The old documents on the historical eruptions as well as the observational facts are precisely examined. Based on the carefull examinations, the typical features of Miyake-jima eruptions were clarified as follows : 1. Usually the eruptions started the flank of the main cone, without any clear precursors except in the activity of 1940. 2. In all cases, the earthquakes and rumblings became sensible only a few hours before the eruption in a small limited area where the eruption broke out. 3. The flank eruptions ended within a few days. 4. The Miyake-jima eruptions were frequently accompanied by post-eruption earthquake swarms. Occasionally these seismic activity caused damage to the houses and inhabitants of the island. 5. The flank eruption was sometimes followed by summit eruption which continued for a considerable period. In this case, post-eruption earthquake activity was not violent. 6. It seems that the occurrence of the eruptions of miyake-jima volcano in historical times were basically under the control of a periodicity of about 22 years.

View full abstract

Relation between the ambient crustal stress field and the distribution of flank eruption sites of an arc front volcano, Miyake-jima is studied. Compared with the adjacent arc volcanoes such as the northerly located Oshima volcano, two characteristic features are observed. 1. the strike of the flank eruptive fissures shows a remarkable bent towards either NW or SE as was the case in 1962 and 1983 fissures. 2. flank erutption sites are widely distributed in Miyake-jima in nearly all the azimuthal directions around the summit crater, whereas they are confined in a narrow straight zone passing through the summit on the northerly located Oshima and Hakone volcanoes. These features indicate the crustal stress field having, 1. northwest-southeast trending Sh_max and 2. horizontal deviatoric stress smaller than in the northern area where the above-mentioned two volcanoes are located. The earthquakes in the region of the present concern are of mostly strike-slip type. The above volcano-inferred stress field is consistent with the earthquake studies. Similar orientation of the Sh_max (P) and Sh_min (T) is indicated by the fault plane solutions of earthquakes. Smaller derivatoric stress in the Miyake-jima area than in the northern area is implied by the existence of normal fault component in the fault plane solutions and by temporarily variable principal stress axes of the quakes. The above nature of the crustal stress field, as inferred from both volcanoes and earthquakes, is understood with the plate tectonic setting of this delta-shaped region which is the northernmost portion of the Philippine Sea plate. The Philippine Sea plate subducts along the Suruga and the Sagami troughs which limit the northwestern and northeastern edge of the delta, respectively. The northern apex of the delta marks the site of the collision to which is attributed the origin of horizontal P axis of the quakes in this region (strike-slip type) rather than horizontal B axis (normal fault type) as is normal for bending earthquakes associated with subduction. Therefore the Sh_max (P) is expected to decrease southward where it is farther from the origin of stronger lateral constraints. Sh_min (T) is also expected larger in the southern Miyake-jima area than in the north because of the larger distance from the troughs where the Sh_min should be the smallest. Thus, the horizontal deviatorie stress (P-T) would be smaller in the south. Inflation and deflation of the entire edifice of Miyake-jima volcano appear to have occurred at least associated with the 1940 and 1962 eruptions. No data are available for the 1983 event, however. Deformation study at Miyake-jima will be useful and necessary for understanding the process of magma storage beneath summit crater and for prediction both for the timing of cessation of the once started eruption and for the preestimate of the amount of the eruptible magma.

View full abstract

Miyakejima volcano erupted at 15 : 15 hours on October 3, 1983 to form fire fountains along a 4.5 km long fissure. The first outbreak was at the altitude of ca. 450 m above sea level, without any noticeable precursors except for felt localized earthquakes. Spinose basaltic scoria was drifted eastward to form a thin blanket across the island while heavy fountaining produced several streams of aa flow one of them reached sea shore. Another flow heading westward entered Ako village on the western coast with 500 houses. About 430 houses were buried or burned by the lava within several hours. The fissure increased its length northward and southward at a rate of 40 m per minute for the first 30 minutes. It stopped at the altitude of 505 m on its northern end but spread southward at a rate of 26 m per second for another one and half hour to reach the south shore. At its southern end below 100 m altitude, violent phreatomagmatic explosions produced large pits and a tuff ring as well as heavy scoria fall over the southeastern shore of the island. The falling blocks were poorly vesiculated subrounded scoria which broke wind shields of the cars and gave heavy damage on crops and houses. The largest earthquakes of M=6.2 occurred at 22 : 33 hours and the eruptive activity in the lower vents was resumed to form a low scoria cone. The eruption ended before 06 : 00 hours on October 4. The mode of eruption, shape and size of the eruptive fissure, magma chemistry, duration of activity and the amount of magma erupted were more or less comparable between the last three eruptions of Miyakejima, i.e., 1940, 1962 and 1983.

View full abstract

A swarm of earthquakes, precursory indication of the eruption, began to be recorded at Miyakejima Weather Station at 1358 on 3 October, 1983, although none was recorded at any other seismic station of JMA until the onset of the eruption. On the other hand, another swarm of earthquakes that resumed after the onset of the eruption was caught by seismographs around Miyakejima. Precursory earthquakes were grouped into two types, i.e., high frequency earthquakes and low frequency ones, according as their predominant frequencies were higher than 2.5 Hz or not. It is unpromissing to try to determine hypocenters of precursory events, but, both the first motions of 5 high frequency earthquakes and the particle motions of 2 low frequency ones in the horizontal plane inidicate that their sources were located to the SW of the seismograph, probably on the island. Gradually increasing continuous tremors started immediately after the earthquake at 1522 which had a predominant frequency of about 1.4 Hz from initial motion through coda. The following continuous tremor had almost the same predominant frequency. Major eruptive activity probably began with this low frequency earthquake. The magnitudes of two large precursory earthquakes were estimated to be about 3.0 by applying the relation between the magnitudes of post-eruption earthquakes and their maximum amplitudes or duration times of vertical component at Miyakejima Weather Station. However, this estimation was not appropriate because earthquakes of such size were large enough to be recorded at seismic stations other than Miyakejima. The seismograph at the sea bottom off Omaezaki (named “TK1OBS” in the seismological bulletin of Japan Meterological Agency), about 180 km W of Miyakejima, detected post-eruption earthquakes of magnitude about larger than 2.5, but did not record any pre-eruption earthquakes. The background noise on 3-4 October had remained at a similar level of 0.02 milikine, which corresponds to the expected maximum velocity of the vertical component on TK1OBS when an earthquake of magnitude 2.4 occurs at Miyakejima. Therefore, precursory earthquakes seem to be of magnitude less than 2.4.

