Subaqueous Volcanic Eruptions Research Articles

Reconstructing the timing and conditions of fragmentation from water speciation in ash during a deep submarine eruption: 2012 eruption of Havre, Kermadec Arc

The lack of visual observations and difficulty recreating ambient conditions in the laboratory mean deep subaqueous volcanic eruptions (>500 m water depth) remain enigmatic. However, the pressure and temperature dependency of water speciation and solubility in volcanic melts and glasses, and water's high diffusion rate, offers a potential window into the processes and conditions during deep subaqueous volcanism.The 2012 submarine eruption of Havre Volcano is the ideal laboratory in which to use water species distributions to understand submarine eruption processes. Post-eruption AUV mapping, and ROV observation and sampling generated an unprecedented amount of contextual information for a deep submarine eruption including on the magma ascent conditions, eruption sequence, and magma fragmentation. Here we present high spatial resolution synchrotron-FTIR measurements of water speciation in glassy ash (125–500 μm) from several subunits (1, 2 and 3) of a widespread ash with lapilli deposit produced in the 2012 eruption.Measurements record OH depletion profiles around vesicles, along with H2Omol enrichment profiles. In several thick vesicle walls far-field regions unaffected by processes at the bubble margins have also been identified. By focusing on OH concentrations, which is effectively non-diffusing below the glass transition, we can remove the effects of secondary rehydration and calculate the pressure and water depth at which ash grain quenched.Far-field regions record quench depths within the 650–900 m water depth range for pre- and post- eruption vent depth, suggesting ash grains equilibrated at vent depth. OH depletion profiles produce quench pressures on vesicle margins typically corresponding to 200–400 m water depth shallower, suggesting a 2–4 MPa decompression before ash quench. Diffusion modelling suggests that at eruption temperature the observed OH depletion profiles formed through exsolution in <2 s. The difference in quench pressure within the ash is not reflective of a difference in the depth of quenching. Rather it is reflective of the rapidity of the quench process relative to the diffusion rate.The results imply transient low-pressure conditions formed during ash generation in the 2012 Havre eruption. We discuss several physical processes in the context of submarine volcanism that may explain the formation of transient low-pressure regions. The results have implications for how we understand submarine eruption processes and highlight the potential of high resolution FTIR for studying submarine volcanism.

The Cretaceous Sedimentary Environments and Tectonic Setting of the Southern East China Sea Shelf Basin

Although the amount of oil and gas reserves and the reservoir properties in the East China Sea Shelf Basin (ECSSB) indicate good prospects for oil and gas exploration in the Mesozoic strata, there has not yet been a significant breakthrough in oil and gas production. An important reason is that there are few detailed studies on the sedimentary paleogeography of the ECSSB which make it difficult to predict the distribution of sand bodies. In this paper, well-core observations, thin-section identifications, electron probe microanalysis, trace elements analysis and seismic interpretations were synthetically applied to study the sedimentary environments and the tectonic setting of the Upper Cretaceous in the southern ECSSB. In the Oujiang Sag, red mudstones and pyroclastic rocks were deposited, accompanied by wedged, chaotic pro-grading reflections in the seismic profile, indicating a volcano-alluvial fan sedimentary system. Abundant sedimentary structures including double-mud drapes, asymmetric herringbone cross-beddings and burrows, such as Planolites and Skolithos, were developed in the Minjiang Sag, typically indicating a high-energy tidal environment. The autogenetic glauconites in situ and the fossil Tintinnid also indicate a shallow marine environment. The resulting sedimentary characteristics are in accordance with the littoral facies revealed by the seismic reflections in the Minjiang Sag. Moreover, volcanic conduit facies, effusive facies, and pyroclastic facies were also recognized in the seismic profile in the Minjiang Sag. Therefore, it was presumed that subaqueous volcanic eruptions occurred in the Minjiang Sag. In the Keelung Sag, the littoral facies were dominant. The tectonic setting of the Upper Cretaceous in the southern ECSSB is the transitional arc accompanied by large-scale volcanic activities. In conclusion, the volcano-alluvial fan sedimentary system was developed in the Oujiang Sag and volcano-littoral sedimentary system was developed from the Minjiang Sag to the Keelung Sag. This study is intended to promote the understanding of the sedimentary paleogeography of the Upper Cretaceous in the southern ECSSB and to provide help in oil and gas exploration.

