid water to exist on the surface.

4.7 Dispersal of Pyroclasts into a Vacuum

The conditions in the region above the vent in an explosive eruption on a planet with an appreciable atmosphere (e.g., Venus, Earth, or Mars) are very different from those when the atmospheric pressure is very small (much less than about 1 Pa), as on the Moon or Io. If the mass of atmospheric gas displaced from the region occupied by the eruption products after the magmatic gas has decompressed to the local pressure is much less than the mass of the magmatic gas, there is no possibility of a convecting eruption cloud forming in eruptions that would have been classed as Hawaiian or Plinian on Earth. In the region immediately above the vent, the gas expansion must be quite complex and will involve a series of shock waves. Relatively large pyroclasts will pass through these shocks with only minor deviations in their trajectories, but intermediate-sized particles may follow very complex paths, and few studies have yet been made of these conditions. The magmatic gas eventually expands radially into space, accelerating as it expands and reaching a limiting velocity that depends on its initial temperature. As the density of the gas decreases, its ability to exert a drag force on pyroclasts also decreases. On bodies the size of the Moon, even the smallest particles eventually decoupled from the gas and fell back to the planetary surface, though in gas-rich eruptions on asteroids they were commonly ejected into space.

These are the conditions that led to the formation of the dark mantle deposits on the Moon, with ultimate gas speeds on the order of 500 m/s, leading to ranges up to 150 km for small pyroclasts 30-100 micrometers in size. They are also the conditions that exist now in the eruption plumes on Io, though with an added complication. The driving volatiles in the Io plumes appear to be mainly sulfur and sulfur dioxide, evaporated from the solid or liquid state by intimate mixing with rising basaltic magma in what are effectively phreato-magmatic eruptions. The plume heights imply gas speeds just above the vent of [?]1000 m/s, and these speeds are consistent with the plume materials being roughly equal mixtures of basaltic pyroclasts and evaporated surface volatiles. However, as the gas phase expands to very low pressures, both sulfur and sulfur dioxide will begin to condense again, forming small solid particles that rain back onto the surface along with the silicate particles to be potentially recycled again in future eruptions.

A final point concerns pyroclastic eruptions on the smallest atmosphereless bodies, the asteroids. Basaltic partial melts formed within these bodies were erupted at the surface at speeds that depended on the released volatile content. This is estimated to have been as much as 0.2-0.3 weight percent, leading to speeds up to 150 m/s. These speeds are greater than the escape velocities from asteroids with diameters less than about 200 km, and so instead of falling back to the surface, pyroclasts would have been expelled into space, eventually to spiral into the Sun. This process explains the otherwise puzzling fact that we have meteorites representing samples of the residual material left in the mantle of at least two asteroids after partial melting events, but have no meteorites from these asteroids with the expected partial melt composition.

5. Inferences about Planetary Interiors

The presence of the collapse depressions called calderas at or near the summits of many volcanoes on Earth, Mars, Venus, and Io suggests that it is common on all of these bodies for large volumes of magma to accumulate in reservoirs at relatively shallow depths. Theories of magma accumulation suggest that the magma in these reservoirs must have an internal pressure greater than the stress produced in the surrounding rocks by the weight of the overlying crust. This excess pressure may be due to the formation of bubbles by gas exsolution, or to the fact that heat loss from the magma to its cooler surroundings causes the growth of crystals that are less dense than the magmatic liquid and so occupy a larger volume. Most commonly, a pressure increase leads to fracturing of the wall of the reservoir and to the propagation of a magma-filled crack, called a dike, as an intrusion into the surrounding rocks. If the dike reaches the surface, an eruption occurs, and removal of magma from the reservoir allows the wall rocks to relax inward elastically as the pressure decreases. If magma does not reach the surface, the dike propagates underground until either the magma within it chills and comes to rest as its viscosity becomes extremely high, or the pressure within the reservoir falls to the point where there is no longer a great enough stress at the dike tip for rock fracturing to continue.

Under certain circumstances, an unusually large volume of magma may be removed from a shallow reservoir, reducing the internal pressure beyond the point where the reservoir walls behave elastically. Collapse of the overlying rocks may then occur to fill the potential void left by the magma, and a caldera (or, on a smaller scale, a pit crater) will form. The circumstances causing large-volume eruptions on Earth include the rapid eruption to the surface immediately above the reservoir of large volumes of low-density, gas-rich silicic (rhyolitic) magma, and the drainage of magma through extensive lateral dike systems extending along rift zones to distant flank eruption sites on basaltic volcanoes. This latter process appears to have been associated with caldera formation on Kilauea volcano in Hawaii, and it is tempting to speculate that the very large calderas on some of the martian basaltic shield volcanoes (especially Pavonis Mons and Arsia Mons) are directly associated with the large-volume eruptions seen on the distal parts of their rift zones. In contrast, we saw earlier that, at the martian volcano Hecates Tholus, a large explosive summit eruption is implicated in the formation of at least one of its calderas.