View full abstract

Eruptive activities of Miyakejima Volcano are studied with reference to tectonics, earthquake swarm activities around the volcano and their focal mechanisms. The earthquakes in this region show a tendency to occur in swarms. Near the volcano, there exist three tectonic ridges with strike of NE-SW ; (1) Niijima-Zenisu ridge, (2) Miyakejima-Onoharajima ridge and (3) Mikurajima-Inanbajima ridge. In the period before the eruption of the volcano, the earthquake swarm tended to concentrate in a narrow part of the ridge (1) or (3), and in the next period, in other part of them. In the course of earthquake swarm activities, the regions of ridges (1) and (3) were extensively occupied by the seismic focal region. The eruption of the volcano started after seismic gap had grown up around the volcano, and after the eruption, a part of the major seismic gap was filled by focal regions. This cycle was commonly observed before and after the eruptions in 1940, 1962 and 1983. Focal mechanism solutions for major shallow earthquakes in ridges (1) through (3) are redetermined in order to study the tectonic stress in this region. Most solutions indicate approximately a pure strike-slip, and a predominant direction of a dislocation closely resembles with the strike of the tectonic ridges which constructed by the volcanoes. Volcanic eruption may have also been caused by the same stress force in the crust. Another phenomenon commonly observed is the complementary relation between the seismic activity just after an eruption and the eruptive activity ; itself seismic activity was at high level when an eruptive activity continued for a short time (approximately two days or less), while it was quite low when an eruptive activity continued for a longer time.

View full abstract

Immediately after the 1983 eruption of Miyakejima, we set up temporary seismic stations on the island. This paper reports some characteristics of the seismic activity after the 1983 eruption and special features of wave forms of volcanic earthquakes observed at these stations. The seismograms recorded at Miyakejima are classified into three types as follows : (1) high-frequency earthquakes. These are classified into two subgroups : (a) high-frequency earthquakes having the same wave form as ordinary tectonic earthquakes with clear P and S phase ; (b) high-frequency earthquakes with no clear S phase. (2) low-frequency earthquakes which show dominant frequency of less than 10 or 15 Hz with no clear S phase. (3) low-frequency volcanic tremors which are also classified into two subgroups : (a) continuous tremor lasting several hours and (b) spasmodic tremor having a duration of a few minutes. The largest earthquake with M 6.2 occurred on October 3, about 7 hours after the biginning of eruption. The number of high-frequency and low-frequency earthquakes decreased gradually after the eruption and declined to only a few events per day on 5 November. On the otherhand. spasmodic tremor started to occur on 7 October. Such tremors were most dominant on 13 October and decayed gradually. Various wave forms of spasmodic tremor were recorded at temporary seismic stations. Hypocenters were determined for 48 high-frequency earthqaukes. Epicenters were concentrated within a small area of about 5 km in diameter, in the southwest part of Miyakejima. The focal depths were distributed between 2 km and 15 km. The push-pull distribution of P-wave first motions is very different from those of ordinary quadrant type earthquakes.

View full abstract

After the 1983 eruption of Miyakejima on Oct. 3, 1983, we carried out a temporary seismic observation at Miyakejima island from Oct. 6 to Nov.1. Four seismic stations were installed in the island. All the data were telemetered to and recorded at a key station. We have observed about 3400 seismic events in this period. The hypocenter determination has been done very precisely and extensively. This is resulting from the adequate configuration of seismic stations which were surrounding the epicentral region and from the high time-resolution due to telemetering observation system. Major results are summarized as follows ; (1) The seismic events are classified into four different categories as short-period earthquakes (Type 1), long-period earthquakes (Type 2), isolated volcanic tremor (Type 3), and continuous volcanic tremor (Type 4). The characteristic period of seismic waves for events of each category is about 0.1, 0.5, 1.0 and 0.8 s, respectively. The amplitude decay of seismic coda for events of each category is also different from one another. (2) Frequencies of both the short- and long-period earthquakes decreased gradually. The continuous volcanic tremor was observed in the morning of Oct. 7 for about five hours. A swarm of isolated volcanic tremor started on Oct. 9, and was gradually activated and attained to the peak activity on 13th. (3) The released seismic energy during the observation period was estimated to be about 3×10¹⁶ erg. The energy partition of the short- and long-period earthquakes and the isolated and continuous volcanic tremors were about 1/2, 1/20, 1/10 and 1/3 of the total, respectively. (4) Hypocenters of short-period earthquakes are distributed in four characteristic regions. Two major regions are located along the old caldera rim. The focal depths of these earthquakes are from 1 to 6 km below the sea level. One of the regions is very close to the newly formed fissure. The other is about 1 km NW of the former. These activities are probably controlled by the local stress strongly related to the structure of the volcano and the movement of underground magma. The third region which is not so active as the formers is around the summit in the depth of about 3 km. The fourth region is to the south-west of the island. Focal depths of earthquakes in the last region are rather deep as much as 15 km. (5) Hypocenters of some long-period earthquakes were determined. They were mostly distributed around the summit in depths shallower than 5 km.

