Search

Understanding the processes that lead to the lithosphere thinning is a key aspect of continental geology research. In this paper, we present essential observations and summarize our understandings on the lithosphere thinning and accompanying magmatism in eastern continental China since the Mesozoic as a straightforward consequence of plate tectonics. We show that the lithosphere thinning in the Mesozoic resulted from basal hydration weakening with the water coming from dehydration of the paleo-Pacific plate in the mantle transition zone. The weakening effect is to convert the basal lithosphere into asthenosphere by reducing its viscosity, having thus thinned the lithosphere while triggering mantle melting and crustal magmatism marked by the widespread Mesozoic basalts and granitoids in space and time. These observations and logical reasoning require the existence and effect of subducted paleo-Pacific plate in the mantle transition zone, whose active subduction ended at ~ 90 Ma with the suture located off the continental China marked by the arc-shaped southeast coastline. As a result, the thinned lithosphere began a 40-Myr period (i.e., ~ 90 to ~ 50 Ma) of basal accretion manifested by compositional systematics of basalts erupted in this period. The initiation of the present-day western Pacific subduction at ~ 50 Ma and its eastward retreat caused eastward drift of continental China, leaving the older portions of the present-day Pacific slab stagnant in the mantle transition zone with resumed water supply in the form of hydrous melt to maintain the thinned lithosphere, which is the same as creating and maintaining the oceanic-type seismic low velocity zone (LVZ) beneath eastern China, responsible for the Cenozoic alkali basalt volcanism in the region. That is, the present-day lithosphere-asthenosphere boundary (LAB) beneath eastern China is a petrological boundary, either as an amphibole dehydration solidus or water-saturated solidus. As predicted, the Cenozoic alkali basalts in eastern China demonstrate that lithosphere thickness (i.e., the LAB depth) controls the compositions of mantle melts, i.e., the lid effect. The latter further confirms the LAB beneath eastern China as a solidus, below which decompression melting happens, and above which melt solidifies or ascends rapidly to the surface. Our studies thus lead us to the unavoidable conclusion that the lithosphere thinning in the Mesozoic, the present-day LAB, the seismic LVZ and the widespread Mesozoic-Cenozoic magmatism in eastern China are all consequences of plate tectonics in response to paleo-Pacific plate subduction, which is of global significance for understanding intra-continental magmatism at present and in Earth’s histories.

NW Iberia includes a rather complete section of a Variscan suture, where different terranes with continental or oceanic affinities appear with clear structural relationships. Three groups of terranes, namely Upper, Ophiolitic and Basal units and a frontal tectonic mélange appear in Galicia, in Cabo Ortegal, Órdenes and Malpica-Tui complexes. They constitute a huge allochthonous pile thrust over the Iberian parautochthonous and autochthonous domains, which represent the section of the Gondwanan margin that escaped continental subduction during the Variscan cycle. Considering the allochthonous character of the nappe pile and the strong deformation associated to the Variscan collision, there are problems to identify the original tectonic setting of the terranes and thence, it is difficult to reconstruct the paleogeographic context during the Variscan and pre-Variscan times in detail. Key features to perform any model for the Variscan convergence should consider the existence of two different high-P metamorphic events (dated at c. 400 and 370 Ma, respectively), separated in time for the generation of mafic-ultramafic sequences at c. 395 Ma which constitute the most common ophiolites described in the Variscan suture. Some dynamic models developed in NW Iberia have important problems to explain the observed tectonothermal evolution, especially the older high-P metamorphic event and the exhumation of deeply subducted transitional-type sections. On the other hand, the recently discovered participation of an older continental crust in the generation of different protoliths of the Middle Devonian ophiolites, makes difficult the interpretation of these ophiolites in relation to open wide oceanic domains. This paper describes the distribution, structure, lithologies, geochronology and chemical composition of the terranes involved in the Variscan suture of NW Spain. The scope of this description ranges from detailed regional aspects to the discussion of the development of the Variscan Orogen in the context of the assembly of Pangea. A two-stage collisional model affecting a wide Gondwanan platform may explain most of the evidences in NW Iberia. The generation of a long pull-apart basin probably occurred after the first collision, where the Devonian ophiolites were formed. These ophiolites have been in general described in the context of the Rheic Ocean, but according to the new existing data they do not seem related to this ocean, but rather they were formed after its closure., Financial support has been provided by the Spanish project CGL2012-34618 (Ministerio de Economía y Competitividad).