Deep-subaqueous implosive volcanism at West Mata seamount, Tonga

Deep subaqueous volcanic eruptions (>500 m below sea level (mbsl)) remain enigmatic due to a lack of visual observations and difficulty recreating ambient conditions in the lab. Eruptive activity at West Mata seamount in May 2009 remains one of two deep subaqueous eruptions to have ever been filmed. A distinct low-intensity eruptive style, termed bubble escape activity, was observed at Hades vent (1200 mbsl) characterized by the ascent and implosion of 0.2 - 1 m diameter volatile-filled vapor bubbles (Resing et al., 2011). Video of this volcanic activity is used to constrain simple numerical models and produce the first subaqueous eruption actualistic model driven by visual observations.Bubble escape activity occurs in three stages defined by changing exsolved volatile and lava behavior. During Stage 1, vapor bubble ascent in a magma filled conduit drives either ductile or brittle deformation of the lava surface at the vent, depending on the timescale of lava cooling prior to bubble escape activity. Fragmentation of the lava during Stage 2 culminates with the vapor bubble coming into direct contact with the ambient water. At this point, Stage 3, bubbles implode through rapid condensation and contraction of the exsolved volatile phase, due to rapid heat loss from the vapor bubble to the ambient water.Numerical modeling of exsolved volatile expansion during conduit ascent to vents across the ocean depth range has identified a transition in exsolved volatile expansion characteristics at 2 - 5 MPa. This transition would produce a fundamental change in eruption processes, from which the characteristics and depth range of deep and shallow end members of bubble escape activity are defined.Bubble escape activity highlights implosive behavior driven by underpressure that develops during exsolved volatile contraction as a key, but often overlooked, component of both pyroclastic and effusive subaqueous volcanism across the entire ocean depth range. This stands in contrast to overpressure driving subaerial explosive eruptions. The fact that exsolved volatiles can expand, contract, or maintain an approximately constant volume in subaqueous volcanism also calls for the careful application of terminology (e.g. explosive) to describe subaqueous eruption processes.

Laboratory Experiments on Tsunamigenic Discrete Subaqueous Volcanic Eruptions. Part 1: Free Surface Disturbances

AbstractA submarine volcanic eruption has the potential to generate a dangerous local tsunami. To better understand the free surface disturbance generated by an underwater volcanic eruption, which will form the initial condition for any subsequent wave generation, we conducted a series of laboratory experiments. In these experiments, compressed air was injected into a tank filled with water to simulate an underwater eruption. The experiments were repeated over a range of different pressures and water depths. Each eruption can be divided into three phases: A momentum‐driven jet, a buoyancy‐driven plume, and a fountain‐generation regime. Our experiments exhibit two fountain regimes (a dome regime and a finger regime), with a transition between them. These fountain regimes have been observed in several real submarine volcanic eruptions. This paper proposes a Froude number criterion to combine the water depths and source conditions together with the aspect ratios of fountains to quantify different fountain regimes. This quantitative relationship holds for two real subaqueous volcanic eruption cases (Myojin‐Sho eruption in 1952 and 1996 eruption in Karymskoye Lake). The fountain of the Myojin‐Sho shallow submarine eruption on September 23,1952 appears to have been in the dome regime, which means it was a relatively weak eruption. Unlike other eruptions from this volcano, which did generate tsunamis, no tsunami waves were detected on September 23 . This study contributes to an enhanced understanding of the usually unseen mechanism of free surface disturbances by volcanic gas injection during submarine eruptions.