View full abstract

A temporary seismic observation was carried out for about one month at Miyakejima island after the eruption on Oct. 3, 1983. About 3400 seismic events were recorded. They are classified into four different categories as short-period earthquakes, long-period earthquakes, isolated volcanic tremors and continuous volcanic tremors. Spectra of long-period earthquakes and continuous volcanic tremors have several distinct peaks, which suggest a mechanism that they are generated by volumetric sources. Isolated volcanic tremors are characterized by monotonous spectra and their source mechanism is considered to be different from that of the above two types of events. Active source regions of short-period earthquakes are along the old caldera rim of the Miyakejima volcano. One region is very close to the newly-formed fissure. Another is about 1-2 km northwest of the fissure. Almost all the earthquakes in the former region show dilatant first motions of P-waves at all the stations in the island, on the other hand, almost all events in the latter region show compressive first motions of P-waves at all the stations. The first motions of P-waves for those earthquakes cannot be explained by such a quadrant type mechanism due to the double couple source. In order to explain the distribution of first motions of P-waves, a tensile crack coupled with shear crack (a tensile-shear crack) is proposed as a source model of short-period earthquakes. Applying this model to observed seismograms, moment tensor elements of tensile and shear cracks are estimated. Results suggest that tensile cracking (opening or closing) is dominant for generating short-period earthquakes in comparison with shear cracking. Strike of tensile cracks is estimated to be nearly parallel to the strike of the newly-opened fissure. Earthquakes that occurred in the northwest region of the fissure are considered to be generated by a sudden opening of a tensile-shear crack due to the excess pressure of intrusive magma. On the contrary, earthquakes occurring beneath the fissure are generated presumably by a sudden closing of a tensile-shear crack, and this suggests that the magmatic pressure under the fissure rapidly decreased after the eruption.

View full abstract

After the eruption of Miyake-jima Volcano in Oct. 1983, magnetic observations were carried out by using proton magnetometers. At a reference point on the northern part of the island, about 3 nT increase of total intensity was observed during 3 months' period after the eruption. This implies expansion of the demagnetized area which is located at the central part of the volcano. Magnetic surveys conducted in Oct. 1983 and Feb. 1984 revealed a total field change well represented by a change in a dipole below the central cone, Mt. Oyama. By use of least squares method, this variation is explained by a change in the moment of a single magnetic dipole located below the north-western flank of Mt. Oyama at a depth of about 2.5 km below sea level. Since the variation occurred only in 2 months' period, we can hardly ascribe the cause of demagnetization to ordinary thermal conductance. A large amount of hydrothermal circulation may be its principal cause. At the northeastern tip of the fissure, geomagnetic measurements were repeated. About 10 nT decrease of total intensity was observed for one month. This decrease is explainable by cooling of surface ejecta as well as that of subsurface intrusion. Some portion of the observed change may be piezomagnetic origin due to closure of the fissure at depth which is associated with the drain back of magma at the post-eruption stage.

View full abstract

Electrical resistivity measurements by means of ELF and VLF MT (magnetotelluric) technique were carried out repeatedly after the eruption of volcano Miyake-jima on Oct. 3, 1983. On the scoria cones newly formed on the south-eastern side of eruptive fissures, VLF surveys were performed along four survey lines. Resistivity changes were detected, which relate to the cooling process of scoria. Cooling rate at each site has a good correlation with the volume of the erupted scoria. The VLF survey, at the north-eastern tip of the fissure along a line across the north-eastward extension of its trend, revealed no remarkable low resistivity structure even 3 days after the eruption. This implies non-existence of high temperature mass at a shallower depth within 10 m or so. One of the purposes of ELF・VLF measurements is to detect changes in the resistivity structure associated with the eruption by comparing the 1980 (YUKUTAKE et al., 1982) survey and the present one. Significant changes were found in the resistivity structure at two observation sites ; one was situated about 400 m north-west from the fissure, and another a fumarole on the central caldera floor. The former can be interpreted as caused by destruction of ground water system, and the latter as a result of hot water supply from the deeper part. The apparent resistivity at the fumarole site is still decreasing even 7 months after the eruption, suggesting the expansion of the hydrothermal system possibly established after the eruption. This low resistivity structure seems to elongate south-eastward from the fumarole site, i.e. in the direction of the eruptive fissure. A phreatic layer does exist beneath the other part of the central caldera, although its resistivity is an order of magnitude higher than the prescribed hydrothermal aquifer.

View full abstract

The Geographical Survey Institute revised the 1 : 5, 000 scale volcanic base maps of Miyake-jima because remarkable geomorphological changes were associated with the 1983 eruption of the Miyake-jima volcano. Leveling along the route skirting the coast of the island was also carried out after the eruption. The geomorphological changes can be revealed by comparing the maps with the ones published in 1981. The outflow consisting mainly of lava in the west of eruption fissures is estimated to be (7.0±2.0)×10^-3 km³ in volume. There are many elevation spots distributed on the side of the mountain. There is no recognizable change in their elevation exceeding the measurement error. However, appreciable ground subsidence is found around a new crater near the Shinmyo pond located in the southern part of the island. A comparison of pre- and post-eruption surveys shows that relative uplift took place in the northeastern and southwestern parts of the coast and subsidence in the northwestern and southeastern parts. The peak to peak displacement is about 24 cm. An almost opposite trend of deformation with much smaller amount is recognized for the period of several years befor the eruption.

