Caldera formation on Santorini and the physiography of the islands in the late Bronze Age

The caldera of Santorini is a composite structure with a subsidence history extending over 100 ka or more. Geomorphological mapping shows that the present-day caldera wall is a complex assemblage of cliff surfaces of different ages, and that collapse at Santorini has repeatedly exhumed earlier caldera cliffs and unconformities. Cliffs bounding the southern, southeastern and northwestern rims of the caldera are morphologically fresh and probably formed during or soon after the Minoan eruption in the late Bronze Age. The well-scalloped shape of these cliffs is attributed to large-scale rotational landslip around the margins of the Minoan caldera. The deposit from one landslip is preserved subaerially. Minoan landslips in southeast santorini detached along the basement unconformity, exposing a cliff of the prevolcanic island. The caldera wall in the north, northeast and east preserves evidence for three generations of cliff: those of Minoan age and two earlier generations of caldera wall. The two early calderas can be dated relative to a well-established statigraphy of lavas and tuffs. The presence of in situ Minoan tephra plastered onto the present-day caldera wall provides evidence that these ancient caldera cliffs had already been exhumed prior to the Minoan eruption. Field relationships permit reconstruction of the physiography of Bronze-Age Santorini immediately before the Minoan eruption. The reconstruction differs from some previously published versions and is believed to be the most accurate to date. Bronze-Age Sa ntorini had a large flooded caldera formed 21 ka ago. This caldera must have acted as an excellent harbour for the Bronze-Age inhabitants of the island. The 3.6 ka Minoan eruption deepened and widened the extant caldera. The volume of Minoan collapse (25 km³) is in good agreement with published estimates for the volume of discharged magma if between 5 and 8 km³ of Minoan ignimbrite ponded as intracaldera tuff.     

Analogue study of particle segregation in pyroclastic density currents, with implications for the emplacement mechanisms of large ignimbrites

Abstract Analogue flume experiments were conducted to investigate the transport and sedimentation behaviour of turbulent pyroclastic density currents. The experimental currents were scaled approximately to the natural environment in three ways: (1) they were fully turbulent; (2) they had a very wide range of particle sizes and associated Rouse numbers (the ratio of particle settling velocity to effective turbulent eddy velocity in the current); and (3) they contained particles of two different densities. Two sets of surge‐type experiments were conducted in a 5 m long, water‐filled lock‐exchange flume at five different volumetric particle concentrations from 0·6% to 23%. In one set (one‐component experiments), the currents contained just dense particles; in the other set (two‐component experiments), they contained both light and dense particles in equal volume proportions. In both sets of experiments, the population of each component had a log‐normal size distribution. In the two‐component experiments, the size range of the light particle population was selected in order to be in hydrodynamic equivalence with that of the dense particles. Dense particles were normally graded, both vertically and downstream, in the deposits from both sets of experiments. The mass loading (normalized to the initial mass of the suspension) and grain size of the dense component in the deposits decreased with distance from the reservoir and were insensitive to initial total particle concentration in the currents. On the other hand, in the two‐component experiments, the light particles were extremely sensitive to concentration. They were deposited in hydrodynamic equivalence with the dense particles from dilute currents, but were segregated efficiently at concentrations higher than a few per cent. With increasing particle concentration, the large, light particles were carried progressively further down the flume because of buoyancy effects. Deposits from the high‐concentration currents exhibited reverse vertical grading of the large, light particles. Efficient segregation of the light component was observed even if the bulk density of the current was less than that of the light particles. In both sets of experiments, marked inflexions in the rate of downstream decline in mass loading and maximum grain size of the dense component can be attributed to the presence of two different particle settling regimes in the flow: (1) particles with Rouse numbers >2·5, which did not respond to the turbulence and settled rapidly; and (2) particles with Rouse numbers <2·5, which followed the turbulent eddies and settled slowly. The results are applied to the transport and sedimentation dynamics of pyroclastic density currents that generate large, widespread ignimbrites. Field data fail to reveal significant departures from aerodynamic equivalence between pumice and lithic clasts in three such ignimbrites: the particulate loads of some large ignimbrites are transported principally in turbulent suspensions of low concentration. In some ignimbrites, the well‐developed inflexions in curves of maximum lithic (ML) size vs. distance can be attributed to the existence of distinct high and low Rouse number particle settling regimes that mark the transition from an overcharged state to one in which the residual particulate load is transported more effectively by turbulence.     

