Studies on the supply system of magma and its behavior based on seismic and geodetic observations are reviewed for recent volcanic activity of the Sakurajima volcano, where three characteristic types of eruptions occurred in the 20th century: the 1914 gigantic flank eruption; the 1946 minor flank eruption; and frequent vulcanian eruptions at the Minamidake summit crater, continuing from 1955 for more than 50 years. The magma plumbing system of Sakurajima is approximated by a deep (～10km) magma reservoir beneath the Aira caldera, a shallow reservoir (3-6km) beneath the central cones, and a conduit connecting the shallow reservoir to the summit crater. From the inflation-deflation pattern and intensity of the vertical displacement, it is estimated that magma has been supplied to the deep reservoir at a rate of 10⁷ m³/year. The shallow reservoir is inferred from vertical displacements, a tilt vector showing the crater side up before the vulcanian eruptions, and an anomalously attenuated zone of seismic waves. The magma migrated from the deep reservoir to the shallow reservoir before the increase in vulcanian eruptivity at the Minamidake crater, as demonstrated by the relocation of the upheaval center of the ground deformation. The increase in seismicity of A-type earthquakes and the migration of their hypocenters toward the shallow part also support the migration of magma. The hypocenters of B-type earthquakes and explosion earthquakes with volumetric sources are concentrated beneath the crater. The hypocenters of A-type earthquakes generated by shear fractures surround the origins of B-type earthquakes and explosion earthquakes. The separation of the hypocenters indicates the existence of a volcanic conduit connecting the shallow reservoir to the bottom of the summit crater. The intrusion of magma into the conduit and smooth ascent up to the crater bottom induce strombolian eruptions and swarms of B-type earthquakes. The magma at the uppermost part of the conduit becomes a cap rock of the conduit against the following intrusive magma, which generates vulcanian eruptions. Upward tilt of the crater side and extensional strain are observed prior to both strombolian activities and the vulcanian activities that follow and turn into downward and contraction strain, respectively, associated with the eruptions. The volatility of the magma has an important role in vulcanian eruptions, in the formation of a gas pocket at the uppermost part of the conduit, and in sudden outgassing triggered by a pressure decrease in the conduit due to gas leakage.

View full abstract

The temporal variation of magmatic chemistry and the evolution of magma chamber system of the Sakurajima volcano and Aira caldera region since 61 ka are studied based on the whole-rock major element, incompatible trace element and rare earth element chemistry of the eruptive products. The magmas of the Sakurajima volcano and Aira caldera region since 61 ka consist of four groups: (1) basaltic to basaltic andesitic magma of the mantle origin, (2) rhyolitic to high silica rhyolitic magma of the crustal origin, (3) dacitic magma and (4) andesitic magma produced by magma mixing of the mafic magma of mantle origin and the crustal felsic magma. Around 61 to 60 ka, basaltic to basaltic andesitic, andesitic and rhyolitic magmas were active in the Aira caldera region, and the Shikine andesite and the Iwato pyroclastic flow deposit were erupted. After a dormant period of about twenty-four thousands of years, the rhyolitic magmatism resumed and the voluminous high silica rhyolitic magma erupted at 29 ka to form the large-scale Osumi pumice fall and Ito pyroclastic flow deposits. The felsic magma produced the Iwato pyroclastic flow deposit and the Osumi pumice fall and Ito pyroclastic flow deposits were similar in composition; the latter high silica rhyolite can be derived from the former rhyolite by crystallization differentiation. The rhyolitic to high silica rhyolitic magma chamber system was stable and long-lived with duration of about thirty thousands of years. The magmatic activity of the Sakurajima volcano began at 26 ka after a quiescent period of about three thousands of years. The Moeshima rhyolitic magma discharged at 13.8 ka in the Aira caldera constitute another magma chamber system different from that of the Sakurajima volcano. The magma chamber system of the Sakurajima volcano was composed of the low Ti-P type and high Ti-P type dacitic and andesitic magmas. The magma chamber system of the low Ti-P type, which was active from about 14 to 4 ka, comprises at least the three sub-systems based on the whole-rock chemistry, while that of the historical eruption since 8^th C is restricted to the high Ti-P type and consists of the three sub-systems, the youngest of which has been active since the subaqueous An-ei eruption at 1779AD. The duration of the activity of each magma chamber sub-system of the Sakurajima volcano is rather short, the time span of which is thousands to several hundreds of years.