View full abstract

Bench marks for precise levels were established along the circling route of Miyakejima Island in 1979 by the Geographical Survey Institute. Since then precise levels were repeated every year and a precise gravity survey was made at the same bench marks once in 1980 before its 1983 eruption. Both the surveys repeated after the 1983 eruption revealed the vertical displacement and the gravity change related to the eruption. During one year preceding the eruption, the island tilted 8.2 seconds of arc towards S 30° W as a whole and a small area ranging about 5 km in width including the eruption sites locally upheaved about 17 cm at the maximum. The gravity change observed at the bench mark of the maximum upheaval is about 65 μgal after the free-air correction. The pressure source responsible for the ground deformation is located at a depth of about 2 km on the rough assumption that the source is a point source. The anomalous mass responsible for the gravity change is determined as an infinite vertical dyke having density contrast 0.3 g/cc to the surroundings, width 7～8 m and depth 2～3 km below the earth surface, on the assumption that the dyke plane approximately coincides with the radial fissures which ejected lavas. It is probable that the pressure source and the distribution of the anomalous mass have the different configurations each other. Although gravity and elevation changes associated with the 1983 eruption of Miyakejima were surely detected, more detailed and accurate discussions remain rather difficult because the observation points are distributed only along the circling route, not covering a wide area of the island and the surveys have not been repeated frequently. We should prepare for the future study by intensifying the observation network, and repeating the surveys or adopting continuous observations.

View full abstract

The volume strainmeter, installed at the bottom of a drill hole at Izu-Oshima, an insular volcano, Izu Islands, Japan shows the characteristic phenomena such as pauses of its contraction tendency and of the intermittent occurrence of minor step-like recordings, which have been detected since the beginning of the observation in 1981, before the taking place of the 1983 October 3 Eruption of Miyakejima, about 80 km SSE of Izu-Oshima. Besides, the amplitude of earth tide recorded by the strainmeter has shown continuous increase almost without phase variations during the observation period. It is confirmed that there is no alteration of the characteristics of the instrumentation on sensitivity etc., through the analysis of the strainmeter’s response against variations of atmospheric pressure and calibrations of the amplification of the system. The phenomena are not seen at the neighbouring strainmeter stations, and there are no relationships between the notable phenomena and the variations of atmospheric pressure, ocean tide level and the temperature at the hole-bottom. It may be possible that the notable phenomena detected by the strainmeter at Izu-Oshima are generated by proceeding volcanic activity in a broad sense at Izu-Oshima or the alternative change of the nature of volcano edifice as the medium materials. However, it will be also possible that there are close relationships between the characteristic phenomena detected by the strainmeter and the active tectonic process in the region of Izu Islands, which shows frequent occurrences of seismic events and the outbreak of volcanic eruption since the end of 1982.

View full abstract

Surface temperature of lava flows, newly opened craters and discolored water on Miyake Sima was obtained at the altitude of 3000 m on Oct. 5, 1983, by the airborne thermal infrared radiometer (AGA Thermovision 780). Measurement range was selected at 2℃ for the area of discolored water and at 100℃ for the lava flows and craters, respectively. Surface temperature of discolored water was measured at 26.4℃, which was only 0.4℃ warmer than that of the ambient sea water (temperature was not corrected in any case). Surface tamperatures of lava flows and fumarole at Sinmyou ike showed complex variations ranging from 45℃ to 92℃ and 46℃ to 84℃, respectively. It is difficult to describe their fine structure because of the steam on lava flows, no correction data, IFOV (instantaneous field of view) of 3.4 milliradian of the instrument and so on.

View full abstract

The volcano Miyake is a volcanic island located at 200 km south of Tokyo. On October 3 in 1983, a fissure eruption took place at the southwestern flank after a quiescence of 21 years since 1962. Several kinds of thermal surveys were carried out before and after the eruption, and the following three items were examined. 1) Geothermal activity at the summit area before and after the eruption : In general, a flank eruption is considered to be caused by a dike intrusion from the central conduit, which connects the summit crater and the magma reservoir at depth. The height of magma head in the conduit is expected to reflect the volcanic activity and the regional stress and to be elevated gradually associated with a supply of magma before the eruption. At the summit area, fumarolic and steaming activities continued discharging thermal energy along the rim of the old crater, which was buried by the lava flow and the scoria cone of the eruption of 1940. It is also known that the thermal activity increased from November, 1963, one year after the eruption of 1962. Therefore, a precursory and/or post eruptive thermal anomaly is expected to occur at the summit area. 2) Cooling of the eruption fissure : The present fissure is about 5 km long, and lava fountains took place at the northern part, while magmatophreatic explosions took place at the southern part. The fissure also intercepts the wall of old caldera. Cooling rates at those areas are expected to reflect the amount of heat source, which remains at shallow part after the eruption, and the difference of geological settings. 3) Thermal anomaly at the area between the northern tip of the eruption fissure and the summit : If the present fissure eruption was caused by a dike intrusion from the central conduit beneath the summit crater, magma or some high temperature matter must have existed and most possibly remains under those areas and thermal energy of those matter may be released through the subsurface cracks possibly created around them. The results are as follows. 1) Geothermal activity of the summit area increased from 1963 to 1970, and turned to decrease since 1975. No precursory thermal anomaly was detected through the surveys before and after the eruption ; at the end of the August in 1983, one month before the eruption, and on Oct., 6, 3 days after the eruption. The summit thermal activity began to increase about 10 days after the eruption and the steepest increase took place during one month after the end of October. Thermal energy release rate was estimated to be at the order of megawatt at the end of November and found to have increased as twice as that before the eruption. These facts are consistent with the results of apparent electrical resistivity surveys and chemical analysis of fumarolic gas. These facts suggest that a hydrothermal system supporting the summit thermal activity is located at shallow depth and the height of magma head was sufficiently deeper than that of the hydrothermal system just before the eruption. 2) Fumarolic temperature decreased rapidly at the northern tip of the eruption fissure and more and more gradually at the southern sites, while that of caldera wall decreased extremely rapidly. It is also indicated that the radius of crater near the caldera wall is significantly larger than that inside the caldera. These facts may reflect the difference of mechanical structure of volcanic body. The difference of cooling rates were not clear between the areas which caused lava fountains and magmatophreatic explosions. 3) Measurement of ground temperature at 70 cm depth was carried out repeatedly along the survey lines perpendicular to the fissure direction at the area between the summit and the northern tip of the eruption fissure. It was found that each profile has two peaks of temperature, and that thermal anomaly elongates about 130 m from the northern tip toward the summit.