Arclogites are eclogite-like rocks formed magmatically as ultramafic residues and cumulates in the roots of thick arcs (Ducea et al., 2021 companion paper). They are inferred to be volumetrically important assemblages to complement subduction-related magmatic rocks at depth, in areas where the upper plate crust is thick. At the surface and in the shallow crust, these arcs form stratovolcanoes found at many sites around the Pacific and batholiths, which are exposed in extinct arcs, such as the western North American Cordillera. Arclogites complement these shallower manifestations of magmatism in subduction zones in that they represent the ultramafic residues left behind following the extraction of intermediate melts. Mass balance calculations constrained by the low silica contents of garnet, amphibole, and iron‑titanium oxides dictate that the silica content of a given arc tracks with the volume of arclogitic residues beneath it. We show that melt extraction takes place in punctuated events of hot-zone evacuation of the lower crust which effectively make up the magmatic flare-ups documented in the mid- to upper-crust. These cyclic events lead to the densification of arclogitic residues and trigger their foundering. We show that in addition to garnet, Fe Ti oxides play an important role in the densification of roots and can trigger foundering even in garnet-free arcs. Recent models show that foundering may be small scale and lateral shearing may have an important role in removing sub arc residues. Consequently, previously postulated large magnitude uplift and large-scale mantle-derived magmatism may not accompany foundering episodes. Partial melting of descending arclogitic bodies is expected to produce small amounts of nepheline (and/or leucite) normative magmas similar in composition to alkaline massifs found in continental interiors, not basaltic or intermediate melts. The foundering rates of arclogites are calculated to be around 20–40 km3/km/Myr in Phanerozoic arcs. The fate of foundered arclogite may include stalling at the 660 km discontinuity, accumulation and mixing with subducted oceanic material at “slab graveyards” along the core-mantle boundary, and/or disaggregation via ductile flow in the mantle. Regardless, there has to be a reservoir in the mantle representing these lower crustal recycled bodies over time. We show that the most likely such reservoir is EM1, one of the well-known and commonly identified recycled material in ocean island basalts, and that 1–3% of the volume of the mantle may be made of this reservoir.

This paper reviews recent progress in tsunami geology and the history of coastal uplift/subsidence along the Pacific coasts of Hokkaido and Tohoku, facing the Kuril and Japan Trenches, respectively. In the last 20 years, research by tsunami geologists in northeast Japan has focused on extending the record of tsunamis beyond the range of historical documents in the region, which span too little time to reveal much of the earthquake history over the last millennia. Along the coast of Hokkaido, recurrent sandy deposits are evidence of historical and prehistoric large tsunamis. At Kiritappu marsh, for example, at least 15 identified tsunami deposits occurred between 200 and 6,000 years ago. The spatial distribution of the most recent prehistoric tsunami deposit, that of the so-called 17th-century earthquake and tsunami, exceeds historical and recent tsunami inundations in eastern Hokkaido. Numerical simulations of this event require a multi-segment fault model with variable slip (Mw > 8.8) encompassing both the Tokachi-oki and Nemuro-oki regions to reproduce the observed deposit distribution in eastern Hokkaido. The recurrence interval of such multi-segment earthquakes along the southern Kuril Trench is estimated to be 100–800 years based on tephrochronological and radiocarbon evidence. Tsunami deposits are also associated with historical earthquakes in 1843 and 1894 in eastern Hokkaido and with the 1640 Komagatake eruption in western Hokkaido. The 17th-century earthquake probably triggered postseismic coastal deformation. Diatom-based transfer functions suggest rapid subsidence before the event and gradual uplift afterward, and historical volcanic ash layers indicate that gradual postseismic uplift occurred over decades. A simple dislocation model can explain such gradual uplift as an aseismic fault extension after the 17th-century event. In the Tohoku region, a tsunami deposit associated with the 869 Jogan earthquake has been identified, and the inundation area of this event has been constrained based on the lateral extent of the tsunami deposit and its stratigraphic position beneath the historical To-a tephra (915 CE). Land-level changes at Odaka, Fukushima prefecture, associated with the Jogan event have been estimated based on changes of fossil diatom assemblages below and above the tsunami deposit. Numerical simulations based on geological evidence and observations of the 2011 Tohoku tsunami deposit suggest that the Jogan event was a plate-boundary rupture at least 200 km long along the Japan Trench (Mw > 8.3–8.6). The 1454 Kyotoku tsunami is also considered to have been generated by a rupture area including the Miyagi-oki region (part of the Jogan rupture). It remains unclear whether the 869, 1454, and 2011 tsunamis were similar in size and source; if they were, the recurrence intervals of such events near Sendai are less than 600 years and more regular than those for multi-segment earthquakes along the Kuril Trench.