View full abstract

We present the results of a detailed petrogenetic study employing newly determined whole-rock major and trace element geochemical analyses and Sr-Nd-Pb isotopic compositions of andesitic and dacitic Quaternary lavas of Sakurajima volcano, a post-caldera volcano situated within the Aira caldera of Japan. Similar geochemical and isotopic investigations are also carried out on basaltic rocks from pre-caldera stage and monogenetic volcanoes from near Sakurajima volcano. Quaternary lavas of Sakurajima volcano analyzed in this study are classified as porphyritic andesites or dacites that contain a mineral assemblage of orthopyroxene, clinopyroxene, and plagioclase, with or without olivine, in a groundmass exhibiting either hyalo-ophitic or hyalopilitic textures. The trace element characteristics of these samples are similar to those of typical island arc magmas, showing clear evidence of Nb depletion along with enrichments in Rb, K, and Pb, which suggests the addition of aqueous fluids to the mantle wedge during melt generation. The Sr, Nd, and Pb isotopic compositions plot close to a mixing curve between MORB-type mantle and sediments of the Philippine Sea Plate, but displaced a bit towards more radiogenic compositions. Plots of Zr versus Nb concentration in these lavas yield a linear trend that falls on a compositional mixing line between the values for mid-ocean ridge basalt (MORB) and average continental crust. Collectively, these observations indicate that the primary source magmas for the Quaternary lavas of Sakurajima volcano were initially generated by partial melting of MORB-type mantle wedge that had already been hydrated by fluids derived from the subducting Philippine Sea Plate. The additional contribution of significant amounts of crustal material during magma evolution is also evident from the Zr/Nb ratios and Sr-Nd-Pb isotopic compositions of the analyzed andesites and dacites of Sakurajima lava samples. From the mixing relation of Sr-Nd-Pb isotopic compositions, it is suggested that the sedimentary rocks of Shimanto Group can be a source of the crustal materials. Although most of the major element oxide compositions of these lavas show a single linear trend on each of the Harker diagrams, two different trends are clearly discernible on each of the P₂O₅, and TiO₂ versus silica variation diagrams, and are subdivided into low-P and high-P geochemical groups. These two groups can also be distinguished when comparing their P₂O₅ and TiO₂ contents and ⁸⁷Sr/⁸⁶Sr ratios, relative to their phenocrystic plagioclase modal abundances. The magma mixing trends of Sakurajima lavas, which seem to be extended from mono andesitic end-member to two different deictic end-members, are observed from the relationships of major element contents and ⁸⁷Sr/⁸⁶Sr ratios. In addition, the low-P versus high-P groups of lavas show distinctive distribution patterns, whereby the high-P lavas are surrounded by low-P lavas in the central to southern parts of the Sakurajima volcano study area. These observations indicate that mixing of andesitic and dacitic magmas played an important role in the genesis of Quaternary lavas of Sakurajima volcano, and that multiple dacitic magma chambers with different geochemical characteristics once existed beneath the Sakurajima area at relatively shallow levels in the crust. From the relations between SiO₂ and Sr isotope ratios, an assimilation and fractional crystallization process is required to originate the andesite and dacite end-members.