View full abstract

The volcanic eruption took place at the Miyake-jima volcano, one of the Seven Izu Islands, on October 3, 1983 after a dormant period lasting about 21 years. The volcanic activities lasted about 15 hours. A study was made on the detectability of the area covered with the volcanic products. Comparison was made between the images taken by Landsat MSS on October 25, 1983, 22 days after the eruption and November 11, 1980 about two years before the eruption. From the difference image of the two periods, selection was made of the area where the remarkable difference was seen in the visible or near infrared band, taking the kinds of land cover into account, and the changes in their spectral characteristics were investigated. Then, classification procedure to these two periods multi-channel images. Based on the analysis of the Landsat MSS data taken before and after the eruption, the total areas covered with the effused lava and the air-fall pyroclastics are roughly estimated as 1.8×10⁶ m² and 1.4×10⁷ m², respectively. It was found that the result comparatively well agreed with the regional distribution area of the volcanic products obtained by the ground survey right after the eruption.

View full abstract

During the 1983 Miyake-jima eruption pyroclasts with the total weight of 1.5×10⁷ ton were ejected, adding to lava flow with the total weight of 1.5×10⁷ ton. Pyroclasts were distributed eastwards and divided stratigraphically into 83 units through field observations. On the basis of the tephra-stratigraphical study on the pyroclasts and the observations including photographs and video, the process of the eruption was clarified as follows. (1) The eruption began with a scoria fall of the E-1 Unit from the E craters on the southern flank of the Oyama-Volcano about 15 : 15, Oct. 3. The fissure extended toward NNE and SSW soon after the openning. Active lava fountains from A to H craters along the fissure gave birth to lava flowing down the western slope of the fissure. On the other hand active fountains and pyroclastic columns formed scoria cones and Miike Scoria Fall (Fig. 4) on the eastern side of the fissure. Soon after 16 : 00, the fissure extended to SSW and I, J and K craters opened before 16 : 30. Lava flowed down to south from the fountains of I and J craters. (2) About 16 : 40, phreatomagmatic eruption occurred from P explosion craters beside Lake Shinmyo. P・Q-2 Scoria Fall (Fig. 7) of poorly-vesiculated scoria, the most voluminous pyroclasts of the 1983 Eruption, was ejected from the craters. The explosive phreatomagmatic eruptions occurred frequently from K, P・Q and S craters until about 20 : 00. R craters whose eruptions were characterized by Strombolian type, began to build the cone near the sea coast about 17 : 10. S crater opened in the sea about 17 : 15 and began to build a ring-shaped cone (tuff ring) composed of surge deposits (S-2 Unit, Fig. 8). About 17 : 30, the eruptions of K craters were replaced by active lava fountain, built a cone and originated lava flows. It suggests the exhaustion of the groundwater. (3) The reopening of A to S craters occurred about 20 : 00-21 : 00 after the short pause. The activity was dominated by small-scale Strombolian eruptions. Among them, however, R-4 Scoria Fall composed of well-vesiculated scoria (Fig. 8) was dispersed far eastwards. S-3 surge deposits indicate the reactivation of S crater in the sea. Other ejecta of this stage were bomb, fine vesicular scoria and ash, showing the wanning stage. The eruption finished before early morning, Oct. 4. (4) Throughout the eruptions from each crater, similar sequential changes of two cycles in lithological facies of pyroclasts and deduced types of pyroclastic columns can be recognized : lower column of the opening stage (fine, excellently vesiculated scoria and ash), active lava fountains and higher column (coarse scoria building cones and generally well-vesiculated scoria dispersing to distal area), middle column (scoria and ash fall dispersing to rather distal area), and in final stage of the second cycle, small-scale eruption ejecting mainly bomb. Pyroclasts, however, resulted from the interaction of magma and external water are characterized by the dominance of non- or poorly-vesiculated scoria. (5) From P・Q and S craters, cock’s tail jets and base surge clouds were observed to be spouting out pulsatingly. The deposits equivalent to those were identified around those craters. Especially, P・Q-4 deposits originated from cock’s tail jets of P・Q craters are characterized by the unsorted deposits composed mainly of fine ash and lithic fragments in block size and by a number of tongue-shaped microtopography.

View full abstract

Stratigraphy and thickness distribution of the pyroclastic fall deposits formed during the eruption of Miyakejima volcano on October 3-4, 1983, were studied immediately after the deposition. Of the total mass of 20 million tons erupted, 8.5 million tons were ejected as basaltic scoria to form a complex set of air-fall deposits east of the fissure vents. One million tons of the latter were ejected from the upper fissures as fire-fountain products. The rest was the product of phreatomagmatic explosions which occurred in the lower fissures where ground water chilled the magma to form dense scoria blocks which devastated villages. Explosion craters and a tuff ring were formed along the N-S trending lower fissures. Account of the general distribution of the deposits, nature of the constituents, mutual stratigraphic correlation and correlation with observed sequence are given.