Laboratory Experiments on Tsunamigenic Discrete Subaqueous Volcanic Eruptions. Part 2: Properties of Generated Waves

AbstractSubmarine volcanic eruptions have the potential to generate tsunamis, which can cause destruction well beyond the range of the eruption itself. Here, we present a series of underwater eruption experiments in which a non‐condensing gas was injected into a water tank with a range of water depths and applied pressures. This study proposes an effective scaled water depth and categorizes underwater eruptions into three types: deep‐water eruptions, intermediate‐water eruptions, and shallow‐water eruptions. In deep‐water eruptions, most of the energy is dissipated within the water column before the plume reaches the surface, and negligible waves are generated. In intermediate‐water eruptions, reductions in water depth reduce the loss of energy to the water column, leaving more energy available for wave generation. This causes an increase in wave heights as water shallows, up to a point. In sufficiently shallow‐water cases, the water depth is so small that almost all of the energy from the eruptive jet or plume passes through the water and is dissipated into the air, so there is only small wave‐making potential, even with relatively intense source strength. Therefore, there exists a critical water depth at which an eruption with a given source intensity will generate the largest waves. That depth lies at the boundary between the intermediate‐ and shallow‐depth regimes, where the energy available for wave generation is at a maximum. This research reveals fundamental wave generation mechanisms related to underwater gas eruptions, thereby extending our understanding of submarine volcanic tsunami generation and providing a foundation for future hazard assessment.

Time constraints on Early Tonian Rifting and Cryogenian Arc terrane-continent convergence along the northern margin of the West African craton: Insights from SHRIMP and LA-ICP-MS zircon geochronology in the Pan-African Anti-Atlas belt (Morocco)

The Neoproterozoic Anti-Atlas belt (Morocco) is a key segment in tracing the history of the northern margin of the West African craton (WAC) from Rodinia breakup to Gondwana assembly. In order to constrain geodynamic events related to rifting and convergence and their stratigraphic records, a radiometric study was carried out on the volcano-sedimentary Tachdamt and Bleïda Formations. The volcaniclastic deposits of Tachdamt Fm. yield an age of ca. 883 Ma and provide a new constraint for the timing of the subaqueous volcanic eruption and the initiation of rifting in the Anti-Atlas. This age is ~100 my older than the previous age of 788 ± 10 Ma obtained from Rb/Sr isotope dating (Clauer, 1976). Available zircon data from the interbedded clastic deposits from Tachdamt Fm. indicate major shifts in provenance during the rifting with sediments sourced from the WAC and a nearby Grenvillian terrane. Detrital zircon ages from metasiltstones of Bleïda Fm. set a maximum depositional age at ca. 700 Ma. The age spectra discloses a major shift in sediment input with provenance from the WAC, Grenvillian terrane and an arc-related terrane (ca. 770–700 Ma) lying along the Anti-Atlas margin. The newly obtained results combined with existing radiometric data allowed the refinement of the stratigraphic and geotectonic framework of the pre-Pan-African strata in the WAC. The Early Tonian syn-rift volcanism has much in common with similar events in other cratons and provides a stratigraphic record for incipient Rodinia breakup. The overlying Bleïda Fm. deposited in a peripheral foreland basin chronicles the Cryogenian arrival of the ca. 770–700 Ma Bou Azzer-Siroua arc-related terrane at the Anti-Atlas margin, and correlates with initial accretion in West Gondwana. The new results indicate that the Grenvillian detrital input in the Anti-Atlas during both geodynamic stages points toward a possible Mesoproterozoic terrane existing along the western margin of the WAC.

