Further paleomagnetic results for the San Juan volcanic field of southern Colorado

Combining paleomagnetic data for 17 new sites from the northwest portion of the (Oligocene) San Juan volcanic field of southern Colorado with data for 29 sites previously published yields a paleomagnetic pole at 85°N, 114°E (with a 95% confidence circle of 7.5° radius). A further combination of the San Juan data with the results of other studies on rocks of Oligocene age from tectonically stable parts of North America gives a mid-Tertiary reference pole located at 81°N, 132.5°E, with a confidence circle of approximately 4°. Mid-Tertiary paleomagnetic poles for the western edge of the continent diverge markedly from this reference pole.     

Flow and deposition of pyroclastic granular flows: A type example from the 1975 Ngauruhoe eruption,New Zealand

Small-volume pyroclastic density currents (PDCs) are generated frequently during explosive eruptions with little warning. Assessing their hazard requires a physical understanding of their transport and sedimentation processes which is best achieved by the testing of experimental and numerical models of geophysical mass flows against natural flows and/or deposits. To this end we report on one of the most detailed sedimentological studies ever carried out on a series of pristine small-volume PDC deposits from the 1975 eruption of Ngauruhoe volcano, whose emplacement were also witnessed during eruption. Using high-resolution GPS surveys, a series of lateral excavations across the deposits, and bulk sedimentological analysis we constrained the geomorphology, internal structure and texture of the deposits with respect to laterally varying modes of deposition.     

Permeability and pore-connectivity variation of pumices from a single pyroclastic flow eruption: Implications for partial fragmentation

The relationship between permeability and vesicularity in volcanic rocks has been used to infer the degassing behavior of hydrous magma. Recent data on natural samples from various eruptions show a wide variation, fitting a power–law relationship of the percolation models with low (< 30%) critical vesicularity (Ф_C). In this study, we present data on permeability and pore-connectivity of juvenile rhyolitic pumice clasts in a pyroclastic flow around Onikobe volcano, NE Japan, and investigate their relationship with vesicularity developed in a single eruption event. The permeability of the pumices having a relatively low abundance of microlites and microphenocrysts shows a trend increasing by 4 orders of magnitude (from 10^− 13.8 to 10^− 10.1 m²) in a high and narrow vesicularity range (from 72 to 80%). This trend intersects at a high angle with the fit to the permeability–vesicularity data in the previous studies that has a low Ф_C, and is located on the extension of the trend for the products of isotropic decompression experiments. The two-dimensional (2D) connectivities of pores for the pumices were also measured from thin sections. From the point of view of percolation theory, connectivity provides information about the probability of percolation. They showed a steep increase from ca. 0 to 0.7 in an almost similar vesicularity range, as compared to their permeabilities. We attribute the increase in 2D connectivity to the increasing amount of ruptured bubble walls, which might have provided less-tortuous paths through larger apertures for gas flow. This, in turn, would cause an effective increase in the permeability. Aggregates of bubble-wall-shaped glass shards were found in the pumices, and their amount and degree of welding are higher in the pumices that have a higher abundance of microlites and microphenocrysts. These pumices have relatively high permeability and 2D connectivity at low vesicularity, which is accounted for by the existence of large irregularly shaped pores. These textural characteristics suggest that a series of partial fragmentation processes, including local rupturing of bubble walls and subsequent foam-collapse with permeable gas flow, might have occurred before the ultimate bulk fragmentation, thus resulting in the increase in permeability. We suggest that the 2D connectivity of pores is a useful parameter to quantify the degree of fragmentation of bubble walls and has the potential for use to assess their permeability.     

Characterization of pyroclastic fall and flow deposits from the 1815 eruption of Tambora volcano,Indonesia using ground-penetrating radar

Ground-penetrating radar (GPR) is used to image and characterize fall and pyroclastic flow deposits from the 1815 eruption of Tambora volcano in Indonesia. Analysis of GPR common-mid-point (CMP) data indicate that the velocity of radar in the sub-surface is 0.1 m/ns, and this is used to establish a preliminary traveltime to-depth conversion for common-offset reflection profiles. Common-offset radar profiles were collected along the edge of an erosional gully that exposed approximately 1–2 m of volcanic stratigraphy. Additional trenching at select locations in the gully exposed the contact between the pre-1815 eruption surface and overlying pyroclastic deposit from the 1815 eruption. The deepest continuous, prominent reflection is shown to correspond to the interface between pre-eruption clay-rich soil and pyroclastics that reach a maximum thickness of 4 m along our profiles. This soil surface is distinctly terraced and is interpreted as the ground surface augmented for agriculture and buildings by people from the kingdom of Tambora. The correlation of volcanic stratigraphy and radar data at this location indicates that reflections are produced by the soil-pyroclastic deposit interface and the interface between pyroclastic flows (including pyroclastic surge) and the pumice-rich fall deposits. In the thickest deposits an additional reflection marks the interface between two pyroclastic flow units.     

Miocene rhyolitic welded tuff infilling a funnel-shaped eruption conduit Shiotani, southeast of Matsue, SW Japan

 At Shiotani, SW Japan, rhyolitic welded tuff forms a steep-sided funnel-shaped body, confined by Paleogene granitic rocks to an elliptical area 1–1.5 km across. The Shiotani welded tuff is pervasively welded and foliated concordantly with the contact that dips inward at angles of 70–90°. In contrast, nearby contemporary volcaniclastic deposits are non-welded and gently inclined. Near the contact with the granite, the tuff is plastically deformed and shows lineations that plunge inward at angles of 40–65°. Lithic and crystal clasts in the rheomorphic outer part are rotated in a plane normal to the foliations and parallel to the lineations indicating downward flow of the welded tuff. The geometry and internal structures suggest that the Shiotani welded tuff was emplaced and welded in a funnel-shaped eruption conduit. Upon collapse of a plinian or phreatoplinian eruption column, the majority of the conduit-filling pyroclasts probably fell back en masse into the conduit. Heat and steam from underlying magma and diffusion of interstitial volatiles into the glass perhaps reduced the viscosity of juvenile pyroclasts and facilitated welding in the conduit, especially at deep levels. The hot welded pyroclasts then flowed down the conduit wall during welding compaction and retreat of the magma. These processes resulted in increased welding toward the contacts and welding foliations concordant with the steep wall. Emplacement of nearby correlative volcaniclastic mass-flow deposits in a shelf to upper bathyal environment suggests a possibility that, when active, the Shiotani conduit was under the sea. Welding compaction would occur even under the sea provided that the steam generated in the upper part of the conduit fill prevented water access. Received: 28 February 1996 / Accepted: 5 May 1997     

Hydrothermal alteration of green tuff belt, Japan: Implications for the influence of seawater/volcanic rock interaction on the seawater chemistry at a back-arc basin

Abstract Chemical data on hydrothermally altered volcanic rocks from a green tuff belt in Japan indicate that the average rate of Mg removal from seawater due to seawater cycling through back-arc basins in the circum-Pacific region during the early to middle Miocene (25–15 Ma) is estimated to be 2.6±1 × 10¹³ g/year. This is similar to that through present-day mid-ocean ridges (2.4 × 10¹³ g/year). Hydrothermal fluxes of K, Ca and Si are calculated to be 4.2±1.6 × 10¹³ g/year, 4.3±1.7×10¹³ g/year and 1.0±0.4 × 10¹⁴ g/year, respectively. These calculated results indicate that the seawater/volcanic rocks interaction at subduction-related tectonic settings have to be taken into account in considering the geochemical mass balance of seawater over geologic time.     

The influence of variable topography on the depositional behaviour of pyroclastic density currents: The examples of the Upper Pollara eruption (Salina Island,southern Italy)

Stratigraphic reconstruction of the Upper Pollara eruption has allowed for the inference of eruptive mechanisms and the distillation of a sedimentological model for pyroclastic density currents (PDCs) moving across variable topography. The pre-eruptive topography in the study area was characterised by a tuff ring-like morphology, with both inward and outward dipping slopes. Highly viscous, moderately porphyritic, dacitic to rhyolitic magmas fed the eruption, which was characterised by a Vulcanian eruptive style. The stratigraphic succession was divided into five eruption units (EUs), which result from different phases of the eruption separated by stases. Sustained columns occurred only during EU1, while PDC generation dominates EU2–5. Lithofacies analysis of the PDC deposits indicates the prevalence of massive coarse-grained deposits on the inner slopes of the Pollara crater, which are interpreted as the deposits of a flow-boundary zone dominated by granular flow or fluid escape regimes. Dune-bedded, massive to stratified lithofacies dominate the outer slopes of the Pollara crater, and are interpreted as the deposits of PDCs with flow-boundary zones in which traction played a major role. Thin, massive PDC deposits are exposed on the sub-horizontal Malfa terrace, and are interpreted as representative of flow-boundary zones dominated by a granular flow regime. The occurrence of stacked deposits indicates that most of the PDCs were characterised by unsteady pulsatory behaviour, with development of trains of pulses during their transport. The downcurrent lithofacies transitions observed for the Upper Pollara deposits have finally been compared with other similar lithofacies associations which have been described for short-lived PDCs at tuff rings, in order to discuss the influence of pre-eruptive topography on lithofacies association.     

Reconstruction of a major caldera-forming eruption from pyroclastic deposit characteristics: Kos Plateau Tuff, eastern Aegean Sea

The 161 ka explosive eruption of the Kos Plateau Tuff (KPT) ejected a minimum of 60 km³ of rhyolitic magma, a minor amount of andesitic magma and incorporated more than 3 km³ of vent- and conduit-derived lithic debris. The source formed a caldera south of Kos, in the Aegean Sea, Greece. Textural and lithofacies characteristics of the KPT units are used to infer eruption dynamics and magma chamber processes, including the timing for the onset of catastrophic caldera collapse.The KPT consists of six units: (A) phreatoplinian fallout at the base; (B, C) stratified pyroclastic-density-current deposits; (D, E) volumetrically dominant, massive, non-welded ignimbrites; and (F) stratified pyroclastic-density-current deposits and ash fallout at the top. The ignimbrite units show increases in mass, grain size, abundance of vent- and conduit-derived lithic clasts, and runout of the pyroclastic density currents from source. Ignimbrite formation also corresponds to a change from phreatomagmatic to dry explosive activity. Textural and lithofacies characteristics of the KPT imply that the mass flux (i.e. eruption intensity) increased to the climax when major caldera collapse was initiated and the most voluminous, widespread, lithic-rich and coarsest ignimbrite was produced, followed by a waning period. During the eruption climax, deep basement lithic clasts were ejected, along with andesitic pumice and variably melted and vesiculated co-magmatic granitoid clasts from the magma chamber. Stratigraphic variations in pumice vesicularity and crystal content, provide evidence for variations in the distribution of crystal components and a subsidiary andesitic magma within the KPT magma chamber. The eruption climax culminated in tapping more coarsely crystal-rich magma. Increases in mass flux during the waxing phase is consistent with theoretical models for moderate-volume explosive eruptions that lead to caldera collapse.     

Paleomagnetism of a pyroclastic flow deposit and its correlative widespread tephra in central Japan: Possible tectonic rotation since the late Pleistocene

Abstract The late Pleistocene Kamitakara Pyroclastic Flow Deposit (KPFD) and its correlative Kasamori (Ks22) Tephra in central Japan are found to preserve stable thermoremanent magnetization (TRM) and detrital remanent magnetization (DRM), respectively. Untilted site‐mean declinations of the KPFD are characterized by a fairly large scatter with easterly deflection, while those of the Ks22 show significantly smaller deflections. Because northerly paleomagnetic directions consistently characterize shallow marine sediments intercalating the Ks22 layer, the directional discordance is not attributed to different acquisition timing between TRM and DRM, but is probably due to a recent tectonic rotation in central Japan. Large scatter in TRM declinations of the KPFD implies that a number of right‐lateral active faults around the depositional area of the pyroclastic flow raised differential rotation of crustal blocks in central Japan, even during the late Pleistocene.     

Pyroclastic flows from the 1991 eruption of Unzen volcano,Japan

Pyroclastic flows from Unzen were generated by gravitational collapse of the growing lava dome. As soon as the parental lobe failed at the edge of the dome, spontaneous shattering of lava occurred and induced a gravity flow of blocks and finer debris. The flows had a overhanging, tongue-like head and cone- or rollershaped vortices expanding outward and upward. Most of the flows traveled from 1 to 3 km, but some flows reached more than 4 km, burning houses and killing people in the evacuated zone of Kita-kamikoba on the eastern foot of the volcano. The velocities of the flows ranged from 15 to 25 m/s on the gentle middle flank. Observations of the flows and their deposits suggest that they consisted of a dense basal avalanche and an overlying turbulent ash cloud. The basal avalanche swept down a topographic low and formed to tongue-like lobe having well-defined levees; it is presumed to have moved as a non-Newtonian fluid. The measured velocities and runout distances of the flows can be matched to a Bingham model for the basal avalanche by the addition of turbulent resistance. The rheologic model parameters for the 29 May flow are as follows: the density is 1300 kg/m³, the yield strength is 850 Pa, the viscosity is 90 Pa s, and the thickness of the avalanche is 2 m. The ash cloud is interpreted as a turbulent mixing layer above the basal avalanche. The buoyant portions of the cloud produced ash-fall deposits, whereas the dense portions moved as a surge separated from the parental avalanche. The ash-cloud surges formed a wide devastated zone covered by very thin debris. The initial velocities of the 3 June surges, when they detached from avalanches, are determined by the runout distance and the angle of the energy-line slope. A comparison between the estimated velocities of the 3 June avalanches and the surges indicates that the surges that extended steep slopes along the avalanche path, detached directly from the turbulent heads of the avalanches. The over-running surge that reached Kita-Kamikoba had an estimated velocity higher than that of the avalanche; this farther-travelled surge is presumed to have been generated by collapse of a rising ash-cloud plume.     

A new regime of volcanic eruption due to the relative motion between liquid and gas

In explosive magma eruptions, magma ascends through a conduit as a Poiseuille flow at depth, and gas exsolves gradually and expands as the pressure decreases (bubbly flow regime). When the volume fraction of gas becomes sufficiently large, liquid or solid parts of magma fragment into droplets or ashes, and the flow dynamics becomes governed by the gas phase (gas–ash flow regime). We propose a new flow regime, which we call fractured-turbulent flow regime, between the bubbly flow regime and the gas–ash flow regime. In the new regime, both liquid magma and gas are continuous phases. The high connectivity of the two phases allows the relative velocity between them to increase significantly. We present one sample calculation, which displays basically explosive characteristics, but has three features distinct from previous models. The explosive characteristics are manifested as the fragmentation of the magma and the high speed jet that issues from the vent. The first distinct feature is a nearly lithostatic pressure distribution, which results from the increase of the height of the fragmentation surface. The second one is the atmospheric pressure at the vent; the flow is not choked. The third one is that the relative velocity between the gas and the ash is large at the vent despite the large interaction force between the two phases. The large relative velocity is established in the fractured-turbulent regime, and is maintained in the subsequent gas–ash flow regime.     

Generation of block and ash flows during the 1990–1995 eruption of Unzen Volcano, Japan

Processes generating block and ash flows by gravitational dome collapse (Merapi-type pyroclastic flow) were observed in detail during the 1990–1995 eruption of Unzen volcano, Japan. Two different types were identified by analysis of video records and observations during helicopter flights. Most of the block and ash flows erupted during the 1991–1993 exogenous dome growth stage initially involved crack propagation due to cooling and flowage of the dome lava lobes. The mass around the crack became unstable, locally decreasing in tensile strength. Finally, a slab separated from the lobe front, fragmented progressively from the base to the top within a few seconds, and became a block and ash flow. Rock falls immediately followed, in response to local instability of the lobe front. Clasts in these rock falls fragmented and merged with the preceding flow. In contrast, block and ash flows during the endogenous dome growth stage in 1994 were generated due to local bulge of the dome. Unstable lava blocks collapsed and subsequently fragmented to produce block and ash flows.     

Tsunami generation by pyroclastic flow during the 3500-year B.P. caldera-forming eruption of Aniakchak Volcano, Alaska

 A discontinuous pumiceous sand, a few centimeters to tens of centimeters thick, is located up to 15 m above mean high tide within Holocene peat along the northern Bristol Bay coastline of Alaska. The bed consists of fine-to-coarse, poorly to moderately well-sorted, pumice-bearing sand near the top of a 2-m-thick peat sequence. The sand bed contains rip-up clasts of peat and tephra and is unique in the peat sequence. Major element compositions of juvenile glass from the deposit and radiocarbon dating of enclosing peat support correlation of the pumiceous sand with the caldera-forming eruption of Aniakchak Volcano. The distribution of the sand and its sedimentary characteristics are consistent with emplacement by tsunami. The pumiceous sand most likely represents redeposition by tsunami of climactic fallout tephra and beach sand during the approximately 3.5 ka Aniakchak caldera-forming eruption on the Alaska Peninsula. We propose that a tsunami was generated by the sudden entrance of a rapidly moving, voluminous pyroclastic flow from Aniakchak into Bristol Bay. A seismic trigger for the tsunami is unlikely, because tectonic structures suitable for tsunami generation are present only south of the Alaska Peninsula. The pumiceous sand in coastal peat of northern Bristol Bay is the first documented geologic evidence of a tsunami initiated by a volcanic eruption in Alaska. Received: 3 December 1997 / Accepted: 11 April 1998     

Lithofacies and eruption ages of Late Cretaceous caldera volcanoes in the Himeji–Yamasaki district, southwest Japan: Implications for ancient large-scale felsic arc volcanism

Abstract The Himeji–Yamasaki region in the Inner Zone of southwest Japan is underlain mainly by Late Cretaceous volcanic rocks called the Ikuno Group or the Hiromine and Aioi Groups. A new stratigraphic and geochronological study shows that the volcanic rocks in this area consist of 15 eroded caldera volcanoes between 82 and 65 Ma; they are, in order of decreasing age, the Hiromine, Hoden, Ibo, Okawachi, Seppikosan, Hayashida, Shinokubi, Fukusaki, Kurooyama, Ise, Fukadanigawa, Nagusayama, Matobayama, Yumesaki and Mineyama Formations. These calderas vary in diameter from 1 to 20 km and are bounded by steep unconformities; they coalesce and overlap each other. The individual caldera fills are composed mainly of single voluminous pyroclastic flow deposits, which are often interleaved with debris avalanche deposits and occasionally underlie lacustrine deposits. The intracaldera pyroclastic flow deposits are made up of massive, welded or non‐welded tuff breccia to lapilli tuff, and are characterized by their great thickness. The debris avalanche deposits are ill‐sorted breccia, generated by the collapse of the caldera wall toward the caldera floor during the pyroclastic‐flow eruption. The large calderas that are more than 10 km in diameter contain original values of approximately 100 km³ of intracaldera pyroclastic flow deposits. These large calderas are similar to the well‐known Valles‐type calderas in their dimensions, although it is uncertain whether their caldera floors are coherent plates or incoherent pieces. Conversely, the small calderas have diatreme‐like subsurface structures. The variety of the caldera volcanoes in this area is caused by the difference in the volume of caldera‐forming pyroclastic eruptions, as the large and small calderas coexisted. The caldera‐forming eruption rates in Late Cretaceous southwest Japan, including the studied area, were similar to those in late Cenozoic central Andes and northeast Honshu arc, Japan, but obviously smaller than those of late Cenozoic intracratonic caldera clusters in western North America and the Quaternary extensional volcanic arcs in Taupo, New Zealand. The widespread Late Cretaceous felsic igneous rocks in southwest Japan were generated by a long‐term accumulation of low‐rate granitic magmatism at the eastern margin of the Eurasian Plate.     

Contrasting styles of explosive eruption at Sete Cidades,São Miguel,Azores, in the last 5000 years: Hazard implications from modelling

The deposits of three eruptions in the last 5000 years are described in detail in order to constrain eruptive parameters and allow a quantitative assessment of the hazard from a range of explosive eruption types at Sete Cidades volcano, São Miguel, Azores. These deposits include: the Caldeira Seca eruption (P17) which occurred around 600 yr BP, which was the last explosive event from inside the Sete Cidades caldera, the P11 eruption, dated at 2220 ± 70 yr BP, and the undated P8 eruption (< 3000 yr BP). These deposits were chosen to represent the range of likely explosive activity from the caldera.     

Tephra-fall deposits from the 1992 eruption of Crater Peak, Alaska: implications of clast textures for eruptive processes

 The 1992 eruption of Crater Peak, Mount Spurr, Alaska, involved three subplinian tephra-producing events of similar volume and duration. The tephra consists of two dense juvenile clast types that are identified by color, one tan and one gray, of similar chemistry, mineral assemblage, and glass composition. In two of the eruptive events, the clast types are strongly stratified with tan clasts dominating the basal two thirds of the deposits and gray clasts the upper one third. Tan clasts have average densities between 1.5 and 1.7 g/cc and vesicularities (phenocryst free) of approximately 42%. Gray clasts have average densities between 2.1 and 2.3 g/cc, and vesicularities of approximately 20%; both contain abundant microlites. Average maximum plagioclase microlite lengths (13–15 μm) in gray clasts in the upper layer are similar regardless of eruptive event (and therefore the repose time between them) and are larger than average maximum plagioclase microlite lengths (9–11 μm) in the tan clasts in the lower layer. This suggests that microlite growth is a response to eruptive processes and not to magma reservoir heterogeneity or dynamics. Furthermore, we suggest that the low vesicularities of the clasts are due to syneruptive magmatic degassing resulting in microlitic growth prior to fragmentation and not to quenching of clasts by external groundwater. Received: 5 September 1997 / Accepted: 1 February 1998     

Incremental growth of a large volume,chemically zoned magma body: a study of the tephra sequence beneath the Rainier Mesa ash flow sheet of the Timber Mountain Tuff

The Rainier Mesa ash-flow is a large (1200 km³), 11.6 My old, chemically zoned unit that ranges in composition from 55 to 76% SiO₂ — one of the largest chemical ranges ever observed in a large volume ash-flow sheet. Two chemical trends occur in this sheet, a low silica (55–66% SiO₂) and a high silica (>66% SiO₂) trend. Ninety per cent of the Rainier Mesa sheet occurs in the high silica trend. Immediately beneath the Rainier Mesa sheet is a thick tephra sequence. The chemical variation of this sequence is nearly equivalent to the high silica portion of the Rainier Mesa ash-flow sheet (about 66–78% SiO₂). Throughout the tephra sequence numerous small ash-flow layers occur, and each ash-flow layer is chemically zoned from more evolved at the base to less evolved at the top. This is consistent with having been erupted from a zoned magma body. The lowest silica tephra units are at the base of the sequence and the highest silica units are at the top — that is, the large-scale chemical trend of the entire sequence is opposite to that of the individual ash-flow layers. These ash-flow layers are of very small volume. The tephra sequence provides a unique record of the incremental development of the zoned, high silica portion of the Rainier Mesa magma body.     

Sequence and eruptive style of the 1783 eruption of Asama Volcano, central Japan: a case study of an andesitic explosive eruption generating fountain-fed lava flow, pumice fall, scoria flow and forming a cone

The 3-month long eruption of Asama volcano in 1783 produced andesitic pumice falls, pyroclastic flows, lava flows, and constructed a cone. It is divided into six episodes on the basis of waxing and waning inferred from records made during the eruption. Episodes 1 to 4 were intermittent Vulcanian or Plinian eruptions, which generated several pumice fall deposits. The frequency and intensity of the eruption increased dramatically in episode 5, which started on 2 August, and culminated in a final phase that began on the night of 4 August, lasting for 15 h. This climactic phase is further divided into two subphases. The first subphase is characterized by generation of a pumice fall, whereas the second one is characterized by abundant pyroclastic flows. Stratigraphic relationships suggest that rapid growth of a cone and the generation of lava flows occurred simultaneously with the generation of both pumice falls and pyroclastic flows. The volumes of the ejecta during the first and second subphases are 0.21 km³ (DRE) and 0.27 km³ (DRE), respectively. The proportions of the different eruptive products are lava: cone: pumice fall=84:11:5 in the first subphase and lava: cone: pyroclastic flow=42:2:56 in the second subphase. The lava flows in this eruption consist of three flow units (L1, L2, and L3) and they characteristically possess abundant broken phenocrysts, and show extensive "welding" texture. These features, as well as ghost pyroclastic textures on the surface, indicate that the lava was a fountain-fed clastogenic lava. A high discharge rate for the lava flow (up to 10⁶ kg/s) may also suggest that the lava was initially explosively ejected from the conduit. The petrology of the juvenile materials indicates binary mixing of an andesitic magma and a crystal-rich dacitic magma. The mixing ratio changed with time; the dacitic component is dominant in the pyroclasts of the first subphase of the climactic phase, while the proportion of the andesitic component increases in the pyroclasts of the second subphase. The compositions of the lava flows vary from one flow unit to another; L1 and L3 have almost identical compositions to those of pyroclasts of the first and second subphases, respectively, while L2 has an intermediate composition, suggesting that the pyroclasts of the first and second subphases were the source of the lava flows, and were partly homogenized during flow. The complex features of this eruption can be explained by rapid deposition of coarse pyroclasts near the vent and the subsequent flowage of clastogenic lavas which were accompanied by a high eruption plume generating pumice falls and/or pyroclastic flows.Editorial responsibility: T. Druitt