国際陸上科学掘削計画のもとオマーンオフィオライト掘削が2016-2018年に実施された．日本の研究者チームが主導して，1億年以上前の海洋底を形成していた海洋地殻・マントル境界を貫く掘削を進めた．その結果，ハンレイ岩層（海洋地殻最下部）からハルツバージャイト層（マントル最上部）へ至る連続コア試料の採取に成功し，地殻・マントル境界部の新鮮なコア試料（左）を入手した．肉眼でも確認できる光沢ある粒子の多くは，もともと中央海嶺下でカンラン石だった結晶が，二次的変質により蛇紋石化したものである．右：写真中央の掘削装置を現地にもち込み，最大で400 m長のコア試料を2本の掘削孔から採取した．また，掘削孔内の物性測定のために，2か所のロータリー掘削も行った．採取したコア試料には，海洋地殻最下部を構成していたハンレイ岩層からモホ遷移帯を貫き，マントル最上部を構成していたハルツバージャイト層が含まれる．今回の過去の地殻・マントル境界を貫く掘削は，現在進めている海洋底でのマントル掘削を目指す上での重要なマイルストーンとなった．

（高澤栄一・小野重明）

 It is now more than half a century since the Deep Sea Drilling Project (DSDP 1968-1983) was launched in the United States. The DSDP was followed by the Ocean Drilling Program (ODP 1983-2003) and subsequently by the Integrated Ocean Drilling Program (IODP 2003-2013). The program has been ongoing since 2013 as the International Ocean Discovery Program (IODP 2013-2023). In October 2020, the long-term scientific outlook to 2050 (Science Framework 2050) for a new deep sea drilling program was published electronically in a follow-up to the second phase of the IODP that will be completed in 2023. Deep-sea drilling science must be one of the most successful international collaborations in the Earth sciences, with more than 1,000 deep-sea scientific drilling sites operated to date, making it an essential approach for the direct exploration of the Earth's interior. The research products have promoted a scientific understanding of the ocean floor in a wide range of fields, including verification of the seafloor spreading theory of plate tectonics in the early days, elucidation of paleoenvironmental changes, and investigations of the subsurface biosphere. However, the so-called “basement drilling” of rocks (mainly basaltic igneous rocks), which are the main material of the oceanic crust covered by deep-sea sediments, has not been developed smoothly. In more than half a century of deep-sea drilling programs, only 38 holes more than 100 m have been drilled through the basement rock, and only 20 holes deeper than 200 m have been drilled. In addition, the total amount of marine crustal material recovered by basement drilling is less than 2% of the total depth of drilling. However, the structure and petrological properties of the oceanic crust have been gradually elucidated from available basement materials, although it is not easy to drill deep seafloor boreholes and to recover drilled rocks. An overview is presented of several important studies on deep-sea basement drilling to date in as much chronological order as possible, including a brief history of scientific drilling prior to the Deep Sea Drilling Program. Moreover, some plans are briefly introduced for ultra-deep basement rock drilling to the mantle in the future.

 In the 1950s, the aim of the original mantle drilling projects was to obtain oceanic mantle samples in order to address the unanswered question of what constitutes the Earth's mantle. However, in the 21st century, it is widely accepted that the uppermost mantle is mainly composed of peridotite. Now, the challenge of mantle drilling is to understand crucial unsolved issues of earth science. Today's Earth is different from other planets due to the existence of life and plate tectonics. It is emphasized that mantle drilling is the only way to obtain the oceanic crust from top to bottom and an active mantle sample from an oceanic plate. The crucial issues that can only be addressed by mantle drilling are: (1) limits of life in an oceanic plate and its controlling factors, and (2) formation process of an oceanic plate and its modification. Modification of an oceanic plate, especially the weakening of plate strength, is required for plate tectonics. These two issues are interrelated. Long seismic profiles of oceanic plates reveal the diversity of Moho seismic reflection regions: clear, unclear, diffuse and non-Moho regions. Faults and/or fracturing in oceanic plates and subsequent seawater flow can modify oceanic plates locally, probably causing the diversity of oceanic Moho, as well as the rheological behavior of oceanic plates. Fluid flows along faults/fractures also extend the biosphere of oceanic plates. The first drilling sample should be a reference to the oceanic crust and the uppermost mantle, and define the nature of the Moho at the site, as well as constrain reasons for the diversity of the Moho in other areas. Deep sampling, such as mantle drilling in an old oceanic plate, can penetrate the biosphere/non-biosphere boundary, which tells us about the controlling factors of the limit of life. This information may help us find extraterrestrial life. After mantle drilling is completed, the borehole is the only window from the ocean floor to the mantle. An in-situ mantle observatory in the mantle hole to monitor plate movement and fluid flow with biological activity within an oceanic plate is also suggested. Detecting geoneutrinos at the mantle site allows the amounts and distributions of radioactive elements from the Earth's mantle to be measured. These provide basic information on the Earth's heat sources and the evolutionary history of the mantle.