Fine-grained volcanic-hydrothermal sedimentary rocks in Permian Lucaogou Formation, Santanghu Basin, NW China: Implications on hydrocarbon source rocks and accumulation in lacustrine rift basins

Abstract Subaqueous volcanic-hydrothermal field is a complex hybrid geological system, where sediments originated from volcanic, hydrothermal, lacustrine, terrestrial, and biotic sources are mixed. Recent petrological studies on such sediments have indicated a potential relationship between volcanic-hydrothermal activities and oil generation and accumulation. Here, we explore the mechanisms of sedimentation and oil accumulation in well-laminated organic-rich fine-grained sedimentary rocks in Permian Lucaogou Formation deposited in the lacustrine Santanghu rift basin, NW China. We conducted petrographic, carbon and oxygen isotopic, and organic geochemical analyses of fifty-four core samples from three wells (forty-eight samples from Well W1, two from W2, four from W3) designed for tight oil exploration. Four main observations and interpretations are: 1) The fine-grained sediments are a mixture of volcanic-hydrothermal detrital feldspar grains, authigenic quartz crystals, and primary lacustrine dolomite, forming interbedded tuffaceous shale and dolostone; 2) authigenic quartz crystals were precipitated from acidic hydrothermal fluids, whereas fine-grained angular alkaline feldspars were derived from subaqueous volcanic eruption; 3) the TOC content and oil generation potential increase with increased mixing of feldspars and carbonate minerals. Hence, the dolomite-bearing tuffaceous shale and tuffaceous dolostone are the best source rocks; and 4) the fine-grained sedimentary rocks are interpreted as a subaqueous volcanic-hydrothermal deposit that controls the source potential. The sedimentary processes of an ancient sediment-hosted geothermal system and its implications on hydrocarbon source rocks and accumulation were first identified. Algal bloom induced by subaqueous eruptions is the main cause for episodic deposition of organic matter and extreme heterogeneity in both lithofacies and source-rock distribution. The results suggest that cone-shaped buildups on seismic sections, which were formed by volcanic-hydrothermal deposits and contain highly-mixed feldspar and carbonate sediments, are the best hydrocarbon reservoirs, and demonstrate a strong relationship between volcanic-hydrothermal activities and hydrocarbon accumulation in an extensional tectonic setting.

In situgranulation by thermal stress during subaqueous volcanic eruptions

Some of the most complex volcanic thermodynamic processes occur when erupting magma interacts with water. In shallow water, “Surtseyan” eruptions are spectacular, and they efficiently fragment magma into fine ash particles. The aviation hazard from these eruptions depends the amount of transportable fine ash that is generated and whether it is aggregated into particle coatings or accretions. To investigate both mechanisms, we analyzed ash-encased lapilli from the Surtseyan eruptions of Capelinhos (Azores, 1957–1958) and Hunga Tonga–Hunga Ha’apai (Tonga, 2014–2015) using X-ray computed microtomography and electron microscopy. We discovered pyroclasts that were not coated, sensu stricto, but had enveloping ash produced by in situ granulation of the particle surface. This was caused by thermal stress as pyroclasts briefly traveled through water and were quenched during eruption. In situ granulation is thus an important secondary disruption process in shallow subaqueous eruptions. Our results imply that ash encasement is not always evidence of particle aggregation and accretion, but it may also result from new ash formation. Shallow-water conditions produce the most efficient ash-generation conditions, leading to the greatest hazard to downwind populations and air traffic.

Mineralogic evidence for subglacial volcanism in the Sisyphi Montes region of Mars

Here we examine the mineral assemblages detected on possible glaciovolcanic edifices in the Sisyphi Planum region of Mars, a high-latitude region in the southern highlands nestled between the Argyre and Hellas impact basins. Minerals were identified utilizing visible/near-infrared spectra from the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM). Analysis of eleven CRISM images located on the volcanic edifices revealed three distinct spectral classes in the region which are interpreted to be: gypsum-dominated, smectite–zeolite–iron oxide-dominated (possibly palagonite), and polyhydrated sulfate-dominated material. While sulfates can form under a variety of alteration conditions, palagonite-like mineral assemblages require low-temperature and high water-to-rock hydrothermal conditions typically found in subglacial or subaqueous volcanic eruptions. The possible palagonite detections on the volcanic edifices, the geomorphology of the region, and the analogous terrestrial mineralogy of subglacial eruptions strongly suggests the formation of these minerals during subglacial eruptions or associated hydrothermal systems. This implies that thick water ice sheets were present in this region in the late Noachian or early Hesperian, and that the subglacial hydrothermal systems could have supported habitable environments with excellent biosignature preservation potential.

