Earth’s mantle constitutes the largest sub-system of the whole Earth system, involving 70% of the total mass, ～80% of the heat capacity, and more than 50% of the internal heat generation by radioactive decay. Therefore, the mantle and the inherited dynamics may control the whole system to a great extent, e.g., in terms of convective motion (including plate motion as its surface expression) and heat transport from the core to the surface, regulating the core cooling and dynamo that eventually affects the surface environment and life. First the basic structures and dynamics of the mantle convection are described, which demonstrate that the surface cooling dominantly drives the convection, creating buoyancy of several to 10 times greater than that generated near the core-mantle boundary. This estimate for the much larger role of near-surface cooling is consistent with the seismic tomography. Then various types of observations on the structures and dynamics of mantle, particularly three boundary layers (i.e., the near-surface, mid-mantle around 660km discontinuity, and core-mantle boundary) have been reviewed and are compared with the simple estimation. Of these, the ’geochemical probe’ approach, which utilizes composition (in particular the isotopic composition) of young basalts that fingerprint geochemical nature of the mantle materials, has been reviewed in conjunction with convective regimes. The latest result of high spatial resolution has revealed that the mantle can be divided into the eastern and western hemispheres, in terms of an anciently (several hundred million years ago) subducted fluid-component. The spatial pattern is strikingly similar to the hemispherical seismic structure of the inner core. Based on these observations, a model for ‘top-down hemispherical dynamics’ is introduced, as a result of focused subduction towards the supercontinents that existed mostly in the eastern hemisphere from ～900 to 250 million years ago (i.e., Rodinia, Gondwana and Pangea). The cooled domain of mantle may absorb heat from the eastern hemisphere of the core, resulting in faster growth and velocity of the eastern half of the inner core. Such ‘top-down’ dynamics is consistent with the various types observations and arguments (made in the first half of this paper) on mantle convection.

抄録全体を表示

The subduction zone system on the Earth is over 40,000km long, comparable to the circumference of the Earth. The active processes -brittle deformation, metamorphism, convection and volcanism- beneath volcanic arcs or continental margins are all linked with fluids derived from subducting oceanic plates. Here I review recent geophysical observations in subduction zones and organize our thoughts on ongoing magmatic processes, focusing on a low-velocity, high-attenuation, and low-resistivity zone in the mantle wedge as an indicator of melt migration paths. In subduction zones where a plate at moderate to old age is subducting, two types of melt migration paths are observed, with an inclined migration path for a gently dipping slab and a sub-vertical migration path for a steeply dipping slab. These observations suggest that melt transportation is primarily governed by the geometry of a mantle upwelling flow developed sub-parallel to the down-dip direction of the slab. However, melts appear to migrate sub-vertically to volcanoes in young subduction zones. Release of slab-derived fluids at shallow depths would produce partial melting only in the fore-arc tip of a mantle upwelling flow, and hence melts may migrate sub-vertically by buoyancy, instead of effective transportations through the upwelling flow as observed in relatively old subduction zones. An understanding of subduction zone processes with reference to the more quantitative integration of all earth science categories (seismology, volcanology, geodesy, petrology, mineralogy, geochemistry, geology, geomorphology, etc.) will be required to improve our knowledge of arc magmatism.

抄録全体を表示

This paper discusses physical phenomena of volcanic eruptions and magma ascent on the basis of a model of magma plumbing system where the conduit flow dynamics and the magma chamber processes including the conditions of start and end of an eruption are taken into consideration. According to the conduit flow models, the relationship of magma discharge rate, M, and the magma chamber pressure, p_ch, (“the M-p_ch relationship”) during explosive eruptions is controlled by the pressure at which the conduit flow changes from a bubbly flow to a gas-pyroclast flow (“the fragmentation pressure”). The fragmentation pressure, in turn, depends on the mechanisms of gas escape and magma fragmentation. The M-p_ch relationship during non-explosive (effusive) eruptions depends on the density change due to gas-escape process and the viscosity change due to crystallization during magma ascent in the conduit. The models of magma chamber processes, on the other hand, suggest that the M-p_ch relationship strongly depends on the effective compressibility of magma chamber and the volume of magma chamber. The effective compressibility of the magma chamber drastically increases when the magma contains gas phase, and hence, it depends on water content of magma and pressure. The diverse features of eruption sequence result from the coupled effects of the conduit flow dynamics and the magma chamber processes. For example, the condition of start and end of an eruption depends on how the conduit flow is driven by the magma chamber pressure, as well as the mechanical stability of magma chamber. Some key parameters of the conduit flow dynamics are determined by the long-term physical processes of magma chamber (e.g., crystallization and differentiation of magma) during the repose period. In order to establish a method to forecast eruption sequences, forward and inverse models of the magma plumbing system are formulated. Because of the above coupled effects of conduit flow dynamics and magma chamber processes, the forward model for the magma plumbing system shows complex behavior of eruption sequences (i.e., various trajectories of the M-p_ch relationship). Such complex behavior, as well as the lack of knowledge of the mechanical stability of magma chamber and the effects of deformation of conduit, makes it difficult to forecast eruption sequence on the basis of the forward model. The difficulty also comes from the fact that the magma chamber volume and the magma chamber pressure cannot be independently determined by the inverse model based on the geodetic and geological observations. Although, because of these difficulties, the model of magma plumbing system is not immediately useful for the volcanic disaster prevention, it certainly provides a frame-work to integrate different volcanological approaches for understanding of the diversity of eruption sequence.

抄録全体を表示

Water plays an important role for magma genesis and frictional properties; consequently, water circulation systems contribute to the variation of magmatic and seismic activity at subduction zones. Although subducting plate transports a large amount of water, most of water is released into the mantle via dehydration reactions at elevated temperature during subduction. Aqueous fluids released from the subducting plate then migrate along the plate boundary due to permeability anisotropy developed in the highly sheared serpentinite. Based on laboratory data, we estimated the fluid migration velocity to be〜7cm/year, which is close to the descending plate velocity, suggesting that polarity of water migration can be different in subduction systems. In northeast Japan, fluid migration velocity is slower than the subduction velocity, and hence water is transported downward into the deeper portions trapped by the mantle corner flow. In contrast, in southwest Japan where the fluid velocity is higher than the subduction velocity, water could be returned to the shallow regions along the subducting plate interface. This model can explain the seismic low velocity anomalies and geochemical signatures in these regions, in which the hydration of the plate interface is observed in shallow mantle wedge in southwest Japan, but is limited to the deeper parts of the mantle in northeast Japan. Water transported to deep levels could contribute to the active arc volcanism in northeast Japan, whereas water circulating at shallow levels in southwest Japan could trigger slow earthquakes due to fluid pressure build-up at the plate boundary.

抄録全体を表示

Arc lavas are characteristically evolved, multiply saturated, and rich in phenocrysts. Recent work in the Mariana arc, however, has shown that small parasitic cones on the flanks of larger volcanoes at the depth of ～2,000m often yield mafic lavas which are not found in the main edifice. Finding and studying such lavas is fundamental to understand the nature of the mantle source and the processes that yield intermediate arc magmas. Additional recent observation is variations in the crustal thickness and its internal structure which is closely related to the average compositions of the lavas from the Quaternary volcanic front in the Izu-Ogasawara arc. I attempt to combine the geophysical findings with the petrology of the volcanic rocks from the oceanic arc in order to understand how the ancient Earth covered by a global ocean developed to the present.

抄録全体を表示

How subduction begins and its consequences for global tectonics remain one of the essential outstanding problems of plate tectonics. Two different endmember mechanisms for subduction initiation have been hypothesized: spontaneous, and induced (or forced). Numerical models suggest that subduction initiation is induced by externally forced compression along a preexisting discontinuity in an oceanic plate such as a fracture zone or transform faults. However, it has been pointed out that spontaneous subduction must have occurred at some points in Earth's history to initiate plate tectonics, and recent numerical models demonstrated that lateral thermal/compositional buoyancy contrast along plate discontinuity or within lithosphere can cause spontaneous subduction initiation. Recent geological and geophysical surveys in the Izu-Bonin-Mariana fore-arc have revealed igneous processes in the initial stages of subduction. The oldest magmatism after subduction initiation generated MORB-like fore-arc basalts, which was associated with seafloor spreading caused by onset of sinking of slab into mantle. Then boninitic magmatism followed by tholeiitic to calc-alkaline arc lavas collectively makes up the extrusive sequence of the fore-arc crust. This magmatic evolution from initial basaltic magmatism to establishment of normal arc magmatism took several million years. Fore-arc stratigraphy observed in the Izu-Bonin-Mariana arc shares some of the key geologic and petrologic characteristics with many supra-subduction zone ophiolite, which implies that fore-arc crustal section produced in the initial stage of oceanic island arc formation could correspond to in-situ section of supra-subduction zone ophiolite prior to obduction. Recent ocean drilling projects targeting initial stage of the Izu-Bonin-Mariana arc inception revealed that subduction initiation to form the Izu-Bonin-Mariana arc took place spontaneously. The drilling results also revealed that the whole arc was established on the ocean crust produced associated with subduction initiation.

抄録全体を表示

Large-scale pyroclastic eruption is one of the most awful natural disasters on the earth. Though their frequency is relatively low compare to the lifetime of human society, large-scale pyroclastic eruption can make serious impact on the global environment. Frequency of the volcanic eruptions shows a negative correlation against their scale: global frequency of the eruptions larger than VEI7 is approximately ten per 10,000 years, whereas more than 10 eruption of VEI4 occur every 10 years. The storage of voluminous magma within a shallow crust is a key process for the preparation for large-scale eruption. Inactive thermal convection in highly-crystallized magma bodies and visco-elastic behavior of the surrounding host rock can allow the stable storage of voluminous felsic magma at the neutral buoyancy level in the upper crust. Segregation of interstitial melt to form a melt pocket in highly-crystallized magma body can cause smaller scale of eruptions, whereas the remobilization of entire part of magma chamber will result a large-scale eruption with caldera collapse. Rupture and collapse of the roof rock of magma chamber induced by rapid decompression of magma chamber is the fundamental process of the eruption of voluminous magmas within short period. The decompression of magma chamber activates the slip of ring fault at the marginal portion of the roof and consequently the caldera starts subsidence. The collapse is controlled by the decompression inside the chamber and the strength of the roof rock. Ring fault turns to an open ring facture through which the voluminous magma can erupt to produce large ignimbrite. The volume of magma erupts during a caldera-forming eruption against the total magma chamber volume show negative correlation against the chamber size. This means that the large fraction of magma can remain even after caldera collapse particularly in large magma chamber. Evaluation of "precursory process" for catastrophic eruption is important to understand the driving mechanism of catastrophic eruption and also the hazard assessment. Accumulation of magma and building of a large-volume magma chamber within the earth’s crust is a long-term preparation process for catastrophic eruption. Short-term process for catastrophic eruption is the destabilization and rupturing process of the magma chamber.

抄録全体を表示

This paper reviews the eruption processes, as well as the processes leading up to the eruption, as revealed by multi-parameter geophysical observations. First, we briefly describe the history and development of geophysical observational methods in Japan. A prototype of multi-parameter geophysical observation had already been established at a few volcanoes in Japan 50 years ago. To obtain high-quality geophysical data, underground tunnels and boreholes had been constructed around volcanoes, and seismometers, tiltmeters, and strainmeters had been installed in them during the 1980s and 1990s. Broadband seismometers and a data logger with a GPS clock that are small and lightweight, were introduced in the mid-1990s. Combined with large data storage capacity in the data loggers and efficient data transmission using IP protocols, simultaneous multi-parameter geophysical observations have been intensively and continuously conducted since the early 2000s in order to reveal the eruption processes. Since then, various analysis methods suitable for multi-parameter observations have been developed. Although the time scales of volcanic eruptions vary widely, from several seconds to several decades, most of these scales are covered by multi-parameter geophysical observation. We review the advances and challenges for understanding not only the eruption processes, but also the precursory processes leading up to eruptions, which have been revealed by multi-parameter geophysical observations during the past decade. The explosive eruption processes, which include Plinian, Vulcanian, and Strombolian eruptions, and effusive eruptions, which include Hawaiian and dome-forming eruptions, are also reviewed. Physical models for these types of eruptions have been developed experimentally and theoretically, and examined later using geophysical observations. For each type of eruption, we briefly introduce the proposed physical models and describe their progress, mainly from the viewpoint of multi-parameter geophysical observations. The eruption processes of Plinian eruptions have been revealed mainly by worldwide seismic and infrasound observation networks and satellite images. Vulcanian and Strombolian eruptions are most intensively studied by multi-parameter geophysical observations, because of their high-frequency nature and accessibility to the proximity of active craters in deploying monitoring instruments. For example, the precursory processes of Vulcanian eruptions are characterized by inflation, relatively stable, and slight deflation stages of the volcano edifices, which are inferred by continuous crustal deformation measurements. Strombolian eruptions are understood as repeating strong gas bursts at the surface of liquid magma. Recently, new monitoring techniques such as high-speed cameras and portable radar units have been introduced to observe the Strombolian surface activities. Hawaiian eruptions are characterized by explosive lava fountains and large lava flows traveling more than several kilometers. Multi-parameter observations are quite useful for monitoring the locations of dike intrusions and lava fountains. Dome-forming eruptions have the longest time-scales among these eruption styles. Temporal fluctuations of dome growth are well monitored by geodetic and photogrammetric observations. Recent seismic observations of dome-forming have revealed the characteristic repeating earthquakes and provided new insights into the physical mechanisms of dome growth. Our understanding of the physical process of phreatic eruptions is quite limited relative to magmatic eruptions. Recently, tilt changes associated with tremors were sometimes observed before phreatic eruptions at several volcanoes, and may be forerunners of phreatic eruptions. We need much more data to understand the processes leading up to phreatic eruptions. Erupted volumes and rates are inferred from multi-parameter geophysical

抄録全体を表示

Volcanic gas plays a crucial role in the dynamics of eruption and ascent of magma. Volatiles degassed from magma are emitted to the surface as high temperature volcanic gases. Lately observation techniques to measure volcanic gas have been developed, allowing us to monitor volcanic activities and to compare the volcanic gas data with geophysical data. The volcanic gas composition and emission rates have been measured so as to elucidate the magma plumbing system. The volcanic gas composition gives us the information of degassing pressure and temperature of the volcanic gas within the volcano. The volcanic gas emission rates reflect the production rates of the degassed magma within the volcano. In many cases, the amount of volcanic gas observed exceeds the gas amount which can be degassed from the erupted magma. At some active volcanoes, a significant amount of volcanic gas is emitted not only during eruptive periods but even during quiescently degassing periods. These results suggest that only a portion of magma is erupted yet the rest is degassed at a depth without discharge. These observational results are known as “excess degassing”. To explain this, degassing models (the permeable flow, magma convective degassing, and gas percolation models) were proposed. Recent studies suggest that the condition of bubble segregation from the magma is a key parameter for the magma degassing process, which controls if an eruption becomes explosive or not. If the bubble separation from the magma does not occur during the magma ascent, the gas volume fraction of the magma increases monotonously, leading to the fragmentation of magma. The transition of the closed- to open-degassing within the conduit was proposed so as to explain the significant volcanic gas emission without eruptions. Recent developments of volcanic gas observation techniques have opened up the possibility to reveal the linkages between degassing and geophysical (seismic or geodetic) phenomena. The relationship between very-long-period seismic events and volcanic gas exhalation was found by multi-disciplinary observations. From the viewpoint of the geodesy, the volcanic gas emission could cause deflation of the volcanic body. The examination of the magmatic and volcanic processes from both viewpoints of geochemical and geophysical studies is important.

抄録全体を表示

Magmas include phenocrysts and bubbles, and sometimes fragment under rapid deformation. Ascent of such a complicated magma generates and destroys surfaces between the melt and other phases. In this system, magma dynamics and physical properties of magma are coupled. The interaction between the magma dynamics and physical properties has not yet explained well, and prevents our understanding of eruption dynamics. Recently, several laboratory experiments have performed, describing interaction between dynamics and physical properties quantitatively. Here, I would like to review those experiments.

抄録全体を表示

This paper presents some basic concepts on possible cooperative framework for contributing to disaster mitigation during volcanic eruptions with the intention of enhancing discussion among members of the Volcanological Society of Japan. At first, this paper describes some examples of problems that have been argued during recent volcanic eruptions because of improper risk communication of volcanologists, and then, outlines the present state of a coordination system for effective disaster assistance by multiple stakeholders with a focus on recent efforts in public health and medical communities. Preliminary ideas on “Expert Assistance Team during Volcanic Crises” are also presented for further discussions.

抄録全体を表示

Deterministic eruption scenarios may mislead taking countermeasures for coming hazards. Preparing the event tree covering all phenomena which may happen in future eruptions even with low probability for the volcano is important not only for forecasting eruptions but also for disaster prevention. Eruption event trees can be prepared in various concepts, for example, eruption type, scale, hazard type, impact direction or area and so on. The probability tree is the event tree equipped with probabilities for the branches. Probability trees by USGS and European scientists include the cumulative trees, trees based on scientists' elicitation and Bayesian trees. Introduction of the eruption event trees into the Japanese volcanologist community began around 2009. Then, event trees were prepared for Izu-Oshima, Miyakejima, Sakurajima, Usu, and Izu-Tobu volcanoes. Reasons for branching and time scales of events were also discussed and shown on the event trees together with probabilities. The event tree for Sinabung volcano, Indonesia, as an example of lava dome-forming eruptions was drawn in 2011, based on the geological study. On-going lava dome/flow eruption at this volcano just followed the most probable scenario. For Sakurajima volcano, a conceptual event tree was drawn for understanding the anomalies controlling the eruption scale.

抄録全体を表示

The main stream of the researches for prediction of volcanic eruption in Japan has been promoted through the national program which was established in 1974. The research has been advanced based on the development of basic volcanology and on the accumulation of practical knowledge obtained through the occasional volcanic eruptions. As it was shown in the review of the recent volcanic eruptions in Japan and the measures taken to mitigate the disasters caused by these eruptions, the prediction of volcanic eruptions in Japan is, however, on the stage of empirical pattern recognition. It is still far away from prediction based on the models of the underlying dynamics of volcano. Even in such situation, public society asks when and where eruption will occur, and how long the eruption will continue. It is difficult to answer these questions; however it is necessary to provide useful information based on the monitoring of eruption and the application of the available knowledge for the mitigation of volcanic disaster. In this context, several issues which might be concerned in the research of prediction of volcanic eruption and in developing the measures to mitigate volcanic disasters are described.

抄録全体を表示

We investigated the geology of Okanodake volcano and the area adjacent to it in the Futago Volcanic Group, NE Kyushu, Japan. Based on the geology and petrology of this area, the formation of Okanodake Volcano (biotitehornblende dacite) began with phreatomagmatic eruptions and ceased with effusion of lava. The age of the eruption is assumed to be 1.19Ma. The volcanic products consist of pyroclastic deposits, pyroclastic surge deposits and block-and-ash flow deposits, and intrusive rock. SiO₂ contents range from 63-67wt.% and K₂O contents range from 1.9-2.4wt.%. Each of the volcanic products is classified within the calc-alkali series. The volume of the various volcanic products were calculated to be ca. 0.07km³, 0.008km³, and 0.002km³, respectively. The total volume is ca. 0.06 DRE km³.

抄録全体を表示

We carried out a paleomagnetic study of the lava flows of Minamidake, which is an active vent of Sakurajima volcano in Kyushu, Japan. The Minamidake volcano started its eruption about 4.5ka, which formed the Older Minamidake and overlying the Younger Minamidake, and several lava flows from parasitic vents in historical time. The oldest lava from the Older Minamidake is Miyamoto lava, whose age was already determined by paleomagnetic study as ca 4 ka. Miyamoto lava is overlain by Kannonzaki lava and Arimura lava. The age of the Arimura lava was estimated to be about 3.1-2.7ka by means of the comparison between paleomagnetic direction of the lava and paleo-secular variation of geomagnetic field. The age of the Kannonzaki lava lying beneath the Arimura lava was consequently determined to be about 3.2-2.7ka by the previous paleomagnetic data. The Kannonzaki and the Arimura lavas accumulated successively from intermittent eruptions during less than 500 years at around 3ka. The volume of lava flows of Minamidake erupted for about 4,000 years was estimated. It became clear the volcanic edifice of Minamidake grew mainly at around 3 ka, as inferred from the accumulation rate of lava flows. The volume of the Nagasakibana lava erupted in A.D. 764, which is the oldest lava in historical time, was estimated to be about 0.9km³, which is greater than the previous estimate. The overall magma effusion rate in the past 240 years and that at around 3ka are faster compared to the effusion rate in other period.

抄録全体を表示