 The ICDP Oman Drilling Project carried out onshore drilling of the world's largest ophiolite, the Oman ophiolite (also known as Samail ophiolite). This drilling project provided an opportunity to explore major key boundaries of the oceanic lithosphere, represented by the Oman ophiolite, by drilling cores and boreholes. Below the layered gabbro at the bottom of the crustal section is the Moho Transition Zone (MTZ), which is mainly composed of dunite with small amounts of gabbroic sills. By drilling at the Wadi Zeeb CM site in the Wadi Tayin massif, cores were successfully collected from a 150 m MTZ. Also collected were fragile altered rocks from wadi outcrops that are easily lost. The core description campaign was carried out aboard deep-sea scientific drilling vessel “Chikyu” anchored at Shimizu Port. The core observations were performed and described according to the IODP procedure, and the analysis was conducted using many instruments. The resulting data provide important insights and will contribute to future drilling of the Mohorovičić discontinuity in the ocean. The most striking fact is that MTZ dunites are strongly influenced by serpentinization. In particular, the upper part of the MTZ just below the boundary with the lower crustal gabbro was most strongly altered, and a fracture zone was also developed. Understanding when and how these alterations occurred at the boundary between the crust and the mantle is an important future task.

 The Izu–Bonin arc has been the target of several hard rock drilling expeditions, including those associated with the Ocean Drilling Program (ODP) and the Integrated Ocean Drilling Program (IODP), as well as a drilling survey for delineating Japanese continental shelves undertaken by Japan's Ministry of Economy, Technology and Industry. In 1989, ODP Legs 125 and 126 successfully recovered cores from the Izu–Bonin forearc area and the backarc rift basin (Sumisu Rift). These cores provided the first opportunity to investigate the early volcanic and tectonic history of the Izu–Bonin arc in the Eocene and Oligocene. They also provided the earliest volcanic products of the Sumisu Rift, which is highly vesicular basalt. Drilling for the Japanese continental shelf survey was conducted in the most reararc side of the Izu–Bonin arc. The cores from the reararc seamounts reveal that the across-arc variations in magma chemistry and age of volcanism observed along the reararc seamount chains continue further toward the spreading center of the Shikoku Basin. The Kinan Escarpment appears to be the westernmost (i.e., most reararc) location where a slab-derived geochemical signature can be recognized in the erupted magma. Three IODP drilling expeditions have been undertaken in the Izu–Bonin arc region. Exp. 350 drilled in a small basin between the reararc seamount chains. Slightly over 1.8 km of sediment of mostly volcanic origin was drilled. This core preserves a continuous magmatic record in the reararc area since cessation of spreading of the Shikoku Basin, and provides critical information about how the reararc volcanoes were reestablished after the middle Miocene. Exp. 351 and 352 aimed to study subduction initiation processes. Exp. 352 was conducted in the Izu–Bonin forearc. It recovered forearc basalt (FAB) and boninites associated with the seafloor spreading at subduction initiation. Based on their ages and geochemical characteristics, fast- and short-lived seafloor spreading is estimated to have occurred. Exp. 351 recovered ocean crust, which is interpreted to be the basement of the arc, from the Amami Sankaku Basin between the Kyushu–Palau Ridge (ancient Izu–Bonin arc) and the Daito Ridge (Mesozoic remnant arc). This basement is similar in age and geochemistry to FAB, which implies the Izu–Bonin arc basement is ocean crust produced following the onset of subduction. This expedition also provided for the first time a continuous volcanic record of early reararc magmatism for the Izu–Bonin arc.

 A significant fraction of the ocean floor is created in back-arc basins, where water plays a major role in generating back-arc basin basalts, contrasting strikingly with magmatic processes at mid-oceanic ridges. Furthermore, much of our understanding of all of the oceanic crust comes from ophiolites, which are largely attributed to supra-subduction zone environments. Therefore, studying the back-arc basin lower crust and uppermost mantle is arguably important, contributing to the overall geology of the oceanic crust. The Godzilla Megamullion, located in the extinct Parece Vela Basin in the Philippine Sea, is the largest known oceanic core complex. It is an elongated massif with a distinct corrugated surface consisting of several individual domal highs. It records the secular evolution of the mantle melting beneath a dying back-arc spreading ridge along the length of the megamullion surface. Furthermore, strong heterogeneity in the P-wave velocity structure is observed along the length of the megamullion, with a normal oceanic crust-like structure in the distal (i.e., near breakaway) to medial parts, and a shallow high-velocity body in the proximal (i.e., near termination) part. The Godzilla Megamullion should arguably be the best place in the world to study the architecture of the back-arc basin lower crust and uppermost mantle, and the actual crust/mantle boundary through the International Ocean Discovery Program (IODP). By locating three 400- to 800 m-deep drill holes along its length, key data are obtained to better understand and constrain the composition of the back-arc basin oceanic crust and uppermost mantle, as well as the architecture of oceanic core complexes. The extinct back-arc basin environment at the Godzilla Megamullion provides a further unique opportunity to explore life in an oceanic crust after extinction of its hydrothermal activity.