View full abstract

On the eruption of 1983, Miyakejima volcano issued lava flows from the fissure, along which many scoria cones were formed at the opposite side of the flowing lava. The fissue is represented by an alignment of many small craters resulted from the drain back of lava to the vent. Some of them are a kind of pit crater formed by the marginal collapse of the primary craters. An explosion crater appearing at the skirt of the volcano and a tuff ring at the seashore were formed by the phreatomagmatic explosions. Difference in shape among the pyroclastic cones is mainly due to the difference in mode of eruptions, that is to say, the difference in ratio of water to magma. Only the explosion crater changed the mode of eruption from phreatomagmatic to magmatic one. Because most of the scoria cones are situated on the slope of the volcano and the volume of the ejecta from the vent is not so large, the volcanic edifice is not a typical, grown up scoria cone surrounded by the talus slope. Many cracks and fissures are opened on the surface of the scoria cones running parallel to the main fissure. Those are a kind of step-fault caused by the shock or the shaking of the earthquakes after the end of the main eruption. Some vents are filled up by the slide mass of the scoria cone. Viscosity of the lava seems to be rather low, and it thinly spreads over the wide area. Lava tree molds and natural levees are frequently found at the marginal part of the lava flow, and many kipukas are at the central part of the flow. The lava is mostly aa type. Only in the small area near the vent, the lava resembles slab pahoehoe, but the slabs themselves have spiny aa surface. The lava gradually changes into block type after a long travel through the lava channel. No lava flow formed by the secondary flowage of the agglutinate is observed. The writer discriminated a peculiar topography resulted from the secondary flowage of the already settled lava caused by earthquakes or collapse of scoria cones, and names it the re-mobilized part of lava. This type of movement may be common in the normal lava flow, but it is hard to distinguish the newly flowed part from the continuously moved lava. On the contrary, in the case of the eruption accompanied by the ejection of a large amount of tephra during or after the outpouring of lava, a peculiar lava topography formed in relation to the presence of tephra layers is usually recognized and it makes the discrimination of the re-mobilized part of lava very easy.

View full abstract

On October 3, 1983, the Miyakejima Volcano, a volcanic island in the Izu Seven Islands, suddenly started a fissure eruption. Eruptive fissures were about 4.5 km in length and ran from the southwest flank of the volcano, about 2 km away from the Oyama Central Cone, to southwest coast of the island during the first 2 hours. The eruption lasted only 15 hours from 15 : 15-15 : 20 and basaltic magma poured out accompaning lava fountaining along the fissures located on the flank higher than 50 meters above sea level and formed several flow lobes of aa lava. Phreato-magmatic eruptions occurred along the fissures near and on the coastal region to form large craters and a tuff ring more than 100 meters in diameter. Lavas poured out from craters A to E flowed down to the west on the slope of the volcano. The largest flow covered the most part of the Ako Village located on the west coast of the island. Lava issued out of craters H to K flowed down to the south and entered the sea near the Abe district. Many surface features on lava flow were identified such as lava wrinkles, lava levees, lava streaks and lava crevasses in vertical aerophotographs. Lava wrinkles appeared near the distal end of lavas. Lava levees are not remarkable except caldera floor areas, but lava streakes are observed characteristically throughout lava flows and commonly two streaks form a pair which represents faster flow inside them. There are many lava crevasses on relatively steep slope. They are thought to have been made by gravitational tension forces after lava supply from fissures stopped. From the eyewitnesses of the opening of fissures and estimation of effusion rate, lava and pyroclastics were fed from 2 or 3 dike systems which intruded laterally from the central conduit beneath the Oyama Central Cone.

View full abstract

A method of the numerical simulation for low viscous lava flows is presented. The lava flows over the digital topographic maps are formulated, expanding the simple solution of Navier-Stokes’ equation concerning the liquid of a constant thickness which flows downward by gravity on an inclined plane. The progress of the fronts of lava flows is, however, nonstationary process. Therefore, the concepts of the minimum thickness of lava flows are introduced to compensate the simplicity of our formulation. The minimun thickness is defined so as to depend on the inclination of the ground, the physical properties of lava and the minimum velocity of lava flows. The method of the numerical calculation was applied to the reproduction of the 1983 lava flows at Miyakejima, assuming the condition of the extrusion of the lava and their physical properties. The topographic map of Miyakejima was digitized with the sampling square of 25 m×25 m. The simulated lava flows fairly coincided with the actual ones in their path and width, when the viscosity of the lave was assumed to be 2×10⁵ poises. The extension of the reproduced lava flows with time agreed with the observation when the successive increase of the viscosity from 5×10⁴ poises to 1×10⁷ poises was assumed. This result suggests that the macroscopic viscosity of the lava flows increased by about one hundred times in their progress. The proposed method of the numerical simulation for lava flows may be available for predicting the paths and the extension of the forthcoming lava flows.

