T. Hincks - Academia.edu (original) (raw)

Papers by T. Hincks

Research paper thumbnail of Probabilistic Modeling of Tephra Dispersion using Parallel Processing

Numerical models of tephra accumulation are important tools in assessing hazards of volcanic erup... more Numerical models of tephra accumulation are important tools in assessing hazards of volcanic eruptions. Such tools can be used far in advance of future eruptions to calculate possible hazards as conditional probabilities. For example, given that a volcanic eruption occurs, what is the expected range of tephra deposition in a specific location or across a region? An empirical model is

Research paper thumbnail of Probabilistic modeling of tephra dispersion : hazard assessment of a multi-phase eruption at Tarawera Volcano , New Zealand

The Tarawera Volcanic Complex (New Zealand) comprises 11 rhyolite domes formed during five major ... more The Tarawera Volcanic Complex (New Zealand) comprises 11 rhyolite domes formed during five major eruptions between 17000 BC and AD 1886, the first four of which were predominantly rhyolitic. The only historical event (AD 1886) erupted about 2 km 3 of purely basaltic tephra fall killing about 150 people. The AD 1305 Kaharoa eruption is the youngest rhyolitic event and erupted a total tephra-fall volume more than two times larger than the AD 1886 eruption. We used data from the AD 1305 Kaharoa eruption to assess the tephra-fall hazard from a future episode at Tarawera. This eruption consisted of a complex sequence of eruptive events with eleven discrete Plinian episodes, characterized by VEI 4. We developed an advection-diffusion model (TEPHRA) that includes grainsize-dependent diffusion law and particle density, a stratified atmosphere, the horizontal diffusion time of particles within the eruptive plume and settling velocities that account for Reynold's Number variations along the particle trajectory. Selected eruption parameters are sampled stochastically, possible outcomes are analyzed probabilistically and simulations are run in parallel on multiple processors. Given the fast computational times, TEPHRA is also characterized by a robust algorithm and high resolution and reliability of resulting outputs (i.e. hazard maps). Therefore, TEPHRA is an example of a class of numerical models that take advantage of new computational tools to forecast hazards as conditional probabilities far in advance of future eruptions. Three different scenarios were investigated on a probabilistic basis for a comprehensive tephra-fall hazard assessment: Upper Limit Scenario, Eruption-Range Scenario and Multiple-Eruption Scenario.

Research paper thumbnail of Global chemical weathering dominated by continental arcs since the mid-Palaeozoic

Research paper thumbnail of Tectonic forcing of global chemical weathering since the mid-Paleozoic

<p&amp... more <p>Weathering of the Earth’s surface has commonly been invoked as a driver of global cooling through geologic time. During the Phanerozoic Eon (541–0 million years ago, Ma), for example, the periodic onset of icehouse conditions has variously been attributed to enhanced weathering fluxes associated with mountain building (e.g. the Himalayas) (<em><strong>1</strong></em>), reductions in the global extent of continental arc volcanoes (e.g. the present-day Andes) (<em><strong>2</strong></em>), and uplift of oceanic crust during arc-continent collisions (e.g. present-day Indonesia and New Guinea) (<em><strong>3</strong></em>). These processes, tied to the global plate tectonic cycle, are inextricably linked.  The resulting collinearity (i.e. independent variables are highly correlated) makes it difficult — using conventional statistical techniques — to isolate the contributions of individual geologic processes to global chemical weathering.   An example of this is the Late Cenozoic Ice Age (34–0 Ma) that corresponds both to uplift of <span>the Tibetan Plateau and Himalaya,</span> and a gradual reduction in the extent of the global continental arc system. </p><p>We developed a machine learning framework to analyse the interdependencies between multiple global tectonic and volcanic processes (e.g., continental distribution, extent of volcanic arcs, mid-ocean ridges etc.) and seawater Sr composition (a proxy for weathering flux) over the past 400 million years. We developed a Bayesian network incorporating a novel algorithm that accounts for time lags for each of the predictor variables, and joint conditional dependence (i.e. how variables combine to influence the environmental outcome). Our approach overcomes problems traditionally encountered in geologic time series, such as collinearity and autocorrelation. Our results strongly indicate a first-order role for volcanism in driving chemical weathering fluxes since the mid-Palaeozoic. This is consistent with the strong empirical correlation previously observed between the strontium isotope composition of seawater and continental igneous rocks over the past billion years (<em><strong>4</strong></em>). Our study highlights how geologic processes operate together — not in isolation — to perturb the Earth system over ten to hundred million-year…

Research paper thumbnail of Oklahoma's induced seismicity strongly linked to wastewater injection depth

Science (New York, N.Y.), Mar 16, 2018

The sharp rise in Oklahoma seismicity since 2009 is due to wastewater injection. The role of inje... more The sharp rise in Oklahoma seismicity since 2009 is due to wastewater injection. The role of injection depth is an open, complex issue, yet critical for hazard assessment and regulation. We developed an advanced Bayesian network to model joint conditional dependencies between spatial, operational, and seismicity parameters. We found that injection depth relative to crystalline basement most strongly correlates with seismic moment release. The joint effects of depth and volume are critical, as injection rate becomes more influential near the basement interface. Restricting injection depths to 200 to 500 meters above basement could reduce annual seismic moment release by a factor of 1.4 to 2.8. Our approach enables identification of subregions where targeted regulation may mitigate effects of induced earthquakes, aiding operators and regulators in wastewater disposal regions.

Research paper thumbnail of Evidence-based real-time hazard evaluation methodology and software for EXPLORIS volcanoes

... Deliverable 6.4 Project number: EVR1-CT-2002-40026 Project Coordinator: Dr. Augusto Neri Deli... more ... Deliverable 6.4 Project number: EVR1-CT-2002-40026 Project Coordinator: Dr. Augusto Neri Deliverable D6.4 Evidence-based real-time hazard evaluation methodology and software for EXPLORIS volcanoes Thea Hincks, Willy Aspinall & Gordon Woo Aspinall & Associates ...

Research paper thumbnail of Bayesian Belief Networks for Hazard Forecasting on Montserrat

Research paper thumbnail of Estimating volcanic plume heights from depositional clast size

Journal of Geophysical Research, 2011

ABSTRACT Tephra deposits retain a considerable amount of information about the nature of volcanic... more ABSTRACT Tephra deposits retain a considerable amount of information about the nature of volcanic eruptions, with plume height commonly inferred from maximum clast size measurements. However, current methods for inferring plume height from maximum clast size lose some of the accuracy in measurements made in the field and have limited application when trying to constrain the uncertainty in these parameters. Here a predictive numerical model is used to determine plume height from maximum clast size found in a deposit. Plume height is an essential parameter for inferring eruption magnitude as it is explicitly related to mass eruption rate. We investigate the effects of different atmospheric and wind profiles, allowing the input conditions to be location specific for each eruption. The use of a predictive model reduces the uncertainty in determining plume height as it retains the detail of measurements made in the field and incorporates the uncertainty in these measurements in the results, in the form of a probability distribution. We have applied this approach to fall deposits from the 1991 eruption of Pinatubo, one of the few Plinian eruptions in which the true height of the plume is well known. The predicted plume heights are in good agreement with those found from satellite measurements. This new approach provides a method for determining plume height from poorly preserved fall deposits, from which a limited amount of data can be sampled leading to large uncertainties in the field measurements.

Research paper thumbnail of Probabilistic modeling of tephra dispersal: Hazard assessment of a multiphase rhyolitic eruption at Tarawera, New Zealand

Journal of Geophysical Research, 2005

1] The Tarawera Volcanic Complex comprises 11 rhyolite domes formed during five major eruptions b... more 1] The Tarawera Volcanic Complex comprises 11 rhyolite domes formed during five major eruptions between 17,000 B.C. and A.D. 1886, the first four of which were predominantly rhyolitic. The only historical event erupted about 2 km 3 of basaltic tephra fall (A.D. 1886). The youngest rhyolitic event erupted a tephra fall volume more than 2 times larger and covered a wider area northwest and southeast of the volcano ($A.D. 1315 Kaharoa eruption). We have used the Kaharoa scenario to assess the tephra fall hazard from a future rhyolitic eruption at Tarawera of a similar scale. The Plinian phase of this eruption consisted of 11 discrete episodes of VEI 4. We have developed an advection-diffusion model (TEPHRA) that allows for grain size-dependent diffusion and particle density, a stratified atmosphere, particle diffusion time within the rising plume, and settling velocities that include Reynolds number variations along the particle fall. Simulations are run in parallel on multiple processors to allow a significant implementation of the physical model and a fully probabilistic analysis of inputs and outputs. TEPHRA is an example of a class of numerical models that take advantage of new computational tools to forecast hazards as conditional probabilities far in advance of future eruptions. Three different scenarios were investigated for a comprehensive tephra fall hazard assessment: upper limit scenario, eruption range scenario, and multiple eruption scenario. Hazard curves and probability maps show that the area east and northeast of Tarawera would be the most affected by a Kaharoa-type eruption.

Research paper thumbnail of Snowball Earth ocean chemistry driven by extensive ridge volcanism during Rodinia breakup

Research paper thumbnail of Improving crisis decision-making at times of uncertain volcanic unrest (Guadeloupe, 1976)

Global Volcanic Hazards and Risk, 2015

Research paper thumbnail of Depositional processes of the basaltic Elie Ness diatreme, East Fife, Scotland

The East Fife coast of Scotland exposes multiple (~100) volcanic vents or diatremes of late Carbo... more The East Fife coast of Scotland exposes multiple (~100) volcanic vents or diatremes of late Carboniferous to early Permian age. Here, we present preliminary results of detailed geological mapping of the Elie Ness (EN) diatreme. The key objective was to map the volcanic structure and lithofacies of the vent-fill, and to determine the eruption styles and key emplacement processes that occur more generally in basaltic maar-diatreme systems. Within the EN diatreme, seven lithofacies and three lithofacies associations (LFA 1-3) were recognised. Preliminary results demonstrate that the diatreme had a protracted history of eruption and infill. The massive lapilli tuffs of LFA 1 are texturally and compositionally homogeneous with occasional degassing structures, making them similar to typical massive volcaniclastic deposits infilling kimberlite pipes. The formation of such deposits are attributed to gas-fluidisation processes operating within the vent. The occurrence within LFA 1 of abundan...

Research paper thumbnail of Retrospective analysis of uncertain eruption precursors at La Soufrière volcano, Guadeloupe, 1975–77: volcanic hazard assessment using a Bayesian Belief Network approach

Journal of Applied Volcanology, 2014

Research paper thumbnail of Risk assessment for silicosis and exposure to volcanic ash on Montserrat

Research paper thumbnail of Using Bayesian Belief Networks and event trees for volcanic hazard assessment and decision support: reconstruction of past eruptions of La Soufrière volcano, Guadeloupe and retrospective analysis of 1975-77 unrest

ABSTRACT Since 1992, mild but persistent seismic and fumarolic unrest at La Soufrière de Guadelou... more ABSTRACT Since 1992, mild but persistent seismic and fumarolic unrest at La Soufrière de Guadeloupe volcano has prompted renewed concern about hazards and risks, crisis response planning, and has rejuvenated interest in geological studies. Scientists monitoring active volcanoes frequently have to provide science-based decision support to civil authorities during such periods of unrest. In these circumstances, the Bayesian Belief Network (BBN) offers a formalized evidence analysis tool for making inferences about the state of the volcano from different strands of data, allowing associated uncertainties to be treated in a rational and auditable manner, to the extent warranted by the strength of the evidence. To illustrate the principles of the BBN approach, a retrospective analysis is undertaken of the 1975-77 crisis, providing an inferential assessment of the evolving state of the magmatic system and the probability of subsequent eruption. Conditional dependencies and parameters in the BBN are characterized quantitatively by structured expert elicitation. Revisiting data available in 1976 suggests the probability of magmatic intrusion would have been evaluated high at the time, according with subsequent thinking about the volcanological nature of the episode. The corresponding probability of a magmatic eruption therefore would have been elevated in July and August 1976; however, collective uncertainty about the future course of the crisis was great at the time, even if some individual opinions were certain. From this BBN analysis, while the more likely appraised outcome - based on observational trends at 31 August 1976 - might have been 'no eruption' (mean probability 0.5; 5-95 percentile range 0.8), an imminent magmatic eruption (or blast) could have had a probability of about 0.4, almost as substantial. Thus, there was no real scientific basis to assert one scenario was more likely than the other. This retrospective evaluation adds objective probabilistic expression to the contemporary volcanological narrative, and demonstrates that a formal evidential case could have been made to support the authorities' concerns and decision to evacuate. Revisiting the circumstances of the 1976 crisis highlights many contemporary challenges of decision-making under conditions of volcanological uncertainty. We suggest the BBN concept is a suitable framework for marshalling multiple observations, model results and interpretations - and all associated uncertainties - in a methodical manner. Base-rate eruption probabilities for Guadeloupe can be updated now with a new chronology of activity suggesting that 10 major explosive phases and 9 dome-forming phases occurred in the last 9150 years, associated with ≥ 8 flank-collapses and ≥ 6-7 high-energy pyroclastic density currents (blasts). Eruptive recurrence, magnitude and intensity place quantitative constraints on La Soufrière's event tree to elaborate credible scenarios. The current unrest offers an opportunity to update the BBN model and explore the uncertainty on inferences about the system's internal state. This probabilistic formalism would provoke key questions relating to unrest evolution: 1) is the unrest hydrothermal or magmatic? 2) what controls dyke/intrusion arrest and hence failed-magmatic eruptions like 1976? 3) what conditions could lead to significant pressurization with potential for explosive activity and edifice instability, and what monitoring signs might be manifest?

Research paper thumbnail of Advancing volcanology through analyses of large datasets: insights from the LaMEVE database

Research paper thumbnail of Risk assessment and management of wildfires

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Risk and uncertainty assessment of volcanic hazards

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Tsunami hazard and risk

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Landslide and avalanche hazards

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Probabilistic Modeling of Tephra Dispersion using Parallel Processing

Numerical models of tephra accumulation are important tools in assessing hazards of volcanic erup... more Numerical models of tephra accumulation are important tools in assessing hazards of volcanic eruptions. Such tools can be used far in advance of future eruptions to calculate possible hazards as conditional probabilities. For example, given that a volcanic eruption occurs, what is the expected range of tephra deposition in a specific location or across a region? An empirical model is

Research paper thumbnail of Probabilistic modeling of tephra dispersion : hazard assessment of a multi-phase eruption at Tarawera Volcano , New Zealand

The Tarawera Volcanic Complex (New Zealand) comprises 11 rhyolite domes formed during five major ... more The Tarawera Volcanic Complex (New Zealand) comprises 11 rhyolite domes formed during five major eruptions between 17000 BC and AD 1886, the first four of which were predominantly rhyolitic. The only historical event (AD 1886) erupted about 2 km 3 of purely basaltic tephra fall killing about 150 people. The AD 1305 Kaharoa eruption is the youngest rhyolitic event and erupted a total tephra-fall volume more than two times larger than the AD 1886 eruption. We used data from the AD 1305 Kaharoa eruption to assess the tephra-fall hazard from a future episode at Tarawera. This eruption consisted of a complex sequence of eruptive events with eleven discrete Plinian episodes, characterized by VEI 4. We developed an advection-diffusion model (TEPHRA) that includes grainsize-dependent diffusion law and particle density, a stratified atmosphere, the horizontal diffusion time of particles within the eruptive plume and settling velocities that account for Reynold's Number variations along the particle trajectory. Selected eruption parameters are sampled stochastically, possible outcomes are analyzed probabilistically and simulations are run in parallel on multiple processors. Given the fast computational times, TEPHRA is also characterized by a robust algorithm and high resolution and reliability of resulting outputs (i.e. hazard maps). Therefore, TEPHRA is an example of a class of numerical models that take advantage of new computational tools to forecast hazards as conditional probabilities far in advance of future eruptions. Three different scenarios were investigated on a probabilistic basis for a comprehensive tephra-fall hazard assessment: Upper Limit Scenario, Eruption-Range Scenario and Multiple-Eruption Scenario.

Research paper thumbnail of Global chemical weathering dominated by continental arcs since the mid-Palaeozoic

Research paper thumbnail of Tectonic forcing of global chemical weathering since the mid-Paleozoic

<p&amp... more <p>Weathering of the Earth’s surface has commonly been invoked as a driver of global cooling through geologic time. During the Phanerozoic Eon (541–0 million years ago, Ma), for example, the periodic onset of icehouse conditions has variously been attributed to enhanced weathering fluxes associated with mountain building (e.g. the Himalayas) (<em><strong>1</strong></em>), reductions in the global extent of continental arc volcanoes (e.g. the present-day Andes) (<em><strong>2</strong></em>), and uplift of oceanic crust during arc-continent collisions (e.g. present-day Indonesia and New Guinea) (<em><strong>3</strong></em>). These processes, tied to the global plate tectonic cycle, are inextricably linked.  The resulting collinearity (i.e. independent variables are highly correlated) makes it difficult — using conventional statistical techniques — to isolate the contributions of individual geologic processes to global chemical weathering.   An example of this is the Late Cenozoic Ice Age (34–0 Ma) that corresponds both to uplift of <span>the Tibetan Plateau and Himalaya,</span> and a gradual reduction in the extent of the global continental arc system. </p><p>We developed a machine learning framework to analyse the interdependencies between multiple global tectonic and volcanic processes (e.g., continental distribution, extent of volcanic arcs, mid-ocean ridges etc.) and seawater Sr composition (a proxy for weathering flux) over the past 400 million years. We developed a Bayesian network incorporating a novel algorithm that accounts for time lags for each of the predictor variables, and joint conditional dependence (i.e. how variables combine to influence the environmental outcome). Our approach overcomes problems traditionally encountered in geologic time series, such as collinearity and autocorrelation. Our results strongly indicate a first-order role for volcanism in driving chemical weathering fluxes since the mid-Palaeozoic. This is consistent with the strong empirical correlation previously observed between the strontium isotope composition of seawater and continental igneous rocks over the past billion years (<em><strong>4</strong></em>). Our study highlights how geologic processes operate together — not in isolation — to perturb the Earth system over ten to hundred million-year…

Research paper thumbnail of Oklahoma's induced seismicity strongly linked to wastewater injection depth

Science (New York, N.Y.), Mar 16, 2018

The sharp rise in Oklahoma seismicity since 2009 is due to wastewater injection. The role of inje... more The sharp rise in Oklahoma seismicity since 2009 is due to wastewater injection. The role of injection depth is an open, complex issue, yet critical for hazard assessment and regulation. We developed an advanced Bayesian network to model joint conditional dependencies between spatial, operational, and seismicity parameters. We found that injection depth relative to crystalline basement most strongly correlates with seismic moment release. The joint effects of depth and volume are critical, as injection rate becomes more influential near the basement interface. Restricting injection depths to 200 to 500 meters above basement could reduce annual seismic moment release by a factor of 1.4 to 2.8. Our approach enables identification of subregions where targeted regulation may mitigate effects of induced earthquakes, aiding operators and regulators in wastewater disposal regions.

Research paper thumbnail of Evidence-based real-time hazard evaluation methodology and software for EXPLORIS volcanoes

... Deliverable 6.4 Project number: EVR1-CT-2002-40026 Project Coordinator: Dr. Augusto Neri Deli... more ... Deliverable 6.4 Project number: EVR1-CT-2002-40026 Project Coordinator: Dr. Augusto Neri Deliverable D6.4 Evidence-based real-time hazard evaluation methodology and software for EXPLORIS volcanoes Thea Hincks, Willy Aspinall & Gordon Woo Aspinall & Associates ...

Research paper thumbnail of Bayesian Belief Networks for Hazard Forecasting on Montserrat

Research paper thumbnail of Estimating volcanic plume heights from depositional clast size

Journal of Geophysical Research, 2011

ABSTRACT Tephra deposits retain a considerable amount of information about the nature of volcanic... more ABSTRACT Tephra deposits retain a considerable amount of information about the nature of volcanic eruptions, with plume height commonly inferred from maximum clast size measurements. However, current methods for inferring plume height from maximum clast size lose some of the accuracy in measurements made in the field and have limited application when trying to constrain the uncertainty in these parameters. Here a predictive numerical model is used to determine plume height from maximum clast size found in a deposit. Plume height is an essential parameter for inferring eruption magnitude as it is explicitly related to mass eruption rate. We investigate the effects of different atmospheric and wind profiles, allowing the input conditions to be location specific for each eruption. The use of a predictive model reduces the uncertainty in determining plume height as it retains the detail of measurements made in the field and incorporates the uncertainty in these measurements in the results, in the form of a probability distribution. We have applied this approach to fall deposits from the 1991 eruption of Pinatubo, one of the few Plinian eruptions in which the true height of the plume is well known. The predicted plume heights are in good agreement with those found from satellite measurements. This new approach provides a method for determining plume height from poorly preserved fall deposits, from which a limited amount of data can be sampled leading to large uncertainties in the field measurements.

Research paper thumbnail of Probabilistic modeling of tephra dispersal: Hazard assessment of a multiphase rhyolitic eruption at Tarawera, New Zealand

Journal of Geophysical Research, 2005

1] The Tarawera Volcanic Complex comprises 11 rhyolite domes formed during five major eruptions b... more 1] The Tarawera Volcanic Complex comprises 11 rhyolite domes formed during five major eruptions between 17,000 B.C. and A.D. 1886, the first four of which were predominantly rhyolitic. The only historical event erupted about 2 km 3 of basaltic tephra fall (A.D. 1886). The youngest rhyolitic event erupted a tephra fall volume more than 2 times larger and covered a wider area northwest and southeast of the volcano ($A.D. 1315 Kaharoa eruption). We have used the Kaharoa scenario to assess the tephra fall hazard from a future rhyolitic eruption at Tarawera of a similar scale. The Plinian phase of this eruption consisted of 11 discrete episodes of VEI 4. We have developed an advection-diffusion model (TEPHRA) that allows for grain size-dependent diffusion and particle density, a stratified atmosphere, particle diffusion time within the rising plume, and settling velocities that include Reynolds number variations along the particle fall. Simulations are run in parallel on multiple processors to allow a significant implementation of the physical model and a fully probabilistic analysis of inputs and outputs. TEPHRA is an example of a class of numerical models that take advantage of new computational tools to forecast hazards as conditional probabilities far in advance of future eruptions. Three different scenarios were investigated for a comprehensive tephra fall hazard assessment: upper limit scenario, eruption range scenario, and multiple eruption scenario. Hazard curves and probability maps show that the area east and northeast of Tarawera would be the most affected by a Kaharoa-type eruption.

Research paper thumbnail of Snowball Earth ocean chemistry driven by extensive ridge volcanism during Rodinia breakup

Research paper thumbnail of Improving crisis decision-making at times of uncertain volcanic unrest (Guadeloupe, 1976)

Global Volcanic Hazards and Risk, 2015

Research paper thumbnail of Depositional processes of the basaltic Elie Ness diatreme, East Fife, Scotland

The East Fife coast of Scotland exposes multiple (~100) volcanic vents or diatremes of late Carbo... more The East Fife coast of Scotland exposes multiple (~100) volcanic vents or diatremes of late Carboniferous to early Permian age. Here, we present preliminary results of detailed geological mapping of the Elie Ness (EN) diatreme. The key objective was to map the volcanic structure and lithofacies of the vent-fill, and to determine the eruption styles and key emplacement processes that occur more generally in basaltic maar-diatreme systems. Within the EN diatreme, seven lithofacies and three lithofacies associations (LFA 1-3) were recognised. Preliminary results demonstrate that the diatreme had a protracted history of eruption and infill. The massive lapilli tuffs of LFA 1 are texturally and compositionally homogeneous with occasional degassing structures, making them similar to typical massive volcaniclastic deposits infilling kimberlite pipes. The formation of such deposits are attributed to gas-fluidisation processes operating within the vent. The occurrence within LFA 1 of abundan...

Research paper thumbnail of Retrospective analysis of uncertain eruption precursors at La Soufrière volcano, Guadeloupe, 1975–77: volcanic hazard assessment using a Bayesian Belief Network approach

Journal of Applied Volcanology, 2014

Research paper thumbnail of Risk assessment for silicosis and exposure to volcanic ash on Montserrat

Research paper thumbnail of Using Bayesian Belief Networks and event trees for volcanic hazard assessment and decision support: reconstruction of past eruptions of La Soufrière volcano, Guadeloupe and retrospective analysis of 1975-77 unrest

ABSTRACT Since 1992, mild but persistent seismic and fumarolic unrest at La Soufrière de Guadelou... more ABSTRACT Since 1992, mild but persistent seismic and fumarolic unrest at La Soufrière de Guadeloupe volcano has prompted renewed concern about hazards and risks, crisis response planning, and has rejuvenated interest in geological studies. Scientists monitoring active volcanoes frequently have to provide science-based decision support to civil authorities during such periods of unrest. In these circumstances, the Bayesian Belief Network (BBN) offers a formalized evidence analysis tool for making inferences about the state of the volcano from different strands of data, allowing associated uncertainties to be treated in a rational and auditable manner, to the extent warranted by the strength of the evidence. To illustrate the principles of the BBN approach, a retrospective analysis is undertaken of the 1975-77 crisis, providing an inferential assessment of the evolving state of the magmatic system and the probability of subsequent eruption. Conditional dependencies and parameters in the BBN are characterized quantitatively by structured expert elicitation. Revisiting data available in 1976 suggests the probability of magmatic intrusion would have been evaluated high at the time, according with subsequent thinking about the volcanological nature of the episode. The corresponding probability of a magmatic eruption therefore would have been elevated in July and August 1976; however, collective uncertainty about the future course of the crisis was great at the time, even if some individual opinions were certain. From this BBN analysis, while the more likely appraised outcome - based on observational trends at 31 August 1976 - might have been 'no eruption' (mean probability 0.5; 5-95 percentile range 0.8), an imminent magmatic eruption (or blast) could have had a probability of about 0.4, almost as substantial. Thus, there was no real scientific basis to assert one scenario was more likely than the other. This retrospective evaluation adds objective probabilistic expression to the contemporary volcanological narrative, and demonstrates that a formal evidential case could have been made to support the authorities' concerns and decision to evacuate. Revisiting the circumstances of the 1976 crisis highlights many contemporary challenges of decision-making under conditions of volcanological uncertainty. We suggest the BBN concept is a suitable framework for marshalling multiple observations, model results and interpretations - and all associated uncertainties - in a methodical manner. Base-rate eruption probabilities for Guadeloupe can be updated now with a new chronology of activity suggesting that 10 major explosive phases and 9 dome-forming phases occurred in the last 9150 years, associated with ≥ 8 flank-collapses and ≥ 6-7 high-energy pyroclastic density currents (blasts). Eruptive recurrence, magnitude and intensity place quantitative constraints on La Soufrière's event tree to elaborate credible scenarios. The current unrest offers an opportunity to update the BBN model and explore the uncertainty on inferences about the system's internal state. This probabilistic formalism would provoke key questions relating to unrest evolution: 1) is the unrest hydrothermal or magmatic? 2) what controls dyke/intrusion arrest and hence failed-magmatic eruptions like 1976? 3) what conditions could lead to significant pressurization with potential for explosive activity and edifice instability, and what monitoring signs might be manifest?

Research paper thumbnail of Advancing volcanology through analyses of large datasets: insights from the LaMEVE database

Research paper thumbnail of Risk assessment and management of wildfires

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Risk and uncertainty assessment of volcanic hazards

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Tsunami hazard and risk

Risk and Uncertainty Assessment for Natural Hazards, 2013

Research paper thumbnail of Landslide and avalanche hazards

Risk and Uncertainty Assessment for Natural Hazards, 2013