Intracontinental mantle plume and its implications for the Cretaceous tectonic history of East Asia

Earth and Planetary Science Letters 479 (2017) 206–218

Contents lists available at ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Intracontinental mantle plume and its implications for the Cretaceous tectonic history of East Asia In-Chang Ryu a , Changyeol Lee b,∗ a b

Department of Geology, Kyungpook National University, Daegu, 41566, Republic of Korea Faculty of Earth Systems and Environmental Sciences, Chonnam National University, Gwangju, 61186, Republic of Korea

a r t i c l e

i n f o

Article history: Received 10 November 2016 Received in revised form 10 September 2017 Accepted 13 September 2017 Available online xxxx Editor: F. Moynier Keywords: intracontinental mantle plume plume-slab interaction numerical model Cretaceous East Asia adakite

a b s t r a c t A-type granitoids, high-Mg basalts (e.g., picrites), adakitic rocks, basin-and-range-type fault basins, thinning of the North China Craton (NCC), and southwest-to-northeast migration of the adakites and I-type granitoids in southern Korea and southwestern Japan during the Cretaceous are attributed to the passive upwelling of deep asthenospheric mantle or ridge subduction. However, the genesis of these features remains controversial. Furthermore, the lack of ridge subduction during the Cretaceous in recently suggested plate reconstruction models poses a problem because the Cretaceous adakites in southern Korea and southwestern Japan could not have been generated by the subduction of the old Izanagi oceanic plate. Here, we speculate that plume-continent (intracontinental plumeChina continent) and subsequent plume-slab (intracontinental plume-subducted Izanagi oceanic plate) interactions generated the various intracontinental magmatic and tectonic activities in eastern China, Korea, and southwestern Japan. We support our proposal using three-dimensional numerical models: 1) An intracontinental mantle plume is dragged into the mantle wedge by corner flow of the mantle wedge, and 2) the resultant channel-like flow of the mantle plume in the mantle wedge apparently migrated from southwest to northeast because of the northeast-to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. Our model calculations show that adakites and I-type granitoids can be generated by increased slab-surface temperatures because of the channel-like flow of the mantle plume in the mantle wedge. We also show that the southwest-to-northeast migration of the adakites and I-type granitoids in southern Korea and southwestern Japan can be attributable to the opposite migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. This correlation implies that an intracontinental mantle plume existed in eastern China during the Cretaceous and that the mantle plume was entrained into the mantle wedge as a channel-like flow. An intracontinental mantle plume can explain the adakitic rocks, A-type granitoids, high-Mg basalts, and basin-and-range-type fault basins distributed in eastern China. Thus, the mantle plume and its interaction with the overlying continent and subducting slab through time plausibly explain the Cretaceous tectonic history of East Asia. © 2017 Elsevier B.V. All rights reserved.

1. Introduction The tectonic history of eastern China during the Cretaceous features complex magmatism, including A-type granitoids (Kim et al., 2016; Wang et al., 2006b; Wu et al., 2002, 2005), high-Mg basalts such as picrites (Gao et al., 2008), and adakitic rocks (Castillo, 2012; Wang et al., 2006b; Xu et al., 2002), as well as a unique tectonic history that includes lithospheric thinning of the North China Craton (NCC) (Menzies et al., 2007; Wang et al., 2006b;

*

Corresponding author. E-mail addresses: [email protected] (I.-C. Ryu), [email protected] (C. Lee). http://dx.doi.org/10.1016/j.epsl.2017.09.032 0012-821X/© 2017 Elsevier B.V. All rights reserved.

Wu et al., 2005; Xu et al., 2002) and NE-trending basin-and-rangetype fault basins, such as the Bohaiwan and Songliao Basins (Okada, 1999; Ren et al., 2002) (Fig. 1a). In addition to the intracontinental magmatism and tectonic history, adakites and A- and I-type granitoids are present in central and southern Korea (Kim et al., 2016; Wee et al., 2006) and southwestern Japan (Iida et al., 2015; Imaoka et al., 2014; Kinoshita, 1995, 2002; Kutsukake, 2002), and the peak magmatism of the adakites and I-type granitoids migrated from southwest to northeast at a rate of ∼3 cm/y (Kinoshita, 2002). To explain the characteristic tectonic history of East Asia during the Cretaceous, passive upwelling of deep asthenospheric mantle or ridge subduction have been suggested. The passive upwelling of deep asthenosphere resulted in various types

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

207

Fig. 1. Distributions of Cretaceous magmatism and basins in East Asia on a present-day map, the trench-normal convergence rate and age of the Izanagi oceanic plate in ancient southwestern Japan through time, and snapshots of the plate reconstruction models during