View full abstract

The An-ei eruption (1779-1782 A.D.) and Taisho eruption (1914-1915 A.D.) were large-scale eruptions of Sakurajima Volcano. The latter, the largest eruption in Japan during the 20^th century, produced about 1.5 km³ DRE of andesitic magma. In both cases, flank eruptions from two sides of the volcano caused pumice fall and lava flows. The An-ei eruption occurred on the northeastern and southern flanks (An-ei NE and An-ei S eruptions, respectively) and the Taisho eruption occurred on the western and eastern flanks (Taisho W and Taisho E eruptions, respectively). In the An-ei NE, two fissures are recognized from the alignments of craters: the main fissure (5 km long) and a minor fissure (1 km long). A large pyroclastic cone consisting of welded pyroclastic materials was formed along these fissures on the upper to middle flank slopes. Old drawings of the An-ei eruption show that a large amount of pyroclastic materials fell from the eruption column in the proximal area. Thus, the cone was considered to have formed simultaneously with the Plinian eruption. The presence of many cracks and a horse-shoe shaped depression on the cone is attributed to the deformation and collapse of the pyroclastic cone due to gravitational instability on the flank slope of the volcano. A stratigraphy of the eruption products shows that many lava lobes were formed after the initial Plinian eruption. In the An-ei S, the existence of a deformed pyroclastic cone sticking on the steep upper slope below the summit crater and clastogenic lava flows on the downslope indicate the syn-Plinian deposition of pyroclasitc materials on the steep slope. After then, effusion of lavas and some explosions occurred from the chains of craters on the middle flank slope. The An-ei eruption progressed continuously in three stages. The initial Plinian eruption for up to two days on the upfissure (Stage 1) was followed by lava flows on the downfissure (Stage 2). Then, submarine eruption occurred intermittently for about two years on the NE offshore (Stage 3). In the case of the Taisho eruption, initial Plinian eruption (Stage 1) was followed by lava flows associated with the intermittent Vulcanian eruptions (Stage 2) and then, gentle lava outflow continued for more than 1.5 years in the Taisho E in Stage 3. In both eruptions, progress from an explosive pyroclastic eruption at a higher flank in Stage 1 to the effusion of lava at a lower flank in Stage 2 could be explained by the propagation of a radial dyke. Although the pyroclastic cone of the An-ei NE and Taisho W are large in scale, the syn-Plinian clastogenic flow and cone collapse of the Taisho W were extensive. On the other hand, the cone of the An-ei S and Taisho E are small in scale. The cone of the An-ei S was deformed on the steep slope. From these variations, the intensity and duration of the Plinian eruption and the gradient of the flank slope would result in the various morphologies of the resultant cone. Concerning the growth of the edifice of Sakurajima Volcano, the proximal process in which the pyroclastic materials plaster the flank slope in Stage 1 does not contribute to the growth of the summit area, but to that of the flank slopes. The intense proximal deposition of pyroclastic materials during the initial vigorous Plinian eruption and the following lava flow mean that a rapid response against fires caused by pyroclast fallout and lava flows from multiple craters on eruptive fissures is needed to minimize damage in future eruptions.

View full abstract

The 1914 eruption of Sakurajima volcano was one of the largest ones in its history and first observed with scientific methods of those days. The eruption flowed out a large volume of lavas, and was accompanied with remarkable deformations on and around Aira caldera. The deformations consist of three stages: pre-eruptive, co-eruptive and post-eruptive ones: The first one is determined by temporal tide-gauge observations at Kagoshima Port but, its origin is still unsettled. The second one is a quasi instant and elastic deformation, and therefore interpretable by a simple model. However, the location of center of depression is ambiguous due to the insufficient number of observations. Prior to the interpretation, it is re-examined considering the results of triangulations in the wider area, and the location changed about 5 km S, on Sakurajima Island. The revised location of depression center can well interpret the observed depression and compromises with seismological and petrological knowledge. The last one is a secular recovery stage, and it has remained highly problematic whether its origin is endogenous or exogenous. The present author interprets that the recovery changes are composed of two factors: In the period before around 1940, viscoelasticity of the crust acted predominantly (endogenous) and later magma pressure accumulated in the reservoir gradually joined to act (exogenous) and its contribution had become noticeable after 1955. The recovery of the depression is interpreted as viscoelastic with retardation time of 16.6 years. Hence, viscosity of the earth crust around Aira caldera is determined at about 3 × 10¹⁹ Pa・s on the assumption that the crust is Kelvin-type. The value is comparable to those obtained with Miyake volcano and by laboratory measurement of a granite piece. The other cases of viscoelastic deformations accompanied with volcanic eruptions are searched: the 1779 eruption of Sakurajima volcano and the 1983 eruption of Miyake volcano are exemplified.

View full abstract

Spatial distribution of parasitic vents is closely related to movements of magmas at a certain depths of main conduit of the central volcano. A simple method for studying distribution of parasitic vents is presented: That is numbers of parasites per unit area according to radial distances from the central vent. On and around Sakurajima volcano, two peaks of the distribution diagram are found at roughly 2.5 and 8.5 km in radial distances. These can be interpreted into that the magmas branched away at different depths of the main vent. The branching mechanism is discussed from a standpoint of material mechanics. In this case, magmatic forces are assumed to be due to point dilatations that have proved effective in interpretation of surface deformations observed at various volcanoes. To interpret formation of parasitic vents, or outward fractures, on the flanks of a polygenetic volcano, the theory of maximum shear stress is adopted. As a result, a parasitic vent branches off from the main conduit at a depth that is related to the radial distance of the parasitic vent from the center of the volcano, and theoretically we may expect twin parasites symmetrically with respect to the center of the volcano. Whether new magmas outburst at the main crater or a new parasitic-vent fractures at the flank may depend on conditions of the main crater, the relative strengths of both the sites, and mechanism of branching. The three largest parasitic eruptions on Sakurajima volcano in historical times, the 1471～76, the 1779～80, and the 1914 eruptions, are examined: Each of these eruptions opened two vents on the opposite flanks of the central summit with a partly exception in the 1779～80 eruption. The exceptional case is suggestive for formation mechanism of twin parasitic cones. Formation of such twin vents is mechanically normal but empirically odd. An empirical fact that parasitic volcanoes only erupt once is hypothetically interpreted: Surroundings of parasitic conduits are probably strengthened mechanically by intrusion of magmas, and the sub-conduit may be tightly choked with lavas. We may say that the next eruption of Sakurajima volcano may take place at the summit crater, and otherwise, statistically, parasitic eruptions may burst probably on the flank and rarely at the sea. The parasitic vents would open at a region of “parasite-gap” on the flank, and would twin at the opposite sides of the summit. To improve the forecast, we need to clarify the formation mechanism of parasites in more detail.

View full abstract

In order to assess the potential for future eruptive activity at Sakurajima volcano, southern Japan, ground deformation around northern Kagoshima Bay (i.e., Aira caldera), especially the height change between leveling stations 2474 and 2480, was re-analyzed. Prior to the large eruption in 1914 at Sakurajima volcano, a remarkable inflation was confirmed at Aira caldera, based on re-surveyed data that were not referred to in previous investigations. Considering those data, the upper limit of the magma storage at Aira caldera just before the 1914 eruption could be obtained. Although the 1914 eruption accompanied a remarkable deflation at Aira caldera, magma started to accumulate again and was likely to have exceeded the level observed in 1900 (14 years before the large eruption). Around the early 1970s, it approached the level before the 1914 eruption. After the ground uplift stopped and slightly reversed during a period of extremely frequent explosions at Sakurajima volcano in the 1970s and 1980s, the inflation seems to again be approaching the inferred level before the 1914 eruption, suggesting the possibility of the next large eruption. In addition, inconsistencies between the inferred amounts of magma supply and observed volumes of erupted materials were discussed and left for further study.

View full abstract

Eruptive activity at the Showa crater of the Sakurajima volcano has steadily increased since it resumed in June 2006, and 2718 vulcanian eruptions occurred during the period from 2008 to 2011. In this paper, we clarify the characteristics of vulcanian eruptions at the Showa crater based mainly on ground deformation. A long-term extension of the ground of the Aira caldera and Sakurajima, repeating quasi-annual cycles of minor inflation-deflation, were obtained by GPS, tilt and strain observations. The inflation event that started in October 2009 was the largest. Major pressure source was estimated to be located at a depth of 12 km beneath the Aira caldera for the inflation and a minor source was obtained a depth of 5 km at the northern flank of Kitadake in addition to a source beneath Minamidake which has been know by previous studies. The magma plumbing system is composed of a major magma reservoir at a depth of ≈ 10 km beneath the Aira caldera and additional magma reservoirs at depths of around 5 km beneath the summit area from the north flank of Kitadake to Minamidake. Strain changes which indicate inflation were detected prior to explosions and the inflation strain lasted mostly 1 h. The strain changes were caused by a shallow pressure source less than 1.5 km. The inflation occasionally continued for more than 7 h with an addition of inflation of a deep source (4 km), which corresponds to the magma reservoir beneath Minamidake. The conduit to the Showa crater may be branched from the magma reservoir beneath Minamidake or from the major conduit connected to it. When inflationary ground deformation progressed at a high rate, the eruptive activity reached a peak from December 2009 to March 2010. This suggests that the accumulation of magma beneath the central cones of the Sakurajima volcano progressed simultaneously to a discharge of magma. The simultaneous progress of the accumulation and discharge of magma and the frequent occurrence of small vulcanian eruptions may be related to the small open conduit.

View full abstract

The recent vertical ground deformation in Sakurajima volcano and around Aira caldera associated with the volcanic activity of this volcano is revealed by the repeated precise leveling surveys conducted in October-December 2007, November 2009 and April and November 2010. The ground uplifts are detected in Sakurajima volcano and around Aira caldera to be centered in the caldera during the period from 1996 to 2010, as the previous results during the period from 1991 to 1996. From the analysis based on a spherical source (Mogi’s) model, the inflation source is located at 8.8 km - 10.8 km depth with the volume change rates of 6.5-8.2×10⁶ m³/year beneath the center of Aira caldera generally through the period of 1996-2010. It is indicated that the magma storage at the inflation source inferred at 10 km depth beneath Aira caldera is progressed during the period. In the period of 2007-2009, a shallow inflation source is located at 4.3 km depth with a volume change rate of about 0.6×10⁶ m³/year beneath the northern part of Sakurajima. It suggests the magma movement towards shallow part of Sakurajima volcano from 10 km depth beneath Aira caldera, although the estimated amount of magma input is as small as about 2.0×10⁶ m³. The magma storage at the magma reservoir beneath Aira caldera has continued since around 1991 when the eruptive activity of Sakurajima volcano was gradually decayed. Explosive eruptions are increasing at Showa crater in Sakurajima volcano especially since 2009. However continuing ground uplifts are observed until November 2010. It is suggested that the amount of supplied magma overcomes that of ejected magma at the magma supply system beneath Aira caldera in spite of the increasing volcanic activity. Considering the estimated volume increase at the inflation sources and the volume of the ejected magma based on the observed amount of the ash-fall deposits, it is indicated that the total of about 1.2×10⁸ m³ magma is inferred to have additionally stored beneath Aira caldera during the period from 1991 to 2010. The ground uplift around the northern part of Sakurajima caused by the progressing magma storage at the time of November 2010 recovers and further exceeds the height level in around 1973, when the intense summit eruptions during the 1970s and the 1980s started. These results suggest the immanent potential of the next intensive eruptive activity of this volcano.

View full abstract

We describe absolute gravity measurements performed from April 2009 through January 2011, and present technical suggestions for carrying out continuous observations in a volcanic area. The results clearly show significant gravity variations of as large as 30 μgal during the observation period. Hydrological simulations reveal that about half of the gravity change is attributable to groundwater disturbance. After correcting for this disturbance, the observed variations in gravity can be divided into 5 separate phases. Phase I is a period with few eruptions, which extends from April to late June 2009 when an abrupt 10 μgal gravity decrease was observed. During the succeeding phase II, from July 2009 to May 2010, gravity oscillated about a mean value with an amplitude of 5 μgal, while the monthly number of explosions at Showa crater dramatically increased from 50 to about 150. In phase III, which was a transient quiescent period, gravity increased by as much as l0 μgal in a single month. This was followed by phase IV, during which there was a steady gravity decrease until November 2010. During the final phase V, gravity remained almost constant until at least January 2011. These five phases are closely linked to the eruptive activity at Showa crater. In fact, excellent correlations are found among the records of absolute gravity, ejected weight of volcanic ash, ground tilt, and infrasound air shock amplitude. The gravity data are transformed into changes in magma head height using a simplified line mass model.

View full abstract

Showa crater of Sakurajima volcano became active in June 2006 after 58 years of quiescence. From multi-parametric geophysical observations, we have identified the processes that typically occur prior to an explosive eruption at the crater. A few hours prior to the onset of an eruption, magma starts to migrate and accumulates at a depth of about 1 km. This accumulation of magma can be clearly observed in strain change records as an inflation process. Several tens of minutes prior to an eruption, the SO₂ gas emission rate gradually decreases, indicating that a sealing process is taking place in the crater bottom as the eruption nears. During the same time period, the volcano's inflation rate starts to accelerate due to the formation of a plug above the conduit that prevents the gas from escaping, with the result that a gas pocket forms beneath the crater. In nighttime events, a volcanic glow is also seen, which weakens and then disappears. A few minutes prior to an eruption, a small tremor starts to occur. Its amplitude grows as the strain changes from inflation to deflation as the stored gas is released through new fractures within the plug that had been confining the gas pocket, leading to a minor depressurization in the conduit. Then, an expansion process starts, that could explain seismically the first motion of an explosion earthquake. This is probably when the effect of depressurization downward from the crater bottom reaches the magma head and a sudden magma expansion with degassing starts. After a short period (about half a second), this expanding magma rises and pushes the gas pocket upward, leading to a swelling of the crater ground along with the radiation of the preceding phase of infrasound waves, and then a breakup occurs. After the plug fails due to deformation, the accumulated gasses and expanding magma are ejected together from the crater as the surface eruption phenomena starts.

View full abstract

We conducted SO₂ flux measurements at Sakurajima volcano, Japan during 2007-2010 to ascertain differences in the degassing activity at two craters of the volcano: Minamidake and Showa craters. Recent developments in SO₂ visualization techniques using a UV camera observation system enabled us to quantify the SO₂ flux from each crater. SO₂ flux reflects conditions of the volcanic conduit. Therefore, a separate estimate of the gas fluxes from different vents is useful to gain insight into conditions within the volcano. Sulfur dioxide flux from Showa crater ranged from a few hundred to several thousand ton/day. In contrast, sulfur dioxide flux of 100-500 ton/day from Minamidake crater remained at a lower level. These results suggest that degassing conditions (i.e. permeability of the conduit or the amount of degassing magma) of Minamidake crater have remained stable. In contrast, the degassing activity of Showa crater is probably variable and sensitive to volcanic activities in the crater. The difference implies that drastic variations of Sakurajima volcano's SO₂ flux data, observed using a conventional traverse method with a UV spectrometer, resulted from the Showa crater’s SO₂ flux, not the Minamidake crater’s SO₂ flux.

View full abstract

On June 2006, Sakurajima volcano in Japan resumed its eruptive activity at Showa crater in Minamidake. We investigated dated ash samples, including lapilli ejecta of several eruptions, to evaluate the temporal relationship between their petrological features and eruptive activity. The activity is characterized by low frequency of explosive eruptions accompanied with small, periodic inflation from June 2006 to August 2009. No juvenile materials existed in the ash during this period (1st period). Frequent explosive eruptions accompanied with large, continuous inflation had continued from September 2009 until March 2010 (2nd period). Juvenile materials such as unaltered scoria and pumice with fresh glass have been recognized since late September 2009. The whole-rock chemistry of these juvenile materials is similar to those from AD 1955 to AD 2000 but is the most mafic (SiO₂=58.5-59.1 wt.%), indicating that the magma system, in which mafic magma was injected into silicic magma, has not changed. The matrix glass compositions of the juvenile materials are dacitic (SiO₂=67.2-72.7 wt.%). During the 2nd period, the proportion of juvenile materials in eruptive ash increased and the SiO₂ contents of glass decreased with time. These temporal changes suggest that the high level of eruptive activity during the 2nd period was caused by the increase of mafic components in eruptive magma. During April to May 2010 (3rd period), the number of eruption had become small without inflation, and the proportion of juvenile materials also decreased. From June 2010 (4th period), although the mode of crustal deformation has changed to deflation, eruptive activity has increased again and the juvenile materials have been mainly ejected. The glass compositions during this period have become slightly higher in SiO₂ content. These temporal changes of eruptive style and glass compositions suggest that the eruptive activity since AD 2006 has not been directly affected by the addition of mafic components in the magma system, but has occurred because of the silicic (andesitic) magma already-supplied beneath the volcano. In this way, the monitoring of the petrological features of dated eruptive materials could provide us useful information to evaluate ongoing eruptive activity as well as geophysical monitoring.

View full abstract

To understand magma degassing processes near the top of volcanic conduit, we investigated a series of volcanic ashes from the Sakurajima volcano, Japan. We describe temporal changes in the color and the amount of leachates (Cl, F, S) of ash erupted from 1981 to 2011. Based on the amount of leachates present, ash samples are classified into two major groups: one is with a molar S/Cl ratio of ～10, and another with S/Cl of ～1 and that is relatively depleted in S. Ash samples that were erupted during 1981-1991 from the Minamidake summit crater belong to the latter group. Ashes erupted from the Showa crater in early 2008 belong the former group, of which the S content was found to decrease systematically through time, although in 2011, ashes of the latter group erupted for the first time from this crater. Based on coloration, the ash samples in this study are classified into two groups: one with a yellowish color, another with a less yellowish color. The coloration of the former group can be explained by the existence of yellowish native sulfur as well as other hydrothermally altered minerals. We observed positive correlations between the interval of successive eruptions and both the yellowness and the amount of ash leachates. Our interpretation is that the observed temporal changes in volcanic ash result from a transition in the amount of fumarolic sulfur accumulation in partly solidified magma near the top of volcanic conduit, which we interpret to reflect the mean residence time of the magma. The magma probably periodically renewed in response to ash eruption and/or magma convection near the top of the volcanic conduit.

View full abstract

We performed refraction analysis for the first P-wave arrival time data observed in the seismic experiment in 2008, and estimated a shallow velocity model up to 3 km depth beneath the Aira caldera and Sakurajima volcano. We found that a basement layer with a velocity of 4.6-5.0 km/s, which corresponds to geologically the Shimanto Group, inclines toward the central part of the Aira caldera. A low velocity zone with a velocity of 4.2-4.4 km/s is located in a depth range 1.5-3 km in the central part of the caldera. This low velocity zone suggests high activity of the magma plumbing system from the deep magma reservoir distributed beneath the caldera. It is found that the basement layer steeply falls down from 1 km to 2.5 km in depth along the northwestern boundary of the Kagoshima graben. The velocity structure in Sakurajima volcano is characterized by a zone with a velocity of 3.6-3.7 km/s. Moreover, we present a possibility that the underground structure strongly restricts an expanse of a focal region of each different type of the volcanic earthquakes.

View full abstract

Seismic reflection structure in the northeast part of Sakurajima Volcano is discussed down to ca. 11 km with three profiles. Data acquisition was conducted along intersecting two lines with 221 stations and eight shot points. The two lines covered the northeastern portion of the volcano. One of the lines was spread along NNW to SSE direction in the east foot, and another line was oriented east to west in the northern flank of the volcano. Data processing were made through the conventional procedure of the reflection seismology, and depth migrated sections were obtained. Four continuous reflection horizons appear in the profiles. One of the reflection horizons around 5.4 km depth disappears in the northeast portion of the coverage area. The interruption of the horizon suggests existence of magma or magma supply path beneath it. The inferred magma supply path locates to the west of that previously presented.

View full abstract

An audio-frequency magnetotelluric (AMT) survey was conducted at the foot of Sakurajima volcano in November 2007. This survey was carried out within the framework of the 7th National Project for Prediction of Volcano Eruptions. The main objective was to clarify the shallow layers of Sakurajima volcano for better understanding of the volcanic activity. We measured electromagnetic fields at frequencies from 1 to 10400 Hz at 27 locations along only three lines set on the northern, western, and southeastern flanks of the volcano. The three profiles enclosed the summit area where access is prohibited because of frequent explosive vulcanian eruptions. The data quality was generally considered to be good at frequencies higher than 2 Hz. Two-dimensional and three-dimensional (3-D) modeling approaches were applied to the data. As a result of 3-D modeling, we obtained the following features in the resistivity model. The surface layer showed a high resistivity from several hundreds to thousands of ohm-meters and corresponded to the lava. This surface layer covered a highly conductive layer with a resistivity of less than tens of ohm-meters, interpreted as containing seawater or groundwater. However, the basement structure was not clearly detected, except for in an area of northern Sakurajima. The first lava layer was distributed thickly in northern and western parts of Sakurajima, and the boundary between the first and second layers was typically located at around sea level. The conductive second layer was found at deeper levels beneath areas such as the Haruta-yama lava dome and the Nabe-yama pumice cone. This depression of the low resistivity layer is probably caused by the fracture zone associated with previous eruptions, which is formed above the conductive layer. In the proximity of Hikino-hira lava dome and the Taisho crater, the second conductive layer was found at a depth shallower than sea level, suggesting that a hydrothermal system has developed underneath this area. In contrast, elevation of the second layer was not seen on the side of Taisho crater located nearest to the edge of Nabe-yama, implying that a heat source is not present beneath the eastern foot of Sakurajima.

View full abstract

An ocean bottom seismometer (OBS) recorded obvious semi-diurnal periodicity of the average velocity amplitudes of ground motions on the seafloor of Wakamiko Caldera (an active submarine volcano) northeast of Sakurajima Volcano in southwestern Kyushu, Japan. The ground motions were probably generated by the activity of the caldera’s seafloor fumaroles, because we found bubbles ascending from those fumaroles just after deployment of the OBS. We compared changes in root-mean-square ground-velocity amplitudes in one-minute windows (RMSAs), tidal gravities (accelerations), and water levels during the observation period to obtain the characteristics of the periodicity. Those characteristics are summarized as follows: 1) We observed clear semi-diurnal periodicity of the RMSAs throughout September, 2007, though sometimes the periodicity was less obvious. 2) The timing of maxima RMSAs corresponded to maximum tidal gravities in the time domain. 3) The frequencies of four peaks seen in the power spectra of the changes in RMSAs were identical with those of the four major tidal components. 4) In detail, changes in RMSAs show saw-tooth shapes, and are irregular in periods of diurnal inequality. 5) Long-term or irregular changes in fumarole activity are possibly dominant in the period. The activity of hydrothermal fluids ascending from the deeper portion toward the seafloor fumaroles, which make up part of the circulation of a hydrothermal system, could be advanced as increasing upward tidal gravities (accelerations).

View full abstract

We propose a ballistic equation of ejected volcanic blocks that is renovated by considering air resistance, wind velocity and direction dependence of the initial velocity. In the equation, we treat the air resistance as a vector that coincides with the opposite direction of flight. By comparing the calculated results based on the equation of motion and the observed data, such as spatial distribution of landed blocks, initial velocities by successive photos and so on, we tried to reproduce kinematic aspects of the ejected blocks at the 1982 and 1983 eruptions of Sakurajima Volcano, Kyushu, Japan.

View full abstract

In studies of volcanic tephra, it is usual that the overall volume of tephra is estimated ashfall volumes based on representative locations within the ashfall area. The precision of the volume estimation largely depends on the number of the locations. However, in the case of ongoing eruptions on island volcanoes, such as Sakurajima volcano, the observation locations are usually limited. We therefore have developed a practical method for estimating ashfall volume and distribution in such case. The method approximates the distribution of ashfall as ellipses, with the distribution area (A) and thickness or weight of deposit (T) determined by A=αT^-1. The ellipse-approximated isopachs can be determined by using the direction of the ellipse axis and ashfall data at two points. In determing the ellipse axis exactly, we usually need additional ashfall amounts from the other locations. We set 37 samplers around Sakurajima volcano, and retrieved the samplers 15 times, from April to December, 2008. Using the propose method, we are able to determine the volume of ash produced by small, continuous eruptions.

View full abstract