View full abstract

Temperature measurements of the lava of 1983 in Miyake-jima in the Ako district were started fifty days after the eruption and have been continued since then. The following three kinds of temperature data have been obtained (Fig. 2). 1. Temperatures at 20 cm depth along a graveled temporary road on the clinkery surface of the lava using mercury and alcohol thermometers. 2. Temperatures at 0.5 to 2.5 m depth in iron pipes inserted into the clinker layer using thermocouples and mercury thermometers. The pipe holes were distributed along the temporary road and at scattered stations on the surface of the lava. 3. We drilled a borehole (DH-1) which penetrates through 5.5 m-thick lava into the previous ground. Temperature was measured at 10 points in the hole using thermocouples. For comparison, similar measurements in the Awabe district were made in pipes with depths up to 2.5 m (Fig. 3). These pipes were buried in the holes dug into the massive part of the lava for electric poles. The temperature data at 20 cm depth and in the pipe holes (Figs. 5-10) indicate that isothermal surfaces in the clinker layer are very complicated. This complexity is explained by rising plumes of hot vapor irregularly present in the lava field. The vapor is produced by degassing process in the massive part of the lava and comes up through newly formed cooling joints. Once a cooling joint is formed, the temperature of the massive part of the lava around the joint fell rapidly because a gas plume effectively transports the heat from the massive part to the surface. But the rate of temperature decrease varies greatly from one station to another. New plumes were formed sporadically and the temperatures of the new plumes were much higher than the decreased temperatures of the older plumes. Some older plumes died out because degassing process ended or the joints were self sealed by sublimates. It is necessary to arrange a number of observation stations and to add stations timely in order to reveal a cooling history of aa lava like the lava of 1983 in Miyake-jima. Around a plume, a convection cell was identified in the clinker layer (Figs. 16-18), which is similar to a hydrothermal convection system usually found in geothermal areas. The change of the temperature-depth profile of DH-1 with time (Figs. 11, 12) clearly shows that the lava heated the underlying previous ground. The peak shape of the profile has become broader and the depth of the maximum temperature has steadily fallen. The change of the temperature-depth profile also suggests that the upper clinker layer prevented rainfall from effective cooling of the massive part of the lava for the first 250 days. During that time, raindrops were evaporated in the clinker layer and did not reach the massive part below the clinker layer. Difference of cooling rate between Awabe lava and Ako lava may be due to the difference of the thickness of the clinker layers (Figs. 15, 19).

View full abstract

The ejecta and lava flows of the 1983 Miyake-jima eruption are basalts with a few percentages of plagioclase and less abundant pyroxene and titanomagnetite phenocrysts. Olivine and hypersthene phenocrysts are rarely found. Variations in bulk chemistry of these ejecta and lava flows are essentially controlled by variations in modal amounts of phenocrysts. Both major and trace element concentrations in these basalts indicate that they are well on the differentiation trend of the Miyake-jima volcano and mildly differentiated. Based on the compositions of the coexisting pyroxenes, the temperature of the magma on eruption is estimated to be 1100℃ or less. Three different types of vesiculated xenoliths are recognized : Type 1 is the least vesiculated and is considered to be exotic materials, type 2 is most vesiculated and has dacitic compositions, and type 3 is composed of basaltic clots set in a well vesiculated dacite glass which is identical to type 2 and has andesite or basaltic andesite composition. Type 2 and 3 often form a composite xenolith. Type 2 xenolith is derived from dacite pumice once erupted on Miyake-jima and type 3 is the mixing product of the dacite pumice and basalt. Trace element chemistry also indicates that type 2 and type 3 xenoliths are genetically related with the Miyake-jima volcano. However, type 3 xenolith has exotic composition from the Miyake-jima volcano.

View full abstract

Petrographic, mineralogical and chemical studies were made for the lavas and pyroclastics of the 1983 eruption of Miyakejima Volcano. Essential eruptive products are phenocryst-poor (8 vol%＞) olivine-bearing augite basalt to basaltic andesite. Products from A and G craters are nearly aphyric with phenocrysts less than 2.5% in volume. Phenocrysts are plagioclase (core : An_92-82), olivine (Fo_68-66), augite (core : Wo₃₉En₄₆-Wo₃₅En₄₅) and titano-magnetite (TiO₂ 8.5-9.3 wt%). Groundmass consists of plagioclase, augite, sub-calcic augite, pigeonite, titano-magnitite, silica minerals and brown glass. From the compositions of groundmass pyroxenes, the eruption temperature is estimated to be about 1120℃. Chemical compositions of essential eruption products are on the differentiation trend of the post-caldera lavas of the Miyakejima Volcano, and their SiO₂ contents vary from 53.3 to 55.2 wt%. SiO₂ contents of A-E craters inside the caldera are 53.3-54.0%, and products from A craters represent their liquid compositions. Products from G-K craters on the southern slope of the precaldera edifice have higher SiO₂ contents, 53.6-55.2 wt%. Products from P-S craters around the sea shore, however, are SiO₂-poor again, 53.4-53.9 wt%, and have similar compositions to those of B-E craters. Above evidence suggests that A-E, G-K and P-S eruptive fissures erupted magmas with different chemical composition coming through different feeders, respectively.

View full abstract

⁸⁷Sr/⁸⁶Sr ratios were determined for 5 volcanic products (3 lava, 1 scoria and 1 xenolith samples) of the 1983 eruption of Miyakejima volcano. The rations were all within the range between 0.70350 and 0.70361, which agrees with those of older volcanic rocks (0.70351-0.70369 ; NOTSU et al., 1983). This suggests that the source material of Miyakejima volcano has not changed significantly since older stages till the 1983 eruption.

View full abstract

A red-hot lava was sampled from a vent fissure of the 1983 eruption of the Miyake-jima volcano, and was subjected to a non-destructive gamma-ray spectrometry within 36 hours of sampling. An observation of the specific activity of ²¹⁴Bi for several weeks indicated a temporal depletion of ²²²Rn possibly through the eruptive event and its growth to an amount in equilibrium with the parent ²²⁶Ra. On the assumption that the ²²⁶Ra-²²²Rn equilibrium had been established in the magma before eruption, the released fraction of ²²²Rn was tentatively estimated to be about 70%. The lava appeared to have continued releasing Rn until it was detached from the vent.

