津久井 雅志

科研費基盤C「江戸時代の史料に基づく噴火履歴の高度化と噴火シナリオ作成」（平成24-28年度）<br />

Old historical documents on 1779 AD An'ei eruption of Sakurajima, southwest Japan were collected from distal places as well as those from neighboring area of the volcano. These records revealed that the ash-fall front traveled to the northeast at about 50-100km/h, reached as far as Tohoku district 1200km from Sakurajima, and covered area of ca. 2.33×10^5km^2. Investigation of old documents helped to improve understanding of behavior of the volcano and environmental effects at the time of infrequent and great eruption. The wide distribution of ash-fall in 1779 Sakurajima eruption suggests that there is a high potential that ash discharged by future eruption of Sakurajima may cover down through the mainland of Japan. We should keep in mind both physical and economical effects of ash-fall in assessing the activity and making the scenario of an eruption.

日本火山学会論文賞

Inflation of Izu-Oshima volcano has been observed since soon after the latest middle-scale eruption occurred in 1986. Such deformation is understood as re-storage of magma beneath the volcano. Because more than 230 years have passed since the last large-scale eruption in 1777, Izu-Oshima is overdue for an average recurrence interval of large-scale eruptions, 100 to 150 years. Preparing for the forthcoming eruption, we aime to improve the resolution of volcanic activities in historic time. Archives on historical eruptions, as well as topography, and stratigraphy of the volcano are reviewed. The results are summarized as follows. (1) Based on historical archives as well as stratigraphy of the volcano, we reviewed sequences of Y_3 (1552 Ten'bun), Y_2 (1684 Jokyo) and Y_1 (1777 An'ei) eruptions. All these three eruptions occurred inside the caldera and proceeded in the order of scoria fallout-lava effusion-ash fall, which is the typical sequence in Izu-Oshima. Duration of main eruption stage, start from basal scoria eject to lava effusion, ranges from 1 to 2 weeks in Y_3 and Y_2 eruptions, to 14.5 months in Y_1 eruption which is bigger than the former two. Those of magma-withdrawal stage, ash-fall stage, lasted 6 years in Y_2 and 9 years in Y_1, respectively. (2) In Y_1 eruption (1777-1792), the local government of Izu grasped the sequence of volcanic activities by frequent reports from the island and by inspections of government officers. The local government submitted reports and mitigation programs to the central government. (3) In case of future large-scale eruption occur in caldera, accurate information of magma-head level is indispensable for the prediction of eruption behavior. (4) We pointed out the significance of topographic lows on the northwestern and the northern flank to predict the flow of lavas.

Based on geological and archaeological data as well as historic documents, we review eruptions and earthquakes which have occurred during the 9th century in central and east Japan. The results reconfirm vigorous activities on Izu Arc, Izu-Oshima(〜838AD < N_3, N_2, N_1 < 886AD), Niijima(〜857AD and 886AD), Kozushima(838AD), Miyakejima(832AD and 850AD), and at Fuji volcano(800AD, 838AD < < 864AD, 864AD) during the 9th century. Beside these events, a big eruption of Niigata Yakeyama volcano had likely occurred in 887AD. Chokai volcano also erupted in 871AD, and 810-823AD. Collapse of Yatsugatake volcano took place in 887, probably invoked by a strong shock in 887. In addition, earthquakes with a magnitude from 7 to 8 had taken place along the Itoigawa-Shizuoka active fault system (in 841 or 762AD), Nagano fault system(887AD), Echigo plains(863AD), Shonai plains(850AD), Akita plains(830AD) and Nankai trough(887AD). As a result, we can point out a linkage of big eruptions and seismic activities in the 9th century over 800km long crossing Japan Arc. Geologically this seems to be a surface expression of East-West compression along eastern margin of Amurian Plate over 800km which was driven by the eastward motion of the plate. It is noteworthy to mention that the very similar volcanic and seismic activities have been occurring in the last 50 years to those happened in the ninth century, in their source areas and manners.