Characterisation of the 1997 Vulcanian explosions of Soufrière Hills Volcano,Montserrat, by video analysis

The activity of Convention at Montserrat Soufrière Hills Volcano, Montserrat, during the period 1995–1999 included numerous violent explosions. Two major cycles of Vulcanian explosions occurred in 1997: a first of 13 explosions between 4 and 12 August and a second of 75 between 22 September and 21 October. The explosions were short-lived events lasting a few tens of seconds during which partial fountain collapse generated pyroclastic surges and pyroclastic flows, and buoyant plumes ascended 3–15 km into the atmosphere. Each explosion discharged on average 3×10⁵ m³ (dense-rock equivalent, DRE) of magma, draining the conduit to depths of 1–2 km. The paper focuses on the first few seconds of three explosions of the 75 that occurred in September/October 1997: 6 October 1997 at 17:50, 7 October 1997 at 16:02 and 9 October 1997 at 12:32. Physical parameters such as exit velocities, magmatic water contents and magma pressures at fragmentation are estimated by following and modelling the ascent of individual momentum-dominated finger jets visible on videos during the initial stages of each explosion. The model treats each finger jet as an incompressible flow sustained by a steady flux of gas and particles during the few seconds of ascent, and produces results that compare favourably with those using a multiphase compressible code run using similar eruptive parameters. Each explosion reveals a progressive increase in eruptive intensity with time, jet exit velocities increasing from 40 m s^–1 at the beginning of the explosion up to 140 m s^–1 after a few seconds. Modelling suggests that the first magma to exit was largely degassed, whereas that discharged after a few seconds contained up to 2 wt% water. Magma overpressures up to ~10 MPa are estimated to have existed in the conduit immediately prior to each explosion. Progressive increases in jet exit velocity with time over the first few seconds of each explosion provide direct evidence for strong pre-eruptive gradients in water content and magma pressure in the upper reaches (probably 100–500 m) of the conduit. Fountain collapse occurred during the first 10–20 s of each explosion because the discharging jets had bulk densities up to 100 times that of the atmosphere and were unable to entrain enough air to become buoyant. Such high eruptive densities were due to the presence of partially degassed magma in the conduit.Editorial responsibility: A. Woods     

Particle fabric in a small, type-2 ignimbrite flow unit (Laacher See, Germany) and implications for emplacement dynamics

 Ignimbrites of the 13-ka Upper Laacher See Tephra were deposited from small, highly concentrated, moderately fluidized pyroclastic flows. Their unconsolidated nature, and the prominence of accidental Devonian slate fragments, make these ignimbrites ideal for clast fabric studies. The upper flow unit of ignimbrite M14 has characteristics typical of a type-2 ignimbrite. Layer 2a and the lower part of layer 2b of the flow unit have strong, upstream-inclined a[p] fabrics (a[p] means long particle axes parallel to flow direction). Only clasts with a/b axial ratios of 2.5 or greater preserve good a[p] fabrics, whereas the a–b planes of flat fragments dip upstream irrespective of axial ratio. The a-axis fabric becomes weaker, flatter, and more girdle-like in the upper half of layer 2b. At one locality the a-axis fabric appears to rotate 40° up through the flow unit, suggesting either shear decoupling of different levels in the moving flow or unsteadiness effects in a flow depositing progressively at its base. The existence of similarly strong a[p] fabrics in layer 2a and the lower half of layer 2b appears inconsistent with the common interpretation that ignimbrite flow units are emplaced as a plug of essentially non-shearing material (layer 2b) on a thin shear layer (layer 2a), and that the entire flow freezes en masse to form the deposit. The data suggest that, if the flow froze en masse, it was shearing pervasively through at least half its thickness. Another possibility is that the flow unit aggraded progressively from the base up, and that the fabrics record the integrated history of shear directions and intensities immediately above the bed throughout the duration of deposition. Received: 13 February 1997 / Accepted: 4 April 1998     