Particle transport in subaqueous eruptions: An experimental investigation

Subaqueous volcanic eruptions are natural events common under the world's oceans. Here we report results from bench-scale underwater explosions that entrain and eject particles into a water tank. Our aim was to examine how particles are transferred to the water column and begin to sediment from it, and to visualize and interpret evolution of the ‘eruption’ cloud. Understanding particle transfer to water is a key requirement for using deposit characteristics to infer behaviour and evolution of an underwater eruption. For the experiments here, we used compressed argon to force different types of particles, under known driving pressures, into water within a container, and recorded the results at 1MPx/frame and 1000fps. Three types of runs were completed: (1) particles within water were driven into a water-filled container; (2) dry particles were driven into water; (3) dry particles were driven into air at atmospheric pressure. Across the range of particles used for all subaqueous runs, we observed: a) initial doming, b) a main expansion of decompressing gas, and c) a phase of necking, when a forced plume separated from the driving jet. Phase c did not take place for the subaerial runs. A key observation is that none of the subaqueous explosions produced a single, simple, open cavity; in all cases, multiphase mixtures of gas bubbles, particles and water were formed. Explosions in which the expanding argon ejects particles in air, analogous to delivery of particles created in an explosion, produce jets and forced plumes that release particles into the tank more readily than do those in which particles in water are driven into the tank. The latter runs mimic propulsion of an existing vent slurry by an explosion. Explosions with different particle types also yielded differences in behaviour controlled primarily by particle mass, particle density, and particle-population homogeneity. Particles were quickly delivered into the water column during plume rise following necking, with minor transfer along initial-jet margins, and for breaching explosions additional delivery from splashdown of tephra jets. Plume rise after necking also draws upward and re-entrains some groups of particles. Most delivered particles participate in initiating vertical sediment-gravity flows, some of which reached the tank floor and began lateral flow within the short duration of our experiments. Particles transferred from plume margins locally were sufficiently well-separated to settle independently from suspension.

Progress on Sedimentation of Subaqueous Volcanic Eruption

Progress on Sedimentation of Subaqueous Volcanic Eruption

Geology, tectonic settings and iron ore metallogenesis associated with submarine volcanism in China: An overview

Submarine volcanogenic iron oxide (SVIO) deposits are one of the most important sources of high-grade iron ores in China. The spatial distribution of the deposits is controlled by the tectonic settings including arc, back-arc and rift environments, with the SVIO deposits mostly concentrated in the western part of China namely, the southwestern Yangtze Craton, Western and Eastern Tianshan, and Altay orogens and the Kaladawan iron ore district in the eastern part of the Altyn Tagh region. The Chinese SVIO deposits range in age from Paleoproterozoic to Mesozoic, and were formed during two main metallogenic epochs in the Proterozoic and Paleozoic. More than 70% of the SVIO deposits formed in the Paleozoic, with three important SVIO-metallogenic provinces recognized, in the Altay, Eastern and Western Tianshan orogens. These SVIO deposits are hosted in lithofacies that are related to submarine magmatism, such as lavas and associated pyroclastic and volcaniclastic-sedimentary rocks. The iron orebodies are hosted in different volcanic lithofacies with different features. Moreover, the different volcanic lithofacies in which the Fe ores are hosted also provide information as to their spatial relationship, ranging from distal to proximal to the eruption center or vent. Many of these deposits are characterized by well developed skarns, and could be interpreted either by a distal position of the ore system in question and/or exposed igneous rocks or active magma chamber, or a relationship to early metamorphism and continuous alteration at relatively high temperature followed by retrograde alteration as temperatures decline. Geological and geochemical evidence suggests that these deposits were formed as a result of submarine magmatic activity, including subaqueous volcanic eruptions, associated volcano-sedimentary lithofacies, and related post-magmatic hydrothermal activity. Iron oxide ore probably formed the hydrothermal fluids which generated the skarns could be a mixture of evolved magma-derived water and convecting sea water driven by the heat from the shallow active magma chamber, whereas volcano-sedimentary deposits could be formed by the fallout of the ore-bearing materials to the sea floor emanating from submarine eruption columns, or fractional precipitation of iron which had been introduced locally into the bottom water by volcanic-origin hydrothermal solutions and by leaching from the relatively iron-rich volcanic rocks. The formation of these various styles of Fe ore deposits is controlled by several key factors, such as magma differentiation, lithofacies of host rocks, temperature and chemical compositions of hydrothermal fluids, as well as the depth of sea water. In combination with their geological characteristics, geodynamic mechanisms and metallogenesis, we propose a genetic model in which the origin of these deposits can be related to the space–time evolution of the submarine volcanism, and their relationship to volcanic lithofacies variation, such as central, proximal and distal environments of ore formation.