Major and trace element abundances, Sr and Nd isotopic ratios, and ²³⁸U-²³⁰Th radioactive disequilibria have been analyzed for samples from five volcanoes, Oshima, Miyakejima, Niijima, Teishi Knoll and Fuji, located in the northern Izu arc to investigate across-arc and regional variations of chemical compositions of mantle and fluids expelled from a subducting slab. The abundances and abundance ratios of both fluid mobile and immobile trace elements show across-arc variations. All samples but one have radioactive disequilibria, with (²³⁸U/²³⁰Th) greater than unity (herein, a ratio in parentheses denotes the activity ratio). The observed disequilibrium is similar to those reported for other arc systems, such as Mariana, Tonga, Kermadec, Chile. It originates from higher mobility of uranium relative to thorium during dewatering of the subducting slab. The degree of radioactive disequilibrium of (²³⁸U/²³⁰Th) decreases with depth to the Wadati-Benioff zone at each volcano. The radioactivity ratios of (²³⁰Th/²³²Th) of the samples from Miyakejima, however, deviate from an across-arc trend formed by Oshima, Niijima, Teishi Knoll and Fuji. The combined data of (²³⁰Th/²³²Th)-(²³⁸U/²³²Th) of the samples from the Izu and the Mariana arc revealed a well defined array in an isochron diagram with the exceptions of the samples from Miyakejima and Alamagan. This finding requires (²³⁰Th/²³²Th) of the sub-arc mantle beneath the arcs of Izu and Mariana except the magma sources of the two islands of Miyakejima and Alamagan was around 1.1 and relatively homogeneous before the final addition of fluid from slab. The homogeneous (²³⁰Th/²³²Th) necessitates input of Th with homogeneous (²³⁰Th/²³²Th) contributed from sediment melt, which dominates the budget of Th in the sub-arc mantle. It also suggests that influence of fluid addition before the final one had not caused a variation in Th isotopic compositions of the sub-arc mantle.

筆者はこれまでに、微量元素との比較から、島弧火山岩のBe同位体比（¹⁰Be/⁹Be比）と流体との関係を明らかにしてきた。本研究では、HFS元素の一つであるNbを用いて、伊豆島弧の流体のBe同位体組成および濃度に関して制約を与えることを目的とする。また、Brenan et al. (1998)の分配係数を用いたモデルから、放出される流体のBe濃度に関しても推定を試みた。

We carried out borehole drilling in the Yamakita area near an active fault of the arc-arc collision zone in the western part of Kanagawa Prefecture. Slime and core samples were obtained. The borehole geology was established by examined them using petrographic analyses and geophysical logging data. We found that the southwestern slope of Mt. Maruyama was down-faulted from the northern slope of the mountain by an unknown fault. After drilling was completed, the borehole was established as a Hi-net seismic observation station. Data are collected by the Hi-net observation system and made available through the INTERNET.

南関東足柄平野北縁の神奈川県山北町において，大都市大震災軽減化特別プロジェクト（大大特）により深度 2,035.4 m のボーリング掘削が実施された．足柄地域ではフィリピン海プレートが北アメリカプレートに対し部分的に衝突しながら沈み込んでいる，とされる． この 2 つのプレートは国府津‐松田断層，神縄断層，中津川断層系，塩沢断層系，日向断層などの衝上断層群で接しており，これらの衝上断層は全体として八の字形に配置している．掘削地点はフィリピン海プレートの北端に位置し，箱根火山の火山麓扇状地の上にある． 本研究では深度 0 m から 1,076.6 m までのオールコア試料を検討した．55 試料の全岩主成分分析と顕微鏡観察結果を基に，コアを 4 層序単位に分けた．最下部のグループI（深度 721.3 m 以深）は，年代的には足柄層群と同時期の堆積物であり，岩石学的に箱根火山とは異なる．一方，上位のグループII ～IV（深度 13 ～ 721.3 m）は箱根外輪山噴出物に相当する．したがって箱根火山の基盤最上部は海面下570 m にある．岩質と岩石学的観察からは，コア中に断層や層序単位の繰り返しを確認することはできなかった． これまでのボーリングのデータも含めて箱根火山の基盤岩の高度分布をまとめると，海抜 500 m 以上の尾根状の高まりが火山体中央部を南北に縦走し，東西山麓では海面下 370 ～ 700 m に没していることが推定された．