Experimental study of dense pyroclastic density currents using sustained,gas-fluidized granular flows

Submarine lava flow morphology is commonly used to estimate relative flow velocity, but the effects of crystallinity and viscosity are rarely considered. We use digital petrography and quantitative textural analysis techniques to determine the crystallinity of submarine basaltic lava flows, using a set of samples from previously mapped lava flow fields at the hotspot-affected Galápagos Spreading Center. Crystallinity measurements were incorporated into predictive models of suspension rheology to characterize lava flow consistency and rheology. Petrologic data were integrated to estimate bulk lava viscosity. We compared the crystallinity and viscosity of each sample with its flow morphology to determine their respective roles in submarine lava emplacement dynamics. We find no correlation between crystallinity, bulk viscosity, and lava morphology, implying that flow advance rate is the primary control on submarine lava morphology. However, we show systematic variations in crystal size and shape distribution among pillows, lobates, and sheets, suggesting that these parameters are important indicators of eruption processes. Finally, we compared the characteristics of lavas from two different sampling sites with contrasting long-term magma supply rates. Differences between lavas from each study site illustrate the significant effect of magma supply on the physical properties of the oceanic upper crust.     

Petrology of the zoned calcalkaline magma chamber of Mount Mazama,Crater Lake,Oregon

Evolution of the magma chamber at Mount Mazama involved repeated recharge by two types of andesite (high-Sr and low-Sr), crystal fractionation, crystal accumulation, assimilation, and magma mixing (Bacon and Druitt 1988). This paper addresses the modal compositions, textures, mineral chemistry and magmatic temperatures of (i) products of the 6845±50 BP climactic eruption, (ii) blocks of partially fused granitoid wallrock found in the ejecta, and (iii) preclimactic rhyodacitic lavas leaked from the chamber in late Pleistocene and early Holocene time. Immediately prior to the climactic eruption the chamber contained ≳ 40 km³ of rhyodacite (10 vol% plag + opx + aug + hb + mt + ilm, ∼880° C) overlying high-Sr andesite and cumulus-crystal mush (28–51 vol% plag + hb ± opx ± aug + mt ± ilm, 880° to ≥950° C), which in turn overlay low-Sr crystal mush (50–66 vol% plag + opx + aug ± hb ± ol + mt + ilm, 890° to ≥950† C). Despite the well known compositional gap in the ejecta, no thermal discontinuity existed in the chamber. Pre-eruptive water contents of pore liquids in most high-Sr and low-Sr mushes were 4–6 wt%, but on average the high-Sr mushes were slightly richer in water. Although parental magmas of the crystal mushes were andesitic, xenocrysts of bytownite and Ni-rich magnesian olivine in some scoriae record the one-time injection of basalt into the chamber. Textures in ol-bearing scoriae preserve evidence for the reactions ol + liq = opx and ol + aug + liq(+ plag?) = hb, which occurred in andesitic liquids at Mount Mazama. Strontium abundances in plagioclase phenocrysts constrain the petrogenesis of preclimactic and climactic rhyodacites. Phenocryst cores derived from high-Sr and low-Sr magmas have different Sr contents which can be resolved by microprobe. Partition coefficients for plagioclase in andesitic to rhyolitic glasses range from 2 to 7, and increase as glass %SiO₂ increases. Evolved Pleistocene rhyodacites (∼30–25,000 BP) and rhyodacites of the Holocene Llao Rock center (7015±45 BP) contain Sr-poor plagioclase and are derivatives from low-Sr magma. Rhyodacites of the Pleistocene Sharp Peak domes, Holocene Cleetwood flow (∼6850 BP), and climactic ejecta contain discrete Sr-rich and Sr-poor plagioclase phenocryst populations and are hybrids produced by mixing low-Sr rhyodacite (containing Sr-poor plag + opx + aug) with a more mafic high-Sr magma (with Sr-rich plag [ + hb?]). The data reinforce the conclusions of crystal-liquid mixing calculations (Bacon and Druitt 1988), and suggest some important refinements to the magma chamber model.     