 Large Igneous Provinces (LIPs), such as the Ontong Java Plateau (OJP) in the western equatorial Pacific, provide information on mantle processes and composition, and their formation may have global environmental consequences. The OJP is the largest oceanic plateau and is probably the most voluminous igneous edifice on Earth. Despite its importance, the size, volume, and formation rate of the OJP are not yet well constrained. The maximum extent of OJP-related volcanism may be even greater than currently estimated, because volcanological studies indicate that long lava flows (or sills) from the OJP may have reached the adjacent Nauru, East Mariana, and possibly Pigafetta basins. Moreover, the similarity in age and some geochemistry of lavas from the Ontong Java, Hikurangi, and Manihiki plateaus suggests that they once may have been part of a single LIP (Ontong Java Nui, OJN). If true, the massive volcanism may have covered > 1% of the Earth's surface. The lack of detailed knowledge of the size, age, and composition of the OJP has given rise to various models, such as a surfacing mantle plume head, bolide impact, and fusible mantle melting, but no model satisfies all observational data and no consensus has been reached on its origin. The OJP is divided into the High Plateau to the west and the Eastern Salient to the east. The basaltic basement of the OJP was cored at seven sites during Deep Sea Drilling Project (DSDP Site 289) and Ocean Drilling Program (ODP Sites 289, 803, 807, 1183, 1185, 1186, and 1187) expeditions, but all sites are exclusively located on the High Plateau. In order to examine the true extent of the OJP (i.e., whether the flows in the Nauru, East Mariana, and Pigafetta basins, as well as the Manihiki and Hikurangi plateaus are parts of the OJN), we propose drilling in the Eastern Salient and adjacent basins to recover basement samples. We also propose drilling through the sedimentary section on the Magellan Rise, a small plateau that formed > 20 Myr before the proposed OJN emplacement. Because of its greater age, the sedimentary sequence on the Magellan Rise may preserve ash layers or other chemical tracers that cover the entire eruptive history of OJN. The sediment layers from the Magellan Rise are also useful for evaluating environmental effects of OJN emplacement, including older and younger perturbations related to other LIPs.

 Water circulation, along with plate subduction, is considered based on the stabilities of hydrous phases and pressure–temperature profiles of the sinking oceanic plate. Water in a rather hot slab like the present one may be largely liberated at shallow depths (< 150 km) and return to the ocean via. arc magmatism. On the other hand, stabilization of dense hydrous minerals under cooler conditions, which current subduction zones will soon experience, causes the transportation or reflux of seawater to the deep mantle, which reduces the total mass of surface seawater. Simple calculations accepting water contents in the subducting slab suggested by a recent seismic velocity structure model indicate that the Earth's oceans are likely to disappear ∼80 million years hence. Significant changes may happen such as the end of plate tectonics and the onset of snowball Earth, with associated catastrophes affecting life. The only way to confirm this picture of the future of the ocean planet Earth is to examine deep hydration taking place along the outer rise through direct analyses of the upper mantle across the Moho.

 The style of crustal extension is governed by M, which is the ratio of magma consumed to relax the crustal strain caused by plate spreading. Bathymetric profiles across ridges are reproduced well by changing M. However, what determines the value of M has not been explained. Fast-spread oceanic crust comprises dense sheet flows underlain by thin dense sheeted dikes compared to magmas. This density structure increases the magma extruded, allowing the crust to be extended solely by magmatic accretion; whereas, the intermediate-spread crust consists of less dense pillow lavas, yielding an apparent level of neutral buoyancy that traps magma to develop the sheeted dikes below. Consequently, the crust extends through dike intrusions in the lower levels and faults at shallow levels. Thus, the density structure of the oceanic crust determines the style of plate spreading, or the value of M. Because the spreading rate or the strain rate does not vary within the same ridge segment, intrasegment variations in crustal structure depend directly on the supply rate of magma, or the value of M, which decreases with the thinning of extrusive layers and the thickening of sheeted dikes along the Galapagos Spreading Center and the East Pacific Rise. This tendency is supported by the crustal architecture observed in holes 504B and 1256D, the Hess Deep and the Oman Ophiolite. The density structure of the upper crust can be discerned by the proportion of sheet flows among extrusive rocks on the ridge axis, which is drastically reduced with spreading rates from 10 cm/a to 7 cm/a. This spreading rate interval coincides with the change in axial magma chamber depth. Throughout this rate interval, the key observation is whether the style of crustal extension from magmatic accretion-dominant to fault displacement-dominant changes gradually, or changes abruptly across a certain threshold that divides these two. To understand how the crustal extension varies through the spreading rate interval, it is proposed to drill into the 78-81 Ma old crust spread at 7 cm/a at one of the candidate sites of MoHole on the North Arch off Hawaii. Crust of this age and spreading rate interval has never been drilled in the history of ocean drilling. This drilling will provide a reference section for the upper oceanic crust and a pilot hole to inform the design of future mantle drilling.