The size of a caldera must be related to the volume of the underlying magma reservoir, or more exactly to the volume of magma removed from it in the caldera-forming event. If the reservoir is shallow enough, the diameter of the caldera is probably similar to that of the reservoir. Diameters from 1 to 3 km are common on basaltic volcanoes on Earth and on Venus, with depths up to a few hundred meters implying magma volumes less than about 10 km[sup.3]. In contrast, caldera diameters up to at least 30 km occur on several volcanoes on Mars and, coupled with caldera depths up to 3 km, imply volumes ranging up to as much as 10,000 km[sup.3]. The stresses implied by the patterns of fractures on the floors and near the edges of some of these martian calderas suggest that the reservoirs beneath them are centered on depths on the order of 10-15 km, about three to four times greater than the known depths to the centers of shallow basaltic reservoirs on Earth. The simplest models of the internal structures of volcanoes suggest that, due to the progressive closing of gas cavities in rocks as the pressure increases, the density of the rocks forming a volcanic edifice should increase, at first quickly and then more slowly, with depth. Rising magma from deep partial melt zones may stall when its density is similar to that of the rocks around it so that it is neither positively nor negatively buoyant, and a reservoir may develop in this way. Because the pressure at a given depth inside a volcano is proportional to the acceleration due to gravity, and because martian gravity is about three times less than that on Earth or Venus, the finding that martian magma reservoirs are centered three to four times deeper than on Earth is not surprising. However, these simple models do not address the reason for the martian calderas being much more than three times wider than any of those on Earth or many of those on Venus. On Io, we see some caldera-like structures, not necessarily associated with obvious volcanic edifices, that are even wider (but not deeper) than those on Mars, though we have too little information about the internal structure of Io's crust to interpret this observation unambiguously. Much is still not understood about the formation and stability of shallow magma bodies.

Evidence for significant shallow magma storage is conspicuously absent from the Moon. The large volumes observed for the great majority of eruptions in the later part of lunar volcanic history, and the high effusion rates inferred for them, imply that almost all of the eruptions took place directly from large bodies of magma stored at very great depth--at least at the base of the crust and possibly in partial melting zones in the lunar mantle. Not all the dikes propagating up from these depths will have reached the surface, however, and some shallow dike intrusions almost certainly exist. Recent work suggests that many of the linear rilles on the Moon represent the surface deformation resulting from the emplacement of such dikes, having thicknesses of at least 100 m, horizontal and vertical extents of [?]100 km, and tops extending to within 1 or 2 km of the surface. Minor volcanic activity associated with some of these features would then be the result of gas loss and small-scale magma redistribution as the main body of the dike cooled.

The emplacement of very large dike systems extending most or all of the way from mantle magma source zones to the surface is not confined to the Moon. It has long been assumed that such structures must have existed to feed the high-volume basaltic lava flow sequences called flood basalts that occur on Earth every few tens of millions of years. These kinds of feature are probably closely related to the systems of giant dikes, tens to hundreds of meters wide and traceable laterally for many hundreds to more than 1000 km, that are found exposed in very ancient rocks on the Earth. The radial patterns of these ancient dike swarms suggest that they are associated with major areas of mantle upwelling and partial melting, with magma migrating vertically above the mantle plume to depths of a few tens of kilometers and then traveling laterally to form the longest dikes. Some of the radial surface fracture patterns associated with the novae and coronae on Venus are almost certainly similar features that have been formed more recently in that planet's geologic history, and on Mars the systems of linear graben, some of which show evidence of localized eruptive vents, extending radially from large shield volcanoes, also bear witness to the presence of long-lived mantle plumes generating giant dike swarms. It seems that there may be a great deal of similarity between the processes taking place in the mantles of all the Earth-like planets; it is the near-surface conditions, probably strongly influenced by the current presence of the oceans, that drive the plate tectonic processes distinguishing the Earth from its neighbors.


Cattermole, P. (1994). "Venus: The Geological Story." Johns Hopkins University Press, Baltimore.

Frankel, C. (1996). "Volcanoes of the Solar System." Cambridge Univ. Press, Cambridge, United Kingdom.

Lopes, R. M., and Gregg, T. K. P., eds. (2004). "Volcanic Worlds--Exploring the Solar System's Volcanoes." Springer-Verlag, Berlin, Heidelberg, New York.

Sgurdsson, H. (2000). "Encyclopaedia of Volcanoes." Academic Press, San Diego, California.
COPYRIGHT 2007 COPYRIGHT 2007 Elsevier Inc. All rights reserved.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Encyclopedia of the Solar System, 2nd ed.
Article Type:Topic overview
Date:Jan 1, 2007
Previous Article:Solar System Dynamics: Regular and Chaotic Motion.
Next Article:Astrobiology.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters