Massey Documents by Type

Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294

Browse

Has files: YesSubject: search.filters.subject.Volcanic eruptions

Search Results

Now showing 1 - 10 of 11
  • Thumbnail Image
    Item
    ‘All Four Engines Have Failed’: A qualitative study of the health impacts, reactions and behaviours of passengers and crew onboard flight BA009 which flew through a volcanic ash cloud in 1982
    (Elsevier Ltd, 2025-06-15) Meach R; Horwell CJ; de Terte I
    This study investigated the experiences, health impacts and behaviours of passengers and crew onboard British Airways flight BA009 which flew through a volcanic ash cloud from Mount Galunggung, Indonesia, in 1982. In addition to secondary data sources, including a book published by one of the passengers, 18 semi-structured interviews were completed (14 passengers, 2 flight crew and 2 cabin crew) which were video recorded and transcribed verbatim. Data were analysed using reflexive thematic analysis to examine the experiences, behaviours, and actions of those onboard, and the health impacts of exposure to volcanic emissions. Our analysis identified five key themes which explain how people onboard flight BA009 responded: 1) Responsibility, 2) Airmanship and prior knowledge of aviation, 3) Upbringing and cultural background, 4) Faith and 5) Behaviour of the crew. Our study found few physical health impacts associated with the exposure to the ‘smoke’ and, despite individual cases of distress, there was no mass panic onboard the aircraft. Our findings highlight valuable information on passenger and crew behaviour in aviation crises, the risks of volcanic ash clouds to aviation, and have practical implications for aviation disaster management, planning and communication.
  • Thumbnail Image
    Item
    Effects of multiphase turbulence on the flow and hazard behavior of dilute pyroclastic density currents : a thesis submitted in partial fulfilment for the degree of Doctor of Philosophy in Earth Science at Massey University, Palmerston North, New Zealand
    (Massey University, 2024-06-16) Uhle, Daniel Holger
    Pyroclastic surges (also dilute pyroclastic density currents or dilute PDCs) are amongst the most hazardous volcanic phenomena associated with explosive volcanic eruptions and hydrothermal explosions. These fast-moving, turbulent, polydisperse multiphase flows of hot volcanic particles and gas occur frequently and have severe impacts on life and infrastructure. This is attributed to a compounding of hazard effects: large flow-internal dynamic pressures of tens to hundreds of kilopascals destroy reinforced buildings and forests; temperatures of up to several hundreds of degrees Celsius pose severe burn hazards; and readily respirable hot fine ash particles suspended inside dilute PDCs cause rapid asphyxiation. Direct measurements inside pyroclastic density currents are largely absent, and previous research has used a combination of detailed field studies on PDC deposits, laboratory experiments on analog density currents, numerical modeling, and theoretical work to interrogate the internal flow structure, gas-particle transport, sedimentation and destructiveness of dilute PDCs. Despite major scientific advances over the last two decades, significant fundamental gaps in understanding the turbulent multiphase flow behavior of dilute PDCs endure, preventing the development of robust volcanic hazard models that can be deployed confidently. Critical unknowns remain regarding: (i) how turbulence is generated in dilute PDCs; (ii) how multiphase processes modify the flow and turbulence structure of dilute PDCs; and (iii) if and how turbulent gas-particle feedback mechanisms affect their destructiveness. To address these gaps in understanding, this PhD research involved high-resolution measurements of velocity, dynamic pressure, particle concentration, and temperature inside large-scale experimental dilute PDCs. It is shown that dilute PDCs are characterized by a wide turbulence spectrum of damage-causing dynamic pressure. This spectrum is strongly skewed towards large dynamic pressures with peak pressures that exceed bulk flow values, routinely used for hazard assessments, by one order of magnitude. To prevent severe underestimation of the damage potential of dilute PDCs, the experimentally determined ratio of turbulence-enforced pressure maxima and routinely estimated bulk pressures should be used as a safety factor in hazard assessments. High-resolution measurements of dynamic pressure and Eulerian-Lagrangian multiphase simulations reveal that these pressure maxima are attributed to the clustering of particles with critical particle Stokes numbers (𝑆𝑡=𝒪(1)) at the margins of coherent turbulence structures. The characteristic length scale and frequency of coherent structures modified in this way are controlled by the availability of the largest particles with critical Stokes number. Through this, spatiotemporal variations in peak pressures are governed by the mass loading and subsequent sedimentation of these clustered particles. In addition to the ‘continuum phase’ loading pressure, the measurements also revealed that the direct impact of clustered margins and high Stokes number particles decoupling from margins with structures generate instantaneous impacts. These piercing-like impact pressures exceed bulk pressure values by two orders of magnitude. Particle impact pressures can cause severe injuries and damage structures. They can be identified as pockmarks on buildings and trees after eruptions. This new type of PDC hazard and the magnitude of pressure impacts need to be accounted for in hazard assessments. Systematic measurements of the evolving experimental pyroclastic surges along the flow runout demonstrate that time-averaged vertical profiles of all flow velocity components and flow density obey self-similar distributions. Variations of the roughness of the lower flow boundary, geometrically scaling ash- to boulder-sized natural substrates, showed the self-similar distributions are independent of the roughness. Mathematical relationships developed from the self-similar velocity and density distribution reveal the self-similar vertical distribution of mean dynamic pressure. This empirical model can inform multi-layer PDC models and estimate the height and values of peak time-averaged dynamic pressure for dilute PDCs of arbitrary scale. Turbulence fluctuations around the mean were investigated through Reynolds decomposition. The large-scale turbulence structure and the dominant source of turbulence generation are shown to be controlled by free shear with the outer flow boundary, while strong density gradients at the basal high-shear flow boundary dampen turbulence generation. The large-scale, shear-induced coherent turbulence structures can be tracked along the runout and were found to be superimposed by smaller turbulence structures. In Fourier spectra of dynamic pressure, flow velocity, and temperature, these sub-structures are observed as discrete frequency bands that correspond to the coarse modes of the spatiotemporally evolving flow grain-size distributions. This can be associated with the preferential clustering of particles at the peripheries of the sub-structures. Following the decoupling of particle clusters, the rapid sedimentation of particle clusters occurs periodically at the characteristic frequency of the turbulence sub-structures. This mechanism of preferential clustering, decoupling and rapid sedimentation of particles with critical particle Stokes numbers is an important mechanism of turbulent sedimentation to explain the spatiotemporally evolving flow grain-size distribution of pyroclastic surges.
  • Thumbnail Image
    Item
    The Arxan-Chaihe Volcanic Field of monogenetic volcanism in intracontinental settings in NE China : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science at Massey University, Palmerston North, New Zealand
    (Massey University, 2024-02) Li, Boxin
    Pliocene to Recent Arxan-Chaihe Volcanic Field (ACVF) is composed of at least 47 vents, representing various types of volcanism, such as Strombolian style, phreatomagmatic explosive, effusive, and lava-fountaining eruptions. These eruptions produced scoria cones, fissure-aligned spatter cones, and tuff rings with a few surrounding maar craters. Field observations imply that the lava-fountaining eruptions are more common on the western side of ACVF, represented by Yanshan (YS)-the Triple Vent, and Daheigou (DHG). In the southwest part of ACVF, lava flows and loose pyroclastic ejecta, such as scoria, mark the eruption events that took place during the Holocene era about 2000 years ago. Dichi (Earth Pond) Lake, with fissures on its eastern side, formed by lava-effusive eruption styles with spatter rows occurring along a fissure, while the low-lying western side of the vent chain is a maar volcano cut into the pre-eruptive lava sheets. Tianchi (Heaven Lake) Lake and Tuofengling (Camel Hump) Lake on the western side of ACVF preserves a range of well-exposed pyroclastic deposits consistent with edifice-building successions. These are composed of scoriaceous pyroclastic materials, yielding construction histories of complex cones (with both "wet" and "dry" explosive eruptive phases). The most significant and largest vent is in the eastern corner of ACVF, Tongxin Lake, a complex phreatomagmatic eruption-style volcano with a maar crater and thick rim deposits. Tongxin Lake is interpreted to be a maar lake that erupted into an intra-montane basin. Intact pyroclastic deposits are preserved within a km from the crater rim and at least 5 meters thick. Stratigraphic and granulometric analyses from five sites around Tongxin Lake indicate the tuff ring of Tongxin was built by processes associated with magma-water interactions that fueled violent explosive eruptions during distinct syn-eruptive stages. Geochemistry is consistent with at least three magma sources contributing to the formation of the complex eruptive products that build the large tuff ring of the maar edifice. Geomorphology terrain analyses performed through GIS-based applications (QGIS) imply that the diverse range of local geology, especially the pre-eruptive topography, was confined and reshaped by the subsequent Pliocene to Recent volcanism in ACVF. Lava flows within ACVF were emplaced over large areas around the two major fluvial systems: Halaha River in the west and Chaoer River in the east of ACVF. The lava flows in the west of ACVF are generally young and can be modelled using the Q-LavHA plug-in of QGIS. The model has been utilized to simulate lava flow inundation and indicates diverse flow along the flow axis as well as lateral and temporal variations during the evolution of the edifice. Other studies of ACVF, e.g., hazard management, concluded that violent phreatomagmatic explosive events had impacted the fluvial valleys that are commonly associated with structural weakness zones in ACVF. In addition, lava-effusive and lava-fountaining eruptions in urban areas and along major utilities (e.g., roads, geopark facilities or powerlines) also could be heavily impacted by fissure-fed lava flows and potential phreatomagmatic explosions controlled by the local hydrology conditions.
  • Thumbnail Image
    Item
    Re-establishing the beast : an investigation into the spatiotemporal evolution of the Y5 phase of the Taupō 232 ± 10 CE eruption, New Zealand : a thesis submitted in partial fulfilment of a Philosophiae Doctor degree in Earth Science, Volcanic Risk Solutions, School of Agriculture and Environment, Massey University, New Zealand
    (Massey University, 2023) Tapscott, Sarah Joanne
    Plinian eruptions are sustained, high-energy explosive eruptions that generate buoyant plumes that reach >20 km into the atmosphere. They often produce devastating pyroclastic density currents (PDC) along with widespread tephra fall out, with significant hazards to communities around the volcanoes. Current computational modelling of Plinian eruptions considers generalized steady versus unsteady column regimes as the explanation for the formation of coeval buoyant Plinian plumes and intraplinian PDCs; however, natural eruption scenarios indicate that these regimes can oversimplify the interpretation of both PDC and plinian fall deposits. The large-Plinian Y5 phase of the Taupō 232 ± 10 CE eruption has an exceptionally widespread and well-preserved deposit that incorporates fall and coeval PDCs. Despite an extensive dataset in place for the Y5, there remain conflicting views on the interpretation of its deposit regarding eruption and sedimentation dynamics. Original studies by Walker (1980) considered the Y5 as a single eruptive unit from the perspective of the widespread fall deposit, without consideration of the intraplinian, coeval Early Flow Units (EFU) identified by Wilson & Walker (1985). Walker’s study determined that the Y5 phase involved a Plinian plume ~50 km high. Bedding characteristics in the fall deposit were considered in detail by Houghton et al (2014), who used their qualitative observations to propose the presence of 26 subunits within the Y5 fall deposited by a fluctuating plume influenced by strong changes in wind direction. Houghton et al.’s study brought the plume height down to (35 – 40 km) and denoted a vent location ~6 km SW of that proposed by Walker (1980). This Ph.D. research presents a comprehensive quantitative dataset of the deposit characteristics in the vertical stratigraphy of the upper phreatoplinian Y4 deposit, and the coeval fall and PDC deposits of the Plinian Y5 phase of the Taupō eruption. The dataset is used to reconstruct the spatiotemporal evolution of the Y5 phase and improve our understanding of Plinian eruption dynamics and sedimentation. Detailed sample collection and analysis was conducted on proximal to medial deposit exposures, whose vertical stratigraphy encompass the final stage of the Y4 (Y4-G), the Y5 fall deposit and its coeval Early Flow Units (EFU). Samples were analysed for grain size distributions, componentry, and juvenile textural characteristics. It is demonstrated that foreign lithic lithologies and their time-relative abundance in relation to other deposit characteristics play an important role in informing vent location, the evolution of the conduit and the nature of generation of erupted facies (i.e., PDC and fall). In this study, foreign lithics were subdivided by their inferred stratigraphic depth of origin below the lake floor into: F1) pre-232 ± 10 CE volcanic material (~0 – 400 m), F2) predominantly Huka Group sediments, minor Whakamaru ignimbrite and hydrothermally altered material (~400 – 3000 m), and F3) plutonic microdiorites and granitoids (>4000 m) At the boundary between the Y4-G and Y5 deposits, a decrease in obsidian abundance of c. 30 wt.%, along with an increase in F2 lithics of c. 20 wt.%, and a drop in pumice vesicularity by c. 30 % indicate a distinct change in vent location between the Y4 and Y5 phases. F2 lithologies in the Y4 differ significantly from those in the Y5, but coincide with those of the plinian Y2 deposit, suggesting similar regions of crustal excavation for Y5 and Y2 and imply a vent location comparable to that of the Y2 phase. Vertical variations in the abundance and relative proportions of different juvenile and lithic pyroclasts, in pyroclast textures and pumice densities identified in the Y5 fall deposit, following the initial clearing of the vent, define three successive stages within a relatively steady, continuous eruption. These stages are: 1) the continuous excavation of the conduit at relatively low mass eruption rate shown through higher lithic:pumice ratios, finer overall grain size and higher pumice densities compared to later stages of the Y5; 2) increasing mass eruption rate towards a climax with relatively steady conduit erosion coinciding with deepening fragmentation, exhibited in increasingly larger grain sizes and relatively lower total lithic abundances, yet higher relative proportions of F2 and F3 lithics; and 3) a moderate decrease in mass eruption rate and the acceleration of conduit erosion (shown through a rapid increase in F1 abundance and decreasing grain size), promoting the potential early onset of caldera collapse that led to the Y6 ignimbrite producing blast event. Vertical bedding features in the Y5 fall deposit are shown to be laterally discontinuous and pinch out over length scales of 101-103 m. This precludes the possibility that coarse-fine fluctuations were caused by mass partitioning of material during partial column collapse, or by variations in wind direction. Instead, I suggest that the bedform features identified in the Y5 deposit result from gravitational instabilities in the umbrella cloud, sedimenting as tephra swathes. Additionally, the intraplinian EFUs were differentiated by their characteristics into two main types: Type 1 centimetre to metre thick, massive, pink-orange to cream coloured, coarser grained deposits that are topographically confined; and Type 2 decimetre to centimetre thick, massive to moderately stratified, white-grey, finer grained deposits that have mounted topography. The anomalously high proportion of ash (<10 µm at 4 – 27 wt.%) in the EFU deposits, in conjunction with a lack of evidence for enrichment of dense clasts (i.e., lithics and crystals), indicates that there was minimal to no mass partitioning that would be expected in the case of partial column collapse. In addition, the inferred high particle concentration of the Type 1 flows and their high temperature emplacement indicates that the materials that propagated to form the EFU PDCs is likely to have originated from lower heights around the jet where entrained air had limited effect to cool the mixture. A lack of variation in the proportion of lithic types and juveniles between Type 1 and Type 2 with relative height compared to the Y5 fall suggests that the EFU are a product of one generation mechanism and that the deposit types 1 and 2 represent contrasts in relative volume, runout distance, and/or topographical constraints on runout of individual flows. The EFU are entirely contained within fall activity and become more abundant, voluminous and/or increase in flow mobility with increasing mass eruption rate during the Y5 phase. The generation mechanism for the EFU PDCs strongly aligns with the modelling and field observations for gargle dynamics, where a dense sheath formed by recycled pre-existing material in a basin-like vent structure develops on an eruptive jet. This dense sheath produces PDCs simultaneous with a sustained Plinian column that occurs seemingly without interruption. Similarities can be drawn with deposits from other, historical large-Plinian eruptions such as the Bishop Tuff, 0.76 Ma and Novarupta, 1912, which also involved phases of coeval fall and PDC deposition analogous to the Y5 and EFUs, and were likely produced through gargle dynamics. This study has shown that through the detailed, quantitative characterisation of deposit features in plinian eruption deposits involving coeval fall and PDCs, the temporal changes in eruptive behaviour, conditions at source and the nature of sedimentation can be identified. Interpretations indicate that the Y5 phase of the Taupō 232 ± 10 CE eruption was a large, steady, and extremely powerful eruption beyond the general depiction of a ‘standard’ Plinian event. Using quantitative analysis such as this may help build upon our knowledge base of the eruption and sedimentation dynamics of large Plinian eruptions by providing a field-based foundation for the reconstruction of the spatiotemporal evolution of such events. This is intended to provide a pathway for the amalgamation of field data and computational eruption models, ultimately improving our ability to forecast and mitigate explosive eruption hazards at similar volcanoes globally.
  • Thumbnail Image
    Item
    Eruption dynamics and frequency-magnitude relationships of explosive eruptions at Mt. Ruapehu, New Zealand over the past 1800 years : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science at Massey University, Palmerston North, New Zealand
    (Massey University, 2020) Voloschina, Marija
    Small to moderate explosive eruptions (VEI ≤ 3) constitute the most frequent eruptions and often involve several phases characterised by different eruption dynamics. Deposits associated with small-scale multi-phase eruptions tend to be underrepresented in geological records and the resulting probabilistic eruption forecast models. This PhD research presents a refined high-resolution tephrostratigraphic framework for the 1800-year Tufa Trig Formation at one of New Zealand’s most active volcanoes, Mt. Ruapehu. This framework is used to characterise short- and long-term changes in eruption behaviour aiming to identify time-variable processes in the volcanic system of a long-lived andesite volcano. Systematic mapping and lithosedimentological characterisation of tephra deposits are combined with geochemical fingerprinting and radiocarbon dating to create a detailed frequency-magnitude record of single- and multi-phase eruptions of the last 1800 years. At least 32 eruptions can be identified, ranging from low to mid-intensity single-phase eruptions (1–10 × 10⁶ m³ deposit volumes) to complex multi-phase eruptions up to two magnitudes larger. The largest eruption is the T13-sequence that comprises at least 5 eruption phases. Multi-lobate dispersal pattern and componentry analyses show that individual eruption phases represent multiple fall events of similar eruption style and magnitude. Major and trace element analyses of juvenile glass display limited syn-sequence variability, while heterogeneous pyroclast and textural characteristics suggest that short-term changes in eruption behaviour are predominantly controlled by shallow conduit processes. The frequency-magnitude record is integrated with geochemistry and statistical modelling, identifying time-variable pattern in Mt. Ruapehu’s eruption behaviour: the time span 1718–1300 cal BP involves low-intensity single-phase eruptions every ~40 years and is followed by a low rate regime (one eruption every 125 years). The largest multi-phase eruptions of the last two millennia occur between 610 and 370 cal BP, while the past 370 years are dominated by smaller multi-phase eruptions every ~40 years, suggesting that long-term changes in Mt. Ruapehu’s eruption behaviour are related to changes in magma supply. This research adds critical complexity to the understanding of the processes and timescales controlling eruption behaviour in the southern Taupo Volcanic Zone and provides insights into the dynamic behaviour of small to moderate multi-phase eruptions. These results will constitute the framework for refining dynamic eruption forecast models at Mt. Ruapehu and other similar volcanoes globally.
  • Thumbnail Image
    Item
    Analyzing seismic signals to understand volcanic mass flow emplacement : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Earth Sciences at Massey University, Palmerston North, Manawatu, Aotearoa New Zealand
    (Massey University, 2017) Walsh, Braden Michael Larson
    Natural hazards are one of the greatest threats to life, industry, and infrastructure. It has been estimated that around a half billion people worldwide are in direct proximity to the danger of volcanic hazards. For volcanic mass flows, such as pyroclastic density currents and lahars, extreme runout distances are common. The close proximity of large population centers to volcanoes requires the implementation of early warning and realOtime monitoring systems. A large portion of the progress towards realOtime monitoring is through the use of geophysical instrumentation and techniques. This research looks into emerging geophysical methods and tries to better constrain and apply them for volcanic purposes. Specifically, multiple types of amplitude source location techniques are described and used for locating and estimating the dynamics of volcanic mass flows and eruptions. Other methods, such as semblance and back projection, are also employed. Applying the active seismic source method to a lahar that occurred on October 13th 2012 at Te Maari, New Zealand, locations and estimations of lahar energy were calculated in an increased noise environment. Additionally, the first ever calibration of the amplitude source location (ASL) method was conducted using active seismic sources. The calibration proved to decrease true error distances by over 50%. More calibration on the ASL method was accomplished by using all three components of the broadband seismometer. Initial results showed that using all three components reduced extreme errors and increase the overall precision of the locations. Finally, multiple geophysical methods (ASL, semblance, back projection, waveform migration, acoustic-seismic ratios) were used to show that a combination of instrumentation could produce more reliable results. This research has filled gaps in the preexisting knowledge for hazards. With these results, more effective hazard warnings can be produced, and systems for real time estimations of locations and dynamics of volcanic events could be developed.
  • Thumbnail Image
    Item
    Understanding the largest-scale explosive volcanism at Mt. Taranaki, New Zealand : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science, Massey University, Palmerston North, New Zealand
    (Massey University, 2017) Torres-Orozco, Rafael
    Over the last 5000 years B.P., at least 53 explosive eruption episodes occurred at Mt. Taranaki, (western North Island, New Zealand) from either the summit-crater (2500 m), or a satellite vent on Fanthams Peak (1966 m). These eruptions are represented in wellpreserved pyroclastic successions on the upper volcano flanks. At least 16 episodes produced deposits with lithostratigraphic characteristics comparable to those of the last sub-Plinian eruption at AD 1655, suggesting an average recurrence of one Plinian/sub-Plinian eruption episode every 300 years. Several large-scale mafic-intermediate (~48-60 wt.% SiO2) eruption episodes sourced from the two vents were studied in detail to determine the “maximum” intensity, magnitude and eruptive styles from this volcano. These episodes comprised climactic phases with sustained and steady, 14-29 km-high eruption columns, often starting and ending with unsteady pulsating, oscillating and collapsing plumes. The columns erupted 0.1-0.5 km3 DRE at mass and volume discharge rates of 107-108 kg/s and 103-104 m3/s, respectively, indicating magnitudes of 4.1-5.1. The unsteady initial, pre- and post-climactic eruptive phases were dominated by domecollapse, column-collapse and lateral-blast pyroclastic density currents (PDCs), with runout distances of 3-19 km and volumes of up to 0.02 km3 DRE. The steadiest phases were associated with eruption of rheologically homogeneous magmas producing homogenous pumice textures. Unsteady phases produced density and porosity pumice gradients by magma stalling in upper conduit levels. Three eruption onset scenarios were developed from this work: a) initial closed-conduit decompression by vent unroofing and domecollapse, b) transient open and clogged conduits produced by repeated plugging-and bursting of chilled or gas-depleted magma, and c) rapid conduit opening with more mafic eruptives. In all scenarios, the climactic phases are comparable, with pyroclastic fallouts covering 1500-2500 km2. The most violent phases of these events, however, are lateralblast PDCs that could reach a broad arc between 14-19 km from source. This reappraisal of the hazardscape at Mt. Taranaki integrates many new details that enable a more realistic hazard management and provides a range of findings that can be applied to other similar andesitic volcanoes prior to reawakening.
  • Thumbnail Image
    Item
    High-precision tephrostratigraphy : tracking the time-varying eruption pulse of Mt. Taranaki, North Island, New Zealand : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science, Massey University, Palmerston North, New Zealand
    (Massey University, 2017) Damaschke, Magret
    In this research it was proposed that a more robust record of volcanic activity for Mt. Taranaki (New Zealand) could be derived from tephras (pyroclastic fall deposits) within cores from several lakes and peatlands across a 120o arc, NE-SE of the volcano, covering a range of prevailing down-wind directions. These data were integrated with previous tephrochronology studies to construct one of the longest and most complete volcanic eruption history records ever developed for an andesitic stratovolcano. Using 44 new radiocarbon dates, electron microprobe analysis of glass shard and ttitanomagnetite chemical composition, along with whole-rock chemistry, a chrono- and chemostratigraphy was established. The new record identifies at least 272 tephraproducing eruptions over the last 30 cal ka BP. Six chemo-stratigraphic groups were identified: A (0.5 – 3 cal ka BP), B (3 – 4 cal ka BP), C (4 – 9.5 cal ka BP), D (9.5 – 14 cal ka BP), E (14 – 17.5 cal ka BP), and F (23.5 – 30 cal ka BP). These were used to resolve previous stratigraphic uncertainties at upper-flank (proximal) and ring-plain (medial) sites. Several well-known “marker tephras” are now recognized as being ~2000 years older than previously determined (e.g., Waipuku, Tariki, and Mangatoki Tephra units) with the prominent Korito Tephra stratigraphically positioned above the Taupo-derived Stent Tephra. Further, new markers were identified, including the Kokowai Tephra unit (~4.7 cal ka BP), at a beach-cliff exposure, 40-km north-east of the volcano. Once age-models were established for each tephra, units were matched between sites using statistical methods. Initial statistical integration showed that the immediate past high-resolution tephrochronological record suffered from a distinctive “old-carbon” effect on its ages (Lake Rotokare). This had biased the most recent probabilistic forecasting and generated artificially high probability estimates (52-59% eruption chance over the next 50 years). Once the Rotokare record was excluded and chemostratigraphy constraints were applied, a reliable multi-site tephra record could be built only for the last ~14 ka BP. The new data confirms a highly skewed distribution of mainly (98% of cases) short intervals between eruptions (mode of ~9 years and average interval ~65 years). Long intervals (up to 580 years) as seen in earlier records were reduced to 2% of the record, but can now be considered real, rather than missing data. The new data confirm a cyclic pattern of varying eruption frequency (with a five-fold range in annual frequency) on a period of ~1000-1500 years. The new time-varying frequency estimates suggest a lower probability for a new eruption at Mt. Taranaki over the next 50 years of 33-42%. The newly established chemostratigraphy was further used to investigate time-related compositional changes. Whole-lapilli analyses highlighted that a specific very evolved Ca-rich and Fe-poor composition was only found within the easterly and south-easterly depositional sites. This was explained by eruption of a stratified magma reservoir, which holds greater modal proportions of plagioclase and lower proportions of pyroxene within low-density, gas-rich upper conduit regions. During the most explosive phases of eruptions, when plumes reach the stratospheric jetstream, the lowest-density pumice is thus dispersed by high-level stable westerly winds. Further, two distinct evolutional trends were seen in the long and new tephrochronological record; from 17.5 to 3 cal ka BP and <3 cal ka BP; with wholelapilli, glass, and titanomagnetite compositions overall evolving over time. The former compositional trend indicates a crystallising and cooling magma source in the deep crust, with multiple, spatially separated magma source regions forming, each generating magmas (i.e., magma batches) with unique titanomagnetite compositions. This trend is interrupted by a distinct shift towards less-evolved compositions and the initiation of a second parasitic vent (Fanthams Peak at the southern flank of Mt. Taranaki).
  • Thumbnail Image
    Item
    Statistical methods for assembling and incorporating volcanic records in hazard estimation : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Statistics at Massey University, Palmerston North, New Zealand
    (Massey University, 2015) Green, Rebecca
    The estimation of hazard arising from volcanic eruptions is a research topic of great interest to New Zealand, given the number and location of active and dormant volcanoes. Probabilistic temporal models are required to handle the stochastic nature of observed records. Such models are usually assembled using point process techniques or renewal theory and most are purely temporal in the sense that they only consider the distribution of event or inter-event times as predictors of further volcanic activity. I demonstrate using a high-resolution eruption record from Mt Taranaki (New Zealand) how geochemical data can be incorporated, using a proportional hazards type approach, to improve the performance of current renewal-type models. Probabilistic forecasting relies on the accuracy and completeness of historical eruption records. This poses the question of how to establish a detailed record of past volcanic events. Multiple sites are needed to build a composite tephra record, but correctly merging them by recognizing events in common and site-speci c gaps remains complex. I present an automated procedure for matching tephra sequences, using stochastic local optimization techniques. Implausible matches are eliminated through careful reasoning, while heuristically searching over the remaining alternatives. Possible matches are veri ed using known tephra compositions and stratigraphic constraints. The method is applied to match tephra records from ve long sediment cores in Auckland, New Zealand. The correlated record compiled is statistically more likely than previously published arrangements from this area. In addition to the matching of tephras found in the Auckland region, the algorithm is applied to stratigraphic records obtained from Mt Taranaki. With more detailed geochemical information available, matches are constrained further by considering principle component analysis of titanomagnetite compositional data. Finally, after combining the amalgamated record of Mt Taranaki events with point thickness measurements, the eruptive volume of Mt Taranaki events is estimated. Utilizing isopach maps and individual point observations a model is formulated, in a Bayesian framework, for the thicknesses of tephra deposits as a function of the distance and angular direction of each location. The model estimates, in addition to eruptive volume, the wind and site-speci c e ects on the thickness deposits. The ndings lead on to methods of incorporating eruptive volumes in hazard estimation.
  • Thumbnail Image
    Item
    Andesitic Plinian eruptions at Mt. Ruapehu (New Zealand) : from lithofacies to eruption dynamics : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science at Massey University, (Palmerston North, Manawatu), New Zealand
    (Massey University, 2012) Pardo Villaveces, Natalia
    A new detailed stratigraphy was developed for a sequence of pyroclastic deposits including the largest known eruptions associated with Mt. Ruapehu, deposited in the period ~27-10 ka BP cal. From the largest Plinian eruption deposits in this sequence, subtle lithofacies variations within componentry, pumice textures and sedimentary features were used to identify a systematic change in eruptive conditions over time. Early eruptions involved steady eruption columns, while younger eruptions involved unsteady, collapsing columns. Isopach and Isopleth (pumice and lithic) mapping of most widespread and distinctive units show that the largest explosive eruptions known from this volcano attained peak column heights between 22 and 37 km, with mass discharge rates reaching 107-108 kg/s. To characterise the conditions controlling the style of Plinian eruptions at this andesitic volcano, and to explain the systematic variation in column stability over time, five key units were sampled in detail, exemplifying the major contrasting lithofacies. The sampled tephras underwent grain-size analysis, along with quantification of componentry, porosimetry and density on particles of a range of size classes, as well as 2D and 3D microtextural analyses of juvenile pumice clasts to define vesicularity and crystallinity. In addition, physiochemical factors such as melt-evolution and volatile-contents were determined by analysing bulk pumice, glass-inclusions and residual glasses with electron microprobe and FTIRspectroscopy. Bulk compositions of these tephras vary from basaltic-andesite to andesite (56-62 wt.%, SiO2), and had minimum pre-eruptive H2O contents of 4-5 wt.%. The evolution of eruption behaviour over time was not correlated to any progressive change in bulk geochemical properties, but instead resulted from variations in physical processes within the conduit. Ascending magmas experienced heterogeneous bubble nucleation, and later-erupted units showed increasing degrees of rheological heterogeneities developed across the conduit. Differences between units were due to changes in the magma decompression rates, the degree of bubble-crystal-melt interactions and bubble shearing, as well as the composition of the residual melt. Conditions that led to the most variable physical states of the magma reaching the fragmentation level resulted in the highest variability in pumice textures, the greatest range in styles of fragmentation, and the most unstable eruption columns. A new model describing the pre-eruptive magma storage region, conduit processes, magma fragmentation, and pyroclastic dispersal during Plinian eruptions at Mt. Ruapehu is proposed. This hypothesises that eruption column unsteadiness and collapse occurs when magma shear reaches extreme levels along the conduit under conditions of low isolated porosity (<3 vol.%). This situation also generates the worst-case hazard scenarios expected for Ruapehu, eruptions, where Plinian columns of over 30 km may produce widespread tephra fall, as well as partially collapse to generate pyroclastic density currents of over 15 km runout.