Emplacement of the 18 May 1980 lateral blast deposit ENE of Mount St. Helens,Washington

Facies variations east-northeast of Mount St. Helens preserve a record of depositional processes in the 18 May 1980 lateral blast cloud. This paper reports new field, grain-size and component data from the ENE sector of the timber-blowdown zone and presents a model for blast flow and sedimentation. The first-erupted ejecta was rich in juvenile components and extends to the distal blowdown limit. The last-erupted ejecta was rich in accidental lithics and reached no further than a few kilometres from the mountain due to waning discharge. The blast cloud was a turbulent stratified flow which transported and deposited sediment in the manner of a high-density turbidity current. The possibility that the blast was emplaced as a giant shearing fluidised bed is not favoured by compositional zoning patterns. Depositional conditions were strongly influenced by the rate of suspended-load fallout from the blast. Within about 8 km from vent rapid sedimentation caused deposition under moderate to high concentration conditions and formation of a basal hindered-settling zone able to detach gravitationally and drain into local depressions. The resulting proximal facies resembles a low-aspect-ratio ignimbrite. Fines depletion in the proximal facies is attributed to a combination of residual turbulence and rapid gas escape during particle settling and compaction through the hindered-settling zone. Component data suggest that the blast head played no significant role in the generation of fines depletion in the blast deposit as suggested by previous workers. With increasing distance from vent the rate of particle fallout declined and sedimentation took place under increasingly dilute and tractional conditions, building up antidune-like bedforms. Wavelengths of these bedforms range from 20 to <1 m, and decrease away from vent. There is a systematic relationship between antidune migration direction and depositional slope. The transition from proximal (ignimbrite-like) to distal (surge-like) facies suggests a possible gradation in transport and deposition processes between conventional pyroclastic surges and high-velocity pyroclastic flows.     

Pre-explosive conduit conditions of the 1997 Vulcanian explosions at Soufrière Hills Volcano,Montserrat: II. Overpressure and depth distributions

A type example of Vulcanian eruptive dynamics is the series of 88 explosions that occurred between August and October 1997 at Soufrière Hills volcano on Montserrat Island. These explosions are interpreted to be caused by the pressurization of a conduit by a shallow highly crystalline and degassed magma plug. We test such an interpretation by combining the pressures and porosities of the pre-explosive magma column proposed by Burgisser et al. (2010, doi:10.1016/j.jvolgeores.2010.04.008) into a physical model that reconstructs a depth-referenced density profile of the column for four mechanisms of pressure buildup. Each mechanism yields a different overpressure profile: 1) gas accumulation, 2) conduit wall elasticity, 3) microlite crystallization, and 4) magma flowage. For the first three mechanisms, the three-part vertical layering of the conduit prior to explosion was spatially distributed as a dense cap atop the conduit with a thickness of a few tens of meters, a transition zone of 400–700 m with heterogeneous vesicularities, and, at greater depth, a more homogeneous, low-porosity zone that brings the total column length to ~ 3.5 km. A shorter column can be obtained with mechanism 4: a dense cap of less than a few meters, a heterogeneous zone of 200–500 m, and a total column length as low as 2.5 km. Inflation/deflation cycles linked to a periodic overpressure source offer a dataset that we use to constrain the four overpressure mechanisms. Magma flowage is sufficient to cause periodic edifice deformation through semi-rigid conduit walls and build overpressures able to trigger explosions. Gas accumulation below a shallow plug is also able to build such overpressures and can occur regardless of magma flowage. The concurrence of these three mechanisms offers the highest likelihood of building overpressures leading to the 1997 explosion series. We also explore the consequences of sudden (eruptive) overpressure release on our magmatic columns to assess the role of syn-explosive vesiculation and pre-fragmentation column expansion. We find that large shallow overpressures and efficient syn-explosive vesiculation cause the most dramatic pre-fragmentation expansion. This leads us to depict two end-member pictures of a Vulcanian explosion. The first case corresponds to the widely accepted view that the downward motion of a fragmentation front controls column evacuation. In the second case, syn-explosive column expansion just after overpressure release brings foamed-up magma up towards an essentially stationary and shallow fragmentation front.     

Evolution of the crustal magma plumbing system during the build-up to the 22-ka caldera-forming eruption of Santorini (Greece)

The formation of shallow, caldera-sized reservoirs of crystal-poor silicic magma requires the generation of large volumes of silicic melt, followed by the segregation of that melt and its accumulation in the upper crust. The 21.8?±?0.4-ka Cape Riva eruption of Santorini discharged >10 km³ of crystal-poor dacitic magma, along with <<1 km³ of hybrid andesite, and collapsed a pre-existing lava shield. We have carried out a field, petrological, chemical, and high-resolution ⁴⁰Ar/³⁹Ar chronological study of a sequence of lavas discharged prior to the Cape Riva eruption to constrain the crustal residence time of the Cape Riva magma reservoir. The lavas were erupted between 39 and 25 ka, forming a ～2-km³ complex of dacitic flows, coulées and domes up to 200 m thick (Therasia dome complex). The Therasia dacites show little chemical variation with time, suggesting derivation from one or more thermally buffered reservoirs. Minor pyroclastic layers occur intercalated within the lava succession, particularly near the top. A prominent pumice fall deposit correlates with the 26-ka Y-4 ash layer found in deep-sea sediments SE of Santorini. One of the last Therasia lavas to be discharged was a hybrid andesite formed by the mixing of dacite and basalt. The Cape Riva eruption occurred no more than 2,800?±?1,400 years after the final Therasia activity. The Cape Riva dacite is similar in major element composition to the Therasia dacites, but is poorer in K and most incompatible trace elements (e.g. Rb, Zr and LREE). The same chemical differences are observed between the Cape Riva and Therasia hybrid andesites, and between the calculated basaltic mixing end-members of each series. The Therasia and Cape Riva dacites are distinct silicic magma batches and are not related by shallow processes of crystal fractionation or assimilation. The Therasia lavas were therefore not simply precursory leaks from the growing Cape Riva magma reservoir. The change 21.8 ky ago from a magma series richer in incompatible elements to one poorer in those elements is one step in the well documented decrease with time of incompatibles in Santorini magmas over the last 530 ky. The two dacitic magma batches are interpreted to have been emplaced sequentially into the upper crust beneath the summit of the volcano, the first (Therasia) then being partially, or wholly, flushed out by the arrival of the second (Cape Riva). This constrains the upper-crustal residence time of the Cape Riva reservoir to less than 2,800?±?1,400 years, and the associated time-averaged magma accumulation rate to >0.004 km³ year^-1. Rapid ascent and accumulation of the Cape Riva dacite may have been caused by an increased flux of mantle-derived basalt into the crust, explaining the occurrence of hybrid andesites (formed by the mixing of olivine basalt and dacite in approximately equal proportions) in the Cape Riva and late Therasia products. Pressurisation of the upper crustal plumbing system by sustained, high-flux injection of dacite and basalt may have triggered the transition from prolonged, largely effusive activity to explosive eruption and caldera collapse.