View full abstract

Miyakejima volcano erupted on October 3, 1983 after 21 years’ interval. This paper deals with a geochemical study on the volcanic gases, sublimates and rock samples collceted after the eruption. Immediately after the eruption, the chemical composition of volcanic gases indicates that these samples are mixtures of gases supplied from magma and residual gases in lava expelled by remaining heat. About ten days after, they show characteristic of the latter mainly. This change may be caused by the drain-back of magma. About five months later, remaining fumarolic gases are solely composed of CO₂. Volcanic sublimates found around the fumaroles were mainly salammoniac (NH₄Cl) and other minerals were a little in their variety and amounts. Insoluble chlorine and total fluorine contents in lava and bomb were determined. They show rather uniform distribution with only one exception of a vesicular volcanic bomb. Their insoluble chlorine contents are distinctly higher than those of the 1962-lava samples. Any evidence of the effect of the interaction of hot lava with seawater on halogen contents is not found.

View full abstract

Eruptions occurred on Miyake Island on the 3rd of October, 1983. An enormous amount of hot lava flowed from the middle slope of Mt. Oyama and also eruptions occurred at Shinmyo Pond and on the coast of the Pacific Ocean. During October 13 to 15, the authors collected water samples from hot and cold springs and lakes on Miyake Island. The analytical results are as follows : Tairo Pond showed 23.9℃ W. temp., 7.1 pH, 485 mg/l evaporated residues, 159 mg/l Cl and 76 mg/l SO₄ ; Shinmyo Pond showed 33.5℃ W. temp., 3.7 pH, 3377 mg/l evaporated residues, 1575 mg/l Cl and 531 mg/l SO₄ ; the new crater lake created by the eruptions showed 64.0℃ W. temp., 7.2 pH, 35965 mg/l evaporated residues, 19000 mg/l Cl and 2771 mg/l SO₄ ; a new pond created on the coast of the Pacific Ocean showed 56.0℃ W. temp., 7.2 pH, 38350 mg/l evaporated residues, 19750 mg/l Cl and 2583 mg/l SO₄. Shinmyo Pond was buried by the eruption which occurred at the edge of the pond on October 3, 1983. Therefore, the water in the pond was acid showing 3.7 pH and 1575 mg/l Cl, while before the eruption, it was alkaline.

View full abstract

The Miyakejima volcanic (basaltic) island is situated about 200 km south-southwest of Tokyo and has a round outline measuring about 8 km in diameter. This paper deals with the groundwater systems in this island. The results of our study can be summarized as follows : (1) Our hydro-geological study comprised the survey of geomorphology, geology, springs, wells, electric prospecting and test drilling in the island with the results shown in Figs. 4 to 8 and Table 1. (2) From these surveys, the groundwater of the island can be subdivided into main-groundwater system and sub-groundwater system. The distribution of the two types of groundwater system are shown in Fig. 9 with schematic profiles. (3) The main-groundwater system is considered to develop as Ghyben-Herzberg's lens within the main somma volcanic deposits along the coastal areas. It may not be distributed in inland area of the island owing to the presence of subsurface impermeable bed. The sub-groundwater system is comprised within shallow parts of somma deposits, the younger parasitic deposits, coastal gravels and sand deposits (Fig. 9). (4) Based on our study, new wells were drilled at four sites, where main-groundwater system was considered to be well developed. Total amount of fresh water measuring 45000 m³/day has been discharged by pumping from these new wells and has been used for water supply of the Miyakejima island. (5) At the time of the 1983 eruption, the phreatic explosions took place mostly underneath the main-groundwater system.

View full abstract

Two calderas are developed in the Miyakejima Volcano, one is on the summit area (ISSHIKI, 1960), and the other is on the western slope. The latter was first recognized and reported by the present authors (SHINDO, 1980 ; CHIHARA et al., 1973). The topographic and geological features of the caldera of western slope are summarized as follows : (1) On the western slope of the volcano, a contrasting drainage pattern (radial valley systems) is seen between the upper and lower slopes, the boundary line running along the contour line of about 300 to 350 m above the sea level (Fig. 1). All radial valleys cut deeply the deposits of older somma stratovolcano. On the other hand, the upper valley system is shallower than the lower valley system and both are independent and discontinuous. (2) Wide and gentle slopes are developed at the altitude of 300 to 450 m above the sea level inside of caldera. The southern flat area is called Kuwanokidaira. (3) Along the western side of the gentle slope, a caldera rim about 4.5 km long, is clearly traced, the height of caldera rim being 0 to 15 m. The amount of collapse of the inside is presumed to be more than 50 m. (4) The inside of the caldera is covered by the aphyric basalt lavas and reddish scoria bed of the younger parastic volcanoes. (5) The caldera of the western slope was formed presumably after the older main stratovolcano and prior to the formation of younger somma and the younger parastic volcanoes. (6) The 1983 basalt lavas flowed down along the caldera rim west of Kuwanokidaira. The lavas partly overflowed the lower parts of the caldera rim and followed just the same course as the 1643? basalt lavas. Therefore, construction of man-made bank in addition to using the natural rim might make it possible to protect the lava disaster for the villages, and at least to change the lava flow course into another valley. These lava flow controls should be planned in the future in the Miyakejima Volcano.

View full abstract

Short-lived but active eruption of Miyakejima in October, 1983, produced a 2.5 km long basaltic lava flow which devastated 350 houses on the western coast. Following the example of Heimaey eruption in 1973 in Iceland, an attempt to control advancing lava flow near the shore line by chilling with sea water was made using portable pumps. About 4700 tons of sea water was sprayed over the 300 m long flow front of the lava for 3 days (16.5 hours net). However, the lava apparently had stopped before the watering operation began and there was practically no movement during the operation. Thus the effect of water spraying on the movement of lava flows was left uncertain. Nevertheless, the operation helped both the scientists and administration in providing many valuable informations on strategies, equipments, methods, etc. for future lava control operations.

View full abstract