Sediment Sorting in the Deposits of Turbidity Currents Created by Experimental Modeling of Explosive Subaqueous Eruptions

Abstract Deposits of turbidity currents induced by subaqueous volcanic eruptions are increasingly recognized in settings ranging from lakes to the deep sea. These “eruption‐fed density currents” have high current temperatures and low‐density particles, which affect current dynamics and deposit features. Field data indicate that deposit thickness and grain size decrease downflow at a changing rate. This trend is similar for deposits of noneruptive turbidity currents, but with volcanic activity, the deposits often exhibit a break in slope of the profile of deposit thickness and grain size distribution. Here we present laboratory results on density currents that address the effects of variations in grain density and water temperature on flow properties and depositional processes. Parallel runs were performed under identical flow conditions using quartz sand and pyroclastic sediment. Each grain type was used in one series of runs in which a hot density current enters a cold environment and one series in which...

Extensive volcaniclastic deposits at the Mid‐Atlantic Ridge axis: results of deep‐water basaltic explosive volcanic activity?

A recent mapping survey followed by submersible diving investigations revealed the existence of extensive volcaniclastic deposits on three adjoining, progressively deeper, segments of the Mid‐Atlantic Ridge, south‐west of the Azores. The maximum water depth at which subaqueous magmatic and hydro‐magmatic explosive processes can occur is controversial. This discovery provides fundamental constraints on the dynamics of deep subaqueous volcanic eruptions. Using multibeam bathymetry, sonar image interpretations, submersible observations, volcaniclastic sample descriptions and chemical analysis, evidence of extensive deep water (up to –1700 m) magmatic explosive activity of basalt at the axis of a spreading Mid‐Ocean Ridge is documented for the first time.

Subaqueous, intermediate to silicic-composition explosive volcanism: a review

Abstract Mass flow deposits of pyroclastic debris commonly consist of two parts, a lower coarse-grained basal section and an upper finer-grained bedded section. Subaqueous pyroclastic flow deposits, particularly those derived from a vertically-directed eruption, appear better-sorted than their subaerial counterparts because of the efficient sorting characteristics of water compared to air. Subaqueous deposits frequently are depleted in bubble-wall pumice that floats off and vitric ash that remains in suspension. Hence, the deposits show a relative enrichment of crystals, especially at the bases, from (1) flotation and suspension of the vitric material and (2) ingestion of water at the head of the flow. A subaqueous volcaniclastic flow deposit may contain evidence indicative of both turbulent and non-turbulent movement. The flow may have had a turbulent head and non-turbulent body, or flow transformations may have occurred from (1) collapse of the eruption column; (2) changes in slope and in the velocity of the flow; and/or (3) the amount of water mixed into the flow. Retention of heat in the subaqueous environment may be enhanced by lateral eruptions, high eruption rates, large eruptive volumes, and little or no mixing of pyroclastic debris with water. The lack of welded subaqueous deposits in the geologic record may be a result of: (1) effusive rather than explosive eruptions in a deep-water environment; (2) explosive phreatic disintegration of pyroclastic flows in a shallow subaqueous environment; and/or (3) the inability of subaerial pumiceous pyroclastic flows to penetrate the land/water interface. Characteristics of explosive, intermediate to silicic-composition, subaqueous volcanic eruptions also have been summarized in terms of: (1) environment of the eruption; (2) orientation of the eruption; and (3) composition of the juvenile debris. These three aspects of explosive subaqueous volcanism are ideal end-members, and gradations no doubt occur.

Origin and evolution of the layered deposits in the Valles Marineris, Mars

Thick sequences of layered deposits are found in the Martian Valles Marineris. They exhibit fine, nearly horizontal layering, and are present as isolated plateaus of what may have once been more extensive deposits. Individual sequences of layered deposits are as thick as 5 km. The greatest total thicknesses of deposits are found in Candor, Ophir, and Melas chasmata. individual layer thicknesses range from about 70 to 300 m. Some tilting of sequences is observed, but at the best image resolutions, no angular unconformities are detectable in the layers. The sequences of events in the canyons, as deduced from morphologic and stratigraphic evidence, was (1) graben formation in response to the tharsis uplift, (2) canyon wall retreat and canyon enlargement, roughly contemporaneous with formation of the layered deposits, (3) deep erosion of the layered deposits, (4) landsliding of the canyon walls, and (5) eolian erosion of the layered deposits, perhaps continuing up to the present. We consider four hypotheses for the origin of the layered deposits: they are eolian deposits, they are remnants of the material that makes up the canyon walls, they are deposits of explosive volcanic eruptions, or they were deposited in standing bodies of water. The rhythmic nature of the layers and their lateral continuity, horizontality, great thickness, and stratigraphic relationships with other units in the canyons all appear most consistent with deposition in an aqueous environment. If standing bodies of water existed in the Valles Marineris, they were almost certainly ice-covered. there are three ways in which sediment could have entered an ice-covered lake: down through the ice cover, up from the lake bottom, or in from the lake margins. Layers of sediment could have been transported downward through an ice cover by foundering or Rayleigh-Taylor instabilities, but it is not clear whether there was a viable mechanism for repeatedly accumulating thick sediment layers on top of the ice cover. Subaqueous volcanic eruption on the lake bottom does not suffer from many of the morphologic arguments that make origin by subaerial volcanism seem improbable. While this mechanism is attractive, there are no eruptive centers observed and there is no other direct evidence to support it. Because canyon enlargement took place at roughly the same time as layer deposition, debris from the canyon walls is an obvious and likely source for some of the material in the layered deposits; however, the volume of material removed from the canyon walls may be insufficient to account for all of the presently observed material. We conclude that there are several geologically feasible, but as yet unproven, mechanisms that could have led to formation of thick deposits in ice-covered paleolakes in the Valles Marineris. Present data are insufficient to choose conclusively among the various possibilities. Several types of data from the Mars Observer mission will be useful in further characterizing the deposits and clarifying the process of their origin. The deposits should be considered important targets for a future Mars sample return mission.

Some remarks on an error respecting the site and origin of Graham’s Island

The author rectifies an erroneous assertion, originating from the report of Captain Larmour, who in the year 1800, when commanding the Wassanaer, a troop-ship on the Egyptian expedition, thought he observed a shoal of four fathoms water with breakers, within a mile of the latitude and longitude of the new volcanic island. The author has determined, by his own observations, that no such reef exists in that spot, nor is the assigned place of this shoal near that of Graham’s Island, which arose considerably to the eastward, from a depth of above a hundred fathoms below the surface of the water. A knoll, with only seven fathoms of water upon it, was discovered not far from the site of these reports. The Adventure bank extends from Sicily nearly to Pentellaria, where the water deepens at once from 763 fathoms to above 375 fathoms, at which no bottom was met with. But, even on the supposition that what Capt. Larmour imagined he saw was the result of a temporary subaqueous volcanic eruption, it could not have justified the assertion of there being breakers with four fathoms upon them; and still less does it afford any foundation for the hypothesis that Graham’s Island was formed by the mere lifting up of such shoal.