Science , — Huppert, H. The role of volatiles in magma chamber dynamics. Scaillet, B. Upward migration of Vesuvius magma chamber over the past 20, years. Owen, J. Explosive subglacial rhyolitic eruptions in Iceland are fuelled by high magmatic H2O and closed-system degassing.
Geology 41 , — Nishimura, K. Melt inclusion analysis of the Unzen — dacite: implications for crystallization processes of dacite magma. Edmonds, M. Pre-eruptive vapour and its role in controlling eruption style and longevity at Soufriere Hills Volcano. Plank, T. Kent, A.
Melt inclusions in basaltic and related volcanic rocks. Gaetani, G. Rapid reequilibration of H2O and oxygen fugacity in olivine-hosted melt inclusions.
Geology 40 , — Wallace, P. Melt inclusion CO2 contents, pressures of olivine crystallization, and the problem of shrinkage bubbles.
Experimental constraints on parameters controlling the difference in the eruptive dynamics of phonolitic magmas: the case of tenerife Canary islands. Non-explosive silicic volcanism. Gonnermann, H. The fluid mechanics inside a volcano. Fluid Mech. Woods, A. Transitions between explosive and effusive eruptions of silicic magmas. Martel, C. Timescales of bubble coalescence, outgassing, and foam collapse in decompressed rhyolitic melts. Scandone, R. Magma supply, magma ascent and the style of volcanic eruptions.
Sparks, R. Nonlinear dynamics of lava dome extrusion. Nature , 37—41 Tarasewicz, J. Magma mobilization by downward-propagating decompression of the Eyjafjallajkull volcanic plumbing system. Castro, J. Schipper, C. Shallow vent architecture during hybrid explosive-effusive activity at Cordon Caulle Chile, : evidence from direct observations and pyroclast textures.
Thomas, M. Understanding which parameters control shallow ascent of silicic effusive magma. Stevenson, J. Mangan, M. Decompression experiments identify kinetic controls on explosive silicic eruptions. Delayed, disequilibrium degassing in rhyolite magma: decompression experiments and implications for explosive volcanism.
Mourtada-Bonnefoi, C. Experimental study of homogeneous bubble nucleation in rhyolitic magmas. Shea, T.
Bubble nucleation in magmas: a dominantly heterogeneous process? La Spina, G. Did magma ascent rate control the explosive-effusive transition at the Inyo volcanic chain, California? Geology 36 , — Endo, E.
The Eruptions of Mount St. Helens, Washington 93—, professional paper U. Alidibirov, M. Magma fragmentation by rapid decompression. Dingwell, D. Volcanic dilemma: flow or blow? Mueller, S. Permeability control on magma fragmentation. Geology 36 , Horwell, C. Kendrick, J. Blowing off steam: tuffisite formation as a regulator for lava dome eruptions. Earth Sci. Boudon, G. What factors control superficial lava dome explosivity? A model of vulcanian eruptions at Galeras volcano, Colombia.
Barmin, A. Periodic behavior in lava dome eruptions. Wright, H. Breadcrust bombs as indicators of Vulcanian eruption dynamics at Guagua Pichincha volcano, Ecuador.
Druitt, T. Rapid and slow: varying magma ascent rates as a mechanism for Vulcanian explosions. Deegan, F. Magma—carbonate interaction processes and associated CO2 release at Merapi volcano, Indonesia: insights from experimental petrology. Troll, V. Crustal CO 2 liberation during the eruption and earthquake events at Merapi volcano, Indonesia.
Zimanowski, B. Analysis of thermohydraulic explosion energetics. The dynamics of bubble formation and growth in magmas: a review and analysis. Thermal vesiculation during volcanic eruptions. Zhang, Y. A criterion for the fragmentation of bubbly magma based on brittle failure theory. Jaupart, C. Gas content, eruption rate and instabilities of eruption regime in silicic volcanoes. Saar, M. Permeability-porosity relationship in vesicular basalts.
Permeability and degassing of dome lavas undergoing rapid decompression: an experimental determination. Takeuchi, S. Permeability measurements of natural and experimental volcanic materials with a simple permeameter: toward an understanding of magmatic degassing processes.
Okumura, S. Shear deformation experiments on vesicular rhyolite: implications for brittle fracturing, degassing, and compaction of magmas in volcanic conduits.
Solid Earth , B Farquharson, J. Permeability and porosity relationships of edifice-forming andesites: a combined field and laboratory study. Colombier, M. The evolution of pore connectivity in volcanic rocks. Rust, A. Permeability of vesicular silicic magma: inertial and hysteresis effects. Magma deformation may induce non-explosive volcanism via degassing through bubble networks.
Mechanisms of bubble coalescence in silicic magmas. The effects of outgassing on the transition between effusive and explosive silicic eruptions.
Lindoo, A. An experimental study of permeability development as a function of crystal-free melt viscosity. Crystal controls on permeability development and degassing in basaltic andesite magma.
Geology 45 , — Textural studies of vesicles in volcanic rocks: an integrated methodology. The percolation threshold and permeability evolution of ascending magmas. Permeability controls on expansion and size distributions of pyroclasts. Solid Earth , 1—17 Degassing conditions for permeable silicic magmas: implications from decompression experiments with constant rates.
Nguyen, C. Explosive to effusive transition during the largest volcanic eruption of the 20th century Novarupta , Alaska. Geology 42 , — Kameda, M. Advancement of magma fragmentation by inhomogeneous bubble distribution. Yoshimura, S. Fracture healing in a magma: an experimental approach and implications for volcanic seismicity and degassing. ADS Google Scholar. Cabrera, A. Melt fracturing and healing: a mechanism for degassing and origin of silicic obsidian.
Geology 39 , 67—70 Permeability reduction of fractured rhyolite in volcanic conduits and its control on eruption cyclicity. Kushnir, A. In situ confirmation of permeability development in shearing bubble-bearing melts and implications for volcanic outgassing.
Stasiuk, M. Tuffen, H. Repeated fracture and healing of silicic magma generate flow banding and earthquakes? Geology 31 , — Kolzenburg, S. Strength and permeability recovery of tuffisite-bearing andesite.
Solid Earth 3 , — Fault textures in volcanic conduits: evidence for seismic trigger mechanisms during silicic eruptions. Flow banding in obsidian: a record of evolving textural heterogeneity during magma deformation. Evidence for the development of permeability anisotropy in lava domes and volcanic conduits. Black, B. Ash production and dispersal from sustained low-intensity Mono-Inyo eruptions. Helens, Washington.
A Volcano Rekindled Renewed Erupt. Helens, — Professional paper U. Geological Survey, Reston, VA, Merapi eruption-chronology and extrusion rates monitored with satellite radar and used in eruption forecasting. Explosive volcanism may not be an inevitable consequence of magma fragmentation. Hale, A. The influence of viscous and latent heating on crystal-rich magma flow in a conduit. De Angelis, S. On the feasibility of magma fracture within volcanic conduits: constraints from earthquake data and empirical modelling of magma viscosity.
Holland, A. Degassing processes during lava dome growth: Insights from Santiaguito lava dome, Guatemala. Volcanic drumbeat seismicity caused by stick-slip motion and magmatic frictional melting. Heap, M. From rock to magma and back again: the evolution of temperature and deformation mechanism in conduit margin zones. Multiple origins of obsidian pyroclasts and implications for changes in the dynamics of the B.
Cordonnier, B. The viscous-brittle transition of crystal-bearing silicic melt: direct observation of magma rupture and healing. Gardner, J. Formation of obsidian pyroclasts by sintering of ash particles in the volcanic conduit. Vasseur, J. Volcanic sintering: timescales of viscous densification and strength recovery.
Wadsworth, F. Nonisothermal viscous sintering of volcanic ash. Solid Earth , — Collinson, A. Gas storage, transport and pressure changes in an evolving permeable volcanic edifice. Kennedy, B. Time-and temperature-dependent conduit wall porosity: a key control on degassing and explosivity at Tarawera volcano, New Zealand.
Schneider, A. Conduit degassing and thermal controls on eruption styles at Mount St. Lamb, O. Ball, J. The hydrothermal alteration of cooling lava domes.
Effects of conduit geometry on magma ascent dynamics in dome-forming eruptions. Transient effects of magma ascent dynamics along a geometrically variable dome-feeding conduit. Costa, A. Effects of wall-rock elasticity on magma flow in dykes during explosive eruptions.
Stability of volcanic conduits during explosive eruptions. Bachmann, O. Voluminous lava-like precursor to a major ash-flow tuff: low-column pyroclastic eruption of the Pagosa Peak Dacite, San Juan volcanic field, Colorado. Michel, J. Incremental growth of the Patagonian Torres del Paine laccolith over 90 k. Strain localization in vesicular magma: Implications for rheology and fragmentation.
Geology 37 , — Rheological transitions in high-temperature volcanic fault zones. Pore-scale mass and reactant transport in multiphase porous media flows. Oppenheimer, J.
Gas migration regimes and outgassing in particle-rich suspensions. Pistone, M. In situ X-ray tomographic microscopy observations of vesiculation of bubble-free and bubble-bearing magmas. Crystal-rich lava dome extrusion during vesiculation: an experimental study. The mechanics of shallow magma reservoir outgassing. Causes and consequences of pressurisation in lava dome eruptions. New geochemical insights into volcanic degassing.
A Math. Spieler, O. Magma fragmentation speed: an experimental determination. Wadge, G. An overview of the eruption of Soufriere Hills Volcano, Montserrat from to Google Scholar. Cioni, R. The impact of pre-existing gas on the ascent of explosively erupted magma. Winson, A. Is there a recipe for Plinian eruptions? Evidence from past events and analogous volcanoes. Union Fall Meet. V41B Time-dependent permeability evolution in compacting volcanic fracture systems and implications for gas overpressure.
What makes a volcano tick-A first explanation of deep multiple seismic sources in ascending magma. Bato, M. Assimilation of deformation data for eruption forecasting: potentiality assessment based on synthetic cases. Surono et al. Moor, J. Turmoil at Turrialba volcano Costa Rica : degassing and eruptive processes inferred from high-frequency gas monitoring.
Solid Earth 1— White, R. Volcano-tectonic earthquakes: a new tool for estimating intrusive volumes and forecasting eruptions. McNutt, S. Voight, B. A method for prediction of volcanic eruptions. Kilburn, C. Multiscale fracturing as a key to forecasting volcanic eruptions. Chastin, S. Statistical analysis of daily seismic event rate as a precursor to volcanic eruptions. Heterogeneity: the key to failure forecasting. Bell, A. Forecasting volcanic eruptions and other material failure phenomena: an evaluation of the failure forecast method.
Segall, P. Volcano deformation and eruption forecasting. Wooster, M. Discrimination of lava dome activity styles using satellite-derived thermal structures. Diefenbach, A. Photogrammetric monitoring of lava dome growth during the eruption of Redoubt Volcano.
Christopher, T. Saunders, K. Linking petrology and seismology at an active volcano. Jay, J. Heiken, G. Fracture fillings and intrusive pyroclasts, Inyo Domes, California. Solid Earth 93 , — Along divergent plate boundaries, such as Oceanic Ridges or spreading centers. In areas of continental extension that may become divergent plate boundaries in the future. Along converging plate boundaries where subduction is occurring.
And, in areas called "hot spots" that are usually located in the interior of plates, away from the plate margins. Since we discussed this in the lecture on igneous rocks, we only briefly review this material here. Active volcanism is currently taking place along all of oceanic ridges, but most of this volcanism is submarine volcanism.
One place where an oceanic ridge reaches above sea level is at Iceland, along the Mid-Atlantic Ridge. Here, most eruptions are basaltic in nature, but, many are explosive strombolian types or explosive phreatic or phreatomagmatic types. As seen in the map to the right, the Mid-Atlantic ridge runs directly through Iceland.
Volcanism also occurs in continental areas that are undergoing episodes of rifting. The extensional deformation occurs because the underlying mantle is rising from below and stretching the overlying continental crust.
Upwelling mantle may melt to produce magmas, which then rise to the surface, often along normal faults produced by the extensional deformation.
Basaltic and rhyolitic volcanism is common in these areas. In the same area, the crust has rifted apart along the Red Sea, and the Gulf of Aden to form new oceanic ridges. This may also be the fate of the East African Rift Valley at some time in the future. Other areas where extensional deformation is occurring within the crust is Basin and Range Province of the western U.
These are also areas of recent basaltic and rhyolitic volcanism. All around the Pacific Ocean, is a zone often referred to as the Pacific Ring of Fire, where most of the world's most active and most dangerous volcanoes occur.
The Ring of Fire occurs because most of the margins of the Pacific ocean coincide with converging margins along which subduction is occurring. These are all island arcs. The Hawaiian Ridge is one such hot spot trace. Here the Big Island of Hawaii is currently over the hot spot, the other Hawaiian islands still stand above sea level, but volcanism has ceased. Northwest of the Hawaiian Islands, the volcanoes have eroded and are now seamounts.
Plateau or Flood basalts are extremely large volume outpourings of low viscosity basaltic magma from fissure vents. The basalts spread huge areas of relatively low slope and build up plateaus. Many of these outpourings appear to have occurred along a zone that eventually developed into a rift valley and later into a diverging plate boundary. Examples of questions on this material that could be asked on an exam.
Physical Geology. Volcanoes and Volcanic Eruptions. Magmas and Lava Since volcanic eruptions are caused by magma a mixture of liquid rock, crystals, and dissolved gas expelled onto the Earth's surface, we'll first review the characteristics of magma that we covered previously.
Viscosity of Magmas Viscosity is the resistance to flow opposite of fluidity. Higher SiO 2 content magmas have higher viscosity than lower SiO 2 content magmas Lower Temperature magmas have higher viscosity than higher temperature magmas.
Solidified Volcanic Rock. Solidified Plutonic Rock. Intermediate or Andesitic. Pahoehoe Flows - Basaltic lava flows with low viscosity start to cool when exposed to the low temperature of the atmosphere. This causes a surface skin to form, although it is still very hot and behaves in a plastic fashion, capable of deformation. Such lava flows that initially have a smooth surface are called pahoehoe flows.
Initially the surface skin is smooth, but often inflates with molten lava and expands to form pahoehoe toes or rolls to form ropey pahoehoe. See figure 9. Pahoehoe flows tend to be thin and, because of their low viscosity travel long distances from the vent.
A'A' Flows - Higher viscosity basaltic and andesitic lavas also initially develop a smooth surface skin, but this is quickly broken up by flow of the molten lava within and by gases that continue to escape from the lava.
This creates a rough, clinkery surface that is characteristic of an A'A' flow see figure 9. Lava Tubes - Once the surface skin becomes solid, the lava can continue to flow beneath the surface in lava tubes. The surface skin insulates the hot liquid lava form further cooling. When the eruption ends, liquid lava often drains leaving an open cave see figure 9.
Pillow Lavas - When lava erupts on the sea floor or other body of water, the surface skin forms rapidly, and, like with pahoehoe toes inflates with molten lava. Eventually these inflated balloons of magma drop off and stack up like a pile of pillows and are called pillow lavas. Ancient pillow lavas are readily recognizable because of their shape, their glassy margins and radial fractures that formed during cooling see figure 9.
Columnar Jointing - When thick basaltic or andesitic lavas cool, they contract. The contraction results in fractures and often times results in a type of jointing called columnar jointing. The columns are usually hexagonal in shape. This often happens when lavas pool in depressions or deep canyons see figure 9. Lava Domes or Volcanic Domes - result from the extrusion of highly viscous, gas poor andesitic and rhyolitic lava.
Since the viscosity is so high, the lava does not flow away from the vent, but instead piles up over the vent. Blocks of nearly solid lava break off the outer surface of the dome and roll down its flanks to form a breccia around the margins of domes. Pyroclastic Material If the magma has high gas content and high viscosity, the gas will expand in an explosive fashion and break the liquid into clots that fly through the air and cool along their path through the atmosphere.
Blocks are angular fragments that were solid when ejected. Volcanic Landforms Volcanic landforms are controlled by the geological processes that form them and act on them after they have formed.
Shield Volcanoes A shield volcano is characterized by gentle upper slopes about 5 o and somewhat steeper lower slopes about 10 o. Most shield volcanoes have a roughly circular or oval shape in map view.
Long periods of repose times of inactivity lasting for hundreds to thousands of years, make this type of volcano particularly dangerous, since many times they have shown no historic activity, and people are reluctant to heed warnings about possible eruptions. Cinder Cones Cinder cones are small volume cones consisting predominantly of ash and scoria that result from mildly explosive eruptions.
They usually consist of basaltic to andesitic material. They are actually fall deposits that are built surrounding the eruptive vent. Slopes of the cones are controlled by the angle of repose angle of stable slope for loose unconsolidated material and are usually between about 25 and 35 o.
On young cones, a depression at the top of the cone, called a crater, is evident, and represents the area above the vent from which material was explosively ejected. Craters are usually eroded away on older cones. Craters and Calderas Craters are circular depressions, usually less than 1 km in diameter, that form as a result of explosions that emit gases and ash.
Calderas are much larger depressions, circular to elliptical in shape, with diameters ranging from 1 km to 50 km. Calderas form as a result of collapse of a volcanic structure. The collapse results from evacuation of the underlying magma chamber. Crater Lake Caldera in southern Oregon is an 8 km diameter caldera containing a lake The caldera formed about years ago as a result of the eruption of about 75 km 3 of rhyolite magma in the form of tephra, found as far away as Canada, accompanied by pyroclastic flows that left thick deposits of tuff on the flanks of the volcano.
Subsequent eruptions have built a cinder cone on the floor of the caldera, which now forms an island called Wizard Island.
Larger calderas have formed within the past million years in the western United States. The Yellowstone caldera is an important example, as it illustrates the amount of repose time that might be expected from large rhyolitic systems, and the devastating effect caldera forming eruptions can have on widespread areas.
Yellowstone Caldera which occupies most of Yellowstone National Park, is actually the third caldera to form in the area within the past 2 million years. The three calderas formed at 2. Thus the repose time is on the average about , years. Tephra fall deposits from the latest eruption are found in Louisiana and into the Gulf of Mexico, and covered much of the Western part of the United States.
The eruption , years ago produced about km 3 of rhyolite in comparison, the eruption of Mt. Helens in May of produced only 0. Magma still underlies Yellowstone caldera, as evidenced by the large number of hot springs and geysers in the area. Volcanic Eruptions In general, magmas that are generated deep within the Earth begin to rise because they are less dense than the surrounding solid rocks.
When the magma reaches the Earth's surface, the gas bubble will simply burst, the gas will easily expand to atmospheric pressure, and a effusive or non-explosive eruption will occur, usually as a lava flow If the liquid part of the magma has a high viscosity, then the gas will not be able to expand very easily, and thus, pressure will build up inside of the gas bubble s. Effusive Eruptions Effusive or Non explosive eruptions are favored by low gas content and low viscosity magmas basaltic to andesitic magmas.
If the viscosity is low, non-explosive eruptions usually begin with fire fountains due to release of dissolved gases. Lava flows are produced on the surface, and these run like liquids down slope, along the lowest areas they can find.
If the magma emerges along a fracture, it results in a fissure eruption, often called a "curtain of fire" Lava flows produced by eruptions under water are called pillow lavas. If the viscosity is high, but the gas content is low, then the lava will pile up over the vent to produce a lava dome or volcanic dome. If the pressure in the bubbles is low, the eruption will produce an eruption column only a few hundred meters high, and most of the pyroclastic material will fall to close to the vent to build a cinder cone.
This type of eruption is called a Strombolian eruption , and is considered mildly explosive. Lahars Volcanic Mudflows A volcanic eruption usually leaves lots of loose unconsolidated fragmental debris. It is important to understand that lahars can occur accompanying an eruption, or can occur simply as the result of heavy rainfall or sudden snow melt, without an eruption.
Volcanic Gases Although the predominant gas erupted from volcanoes is H 2 O vapor, other gases are erupted can have disastrous effects on life. Sulfur gases in the atmosphere, along with volcanic ash, reflect incoming solar radiation back into space and have a cooling effect on the atmosphere, thus lowering average global temperatures.
The effect lasts only as long as the gases and ash remain in the atmosphere, normally a few years at the most. CO 2 gas, produces the opposite effect. It is a greenhouse gas which absorbs solar radiation and causes a warming effect. Eruptions in the past that produced huge quantities of this gas may have been responsible for mass extinction events. As the first block, began to slide downward, the magma chamber beneath the volcano became exposed to atmospheric pressure.
The gas inside the magma expanded rapidly, producing a lateral blast that moved outward toward the north. As the second slide block began to move downwards a vertical eruption column began to form above the volcano. The lateral blast rapidly overtook the slide block and roared through an area to the north of the mountain, knocking down all trees in its path and suffocating all living things, Within the next 10 seconds the third slide block moved out toward the north.
The landslide thus became a debris avalanche and left a deposit extending about 20 km down the valley see map below. Our exchangeability assumption is based on the fact that each volcano in our analysis is located in an arc possible of producing large-magnitude volcanic eruptions between Magnitudes 4 and 7.
Consequently, implicit in our assumption of exchangeability is that the injection of magma does not alter the thermomechanical behavior of the upper crust and influence magma accumulation, which occurs for the largest magmatic reservoirs feeding super-eruptions de Silva and Gregg, We refine the assumption by removing volcanic arcs undergoing significant extension e.
The assumption of exchangeability then allows us to investigate tectonic parameters that may influence variability in the behavior of each volcano, but not specifically control the size or style of eruptive activity of a single volcano.
As we are interested in the effect of regional-scale tectonic processes it is important to average the behavior of many volcanoes e. We are not attempting to explain why an individual volcano had a large or a small eruption. Rather, we are attempting to understand whether groups of volcanoes have larger or smaller eruptions on average, and if this different average behavior relates to their subduction setting.
Grouping multiple volcanoes reduces the variability arising from local processes that may influence an individual volcanic record over a short time period. Selecting volcanoes based on their geographic location is a natural approach to grouping e. For each volcanic arc we compare the mean of each tectonic parameter with the average eruptive activity recorded Supplementary Table A2. The mean represents an average of all possible eruptive states for a subduction zone magmatic system.
In total, our analysis consists of individual volcanic records constrained by the availability of tectonic parameters Figure 1 and effects of under-recording, which is described in the following section.
Due to the multi-dimensional character of subduction zones, in which parameters are likely related to each other, we first use a structure-learning algorithm to define a graphical model, which quantifies the probabilistic relationships between different variables, and guides our comparison of the eruptive behavior in different geographical regions.
We use a graphical model called a Bayesian network to define the probability of tectonic parameters and eruption observations to be related for all volcanoes in our analysis.
A graphical model is defined by a directed acyclic graph in which the variables of interest are represented by nodes i. If the data show significant departures from a normal distribution, they are usually discretised for mathematical convenience and computational speed, which is the case in this study.
Consequently, the local distribution for each node is defined using conditional probability tables CPT. A CPT defines the probability of a node, conditional on the other nodes to which it is connected. To test whether variables are probabilistically related, and thus whether they should be connected by an edge in the graph, we use a structure-learning algorithm.
Two main classes of algorithms have been developed for this task. The first, a score-based approach searches across all possible structures of a network by adding, removing and reversing arcs until it finds one that maximizes a score function that defines how well the model fits the data Russell and Norvig, The second is a constraint-based method, based on the principle of the Inductive Causation IC algorithm Verma and Pearl, , which provides a framework to test the conditional independence between nodes.
We use the first approach, specifically the Tabu search algorithm Glover, , which uses a greedy search to identify a network structure to maximize a scoring function. We choose a predictive log-likelihood score Scutari et al. The analysis is bootstrapped times using combinations of different arcs to learn and score the network, from which the frequency of each edge is calculated, and edges can be removed if their frequency falls below a data-driven threshold Scutari and Nagarajan, This approach allows us to estimate the regional-scale behavior and remove local effects under the assumption that such behavior is unlikely to be replicated across different regions.
In order to construct each bootstrap sample, we randomly sample volcanoes from 15 to 21 sub-regions from a total of 30 sub-regions, as defined by the LaMEVE database. Volcanoes in the remaining regions are used to score the network. This is because we are interested in crustal scale processes related to magma generation and transport.
These larger events characterize the tail of the frequency-magnitude distribution for volcanic eruptions, and will be less sensitive to shallow, local, short-term processes that may affect the occurrence of smaller eruptive events.
To overcome rounding bias in the eruption record magnitudes are rounded 4, 5, 6, and 7. Under-recording of large magnitude eruptions also exists due to either natural causes e. As the analysis of the eruptive record is commonly performed for a set of volcanoes that are grouped systematically e. Instead, the hierarchical structure we adopt in our analysis enables us to assess under-recording based on each individual volcanic record.
This limits the influence of any possible epistemic biases. The record of a volcano is included if it contains at least one event older than a specified date t unique. The principle of the method is to then search for a time interval in which changing the value of t unique does not affect the macro-scale properties of the group in our case the proportionality of the different magnitude eruptions.
The details of this approach are presented in Sheldrake and Caricchi , which shows for the Holocene period Volcanoes that only have very recent records are biased due to the historical nature of their record, which means that the smallest events Magnitude 4 are recorded preferentially Rougier et al.
Using the Holocene record and removing volcanoes that only have a historical record i. A slight temporal variation still exists, but is not related to eruption magnitude, geographical location or maximum age of the eruption record at an individual volcano Figure 2.
This is supported by the observation that the LaMEVE database is not overly biased by the activity of a few volcanoes Deligne et al. Figure 2. Eruption record for all volcanoes in this analysis, showing no clear bias in event age or geographical location with respect to eruption magnitude. The volcanoes are ordered alphabetically according to the classification by volcanic arc in Figure 1.
The order of volcanoes is the same order as in the Supplementary Table A1. In total, the eruption records of volcanoes satisfy the under-recording criterion presented in Sheldrake and Caricchi Of these, volcanoes have corresponding tectonic parameters that we analyze in this study. Variability in the magnitude of volcanic eruptions is calculated using the Holocene eruption record during the previous 11, years before C.
We also record the latitude and longitude of each volcano, which allows us to calculate the modeled crustal thickness at that location Laske et al. We do not consider the impact of climate on volcanism e. Figure 3. Pairs plot of all variables for the volcanoes used in this study. Darker regions in each graph represent a higher density of volcanoes.
The upper right panel contains the continuous data, with the relevant units labeled in the diagonal. The bottom left quadrant is the discretized data.
The bounds of each interval are chosen to distinguish clusters of data in each parameter and can be found in the Supplementary Figure 1.
Consequently, the bottom left panel is broadly a mirror image of the upper right panel. This can be observed by comparing the relationship between convergence obliquity O and the parallel component of subduction Sp.
Descriptions of each parameter can be found in Table 1. Kinematic parameters characterize the overall motion of subduction at each plate boundary, which is calculated using the absolute motion of the overriding and subducting plates, and the arc-trench. In our analysis we focus on velocity of subduction Figure 4 , which provides a first-order indication of the rate at which oceanic lithosphere is subducted beneath the overriding plate.
We also examine the influence of convergence obliquity, which quantifies the angle of subduction relative to the trench-normal direction, measured in degrees. This is calculated using an orthogonal component i. Figure 4. As can be seen in the figure, the rate of subduction is broadly the same as the rate of convergence. We have labeled regions that do not fall on the line between convergence rate and subduction rate; C A plot of the thermal parameter Eq. At the global scale, variation in the thermal parameter is dominantly controlled by the age of the subducting slab.
We label the Antilles, which is the only region whose thermal parameter is strongly controlled by the vertical rate of subduction. The thermal structure of the mantle is characterized by the thermal parameter Kirby et al. At a global scale, the dominant parameter is the age of the slab at the trench A; Figure 4C , which is also included in our analysis.
We do not incorporate the subducting slab angle into our analysis, which influences the distance between the trench and the volcanic arc England et al. Instead, we include the crustal thickness Z; km of the overriding plate at each volcano as a bulk measurement of upper crust variability. Quantitative uncertainties are not reported in the tectonic datasets that we use, but uncertainties are orders of magnitude lower than the parameter variance and so will not impact the validity of the correlations that we identify.
Variability in the sizes of volcanic eruptions is calculated using the results of a Bayesian hierarchical analysis of the volcanoes that satisfy the under-recording criteria described in Sheldrake and Caricchi The Bayesian analysis is performed using the principle of exchangeability, so that the probability of recording each eruption magnitude for each volcano is a result of its individual record the proportion of different eruption magnitudes and total number of events recorded for each volcano and the global record.
As these probabilities exhibit power law characteristics we perform this regression using the logarithm of each magnitude probability. We report the average of each tectonic parameter, as well as the standard deviation in the Supplementary Data Supplementary Table A2. Figure 5. Probability of recording different eruption magnitudes for; A all volcanoes in our analysis; B Izu-Bonin and Java-Bali dashed-lines , with the average distribution for each arc fitted using a power law solid lines.
The results of the structural learning algorithm are presented in Figure 6 , in which the darkness of the edge is proportional to its probabilistic strength. In total only 22 possible networks were identified by the learning algorithm Figure 6A. The strength of each edge was calculated by averaging its frequency across all networks Figure 6B. Solid lines indicate edges have passed the data-driven significance threshold, which is 0.
The variable that shows the greatest probabilistic relationship with eruption size is the obliquity of convergence. The subduction-parallel component is not significant. Figure 6. Graphical model that identifies which nodes circles representing parameters are linked by edges lines.
A A plot of the cumulative number of graphical structures and number of graphical models estimated using the bootstrap approach. In total twenty-two networks are estimated, with two of the networks identified n is equal to the number of times that network is identified as best-fitting the data.
By performing this bootstrap times we reduce the effects of noise on our analysis, and improve the strength of the probabilistic correlations. However, the relative strengths of the edges, is what we are most interested in, and these remain, even if the absolute values of the probabilistic relationships varies slightly; C The resulting graphical model from the bootstrapped networks.
The darkness of the edge represents the number of times that edge appears in the bootstrapped graphs. Edges that do not pass a significance threshold are labeled with a dashed line. The role of convergence obliquity is observed when we compare the eruptive behavior of different volcanic arcs. This is due to the roles of other parameters on eruption size, such as the age of the subducting slab and the rate at which the slab is subducting. This corroborates the results of the graphical model that the obliquity of convergence has the greatest probabilistic strength with eruption size Figure 6.
Finally, we can see the influence of crustal thickness for some groups of volcanoes, such as those with moderate convergence obliquity Figure 7A or the oldest slab ages Figure 7B , where the value of alpha decreases as crustal thickness increases. Figure 7. The graphs are color contoured for the average value of; A the obliquity of convergence O ; B the age of the slab A ; C the rate of subduction perpendicular to the volcanic arc S n. Hence, the median size of volcanic eruptions increases to the right.
To explore the combined effect of these tectonic parameters on volcanic activity in more detail we distinguish volcanic arcs into two groups based on the following parameter,. We utilize this combined parameter H to represent the average size of magma reservoir in the crust.
It allows us to distinguish the effects of high mantle productivity and low convergence obliquity, which combined favor the development of larger crustal reservoirs. Mantle productivity increases with the age of the slab A in Eq. Neglecting the cases of arcs experiencing significant extension, increasing normality of convergence Qs in Eq. Therefore, with increasing H the likelihood of formation of larger reservoirs crust also increases. The value of 3 separating the two regimes was identified using a k-means clustering algorithm.
Figure 8. Volcanic arcs are distinguished into two regimes according to the age of the slab and subduction motion Eq. We estimated graphical models for volcanoes in each of these two groups, which enables us to identify within-group distinct dominant behaviors Figure 9. For the high-H group the rate of subduction has no significant correlation with the magnitude distribution of volcanic eruptions Figure 9A.
Instead, the age of the slab becomes the dominant parameter followed by the obliquity of convergence and then the thickness of the crust Figure 9A. This is also visible in the results of the graphical model, which indicates that the strongest probabilistic link with eruption size is with crustal thickness Figure 9C.
Here the only exception is the Central Aleutians, which has exceptionally large values of S p for this group Figure 9D. Figure 9. The dominant parameters for the two regimes defined in Eq. Aleutians iii , which has high rates of subduction parallel to the volcanic arc. Arcs are labeled as follows: i Antilles, ii Cascades, iii C. We have compared the tectonic setting to the eruptive records of individual volcanoes, using a Bayesian network.
We have further compared the tectonic setting of different volcanic arcs to their average eruptive behavior. These two approaches have provided strikingly similar results and identified the obliquity of convergence, combined with the age of the slab and crustal thickness, as the most important parameters controlling the size distribution of volcanic eruptions Figures 6 — 9.
Using these results we discuss the possible processes that influence the size of volcanic eruptions up to magnitude 7 in different volcanic arcs.
Magma chambers are formed from the assembly of multiple sills, which will accumulate when vertically orientated dikes are retarded at the boundaries between crustal heterogeneities Gudmundsson, However, regional compression can also change the trajectory of dikes to form sills Menand et al.
When subduction is more normal i. For the set of volcanoes included in this analysis, lower convergence obliquity is more commonly associated with compressive rather than extensive stress regimes Figure Hence, we suggest that in low-obliquity arcs in our study, compression may also reduce the vertical transport of magma, promoting the formation of larger magmatic reservoirs and larger eruptions i.
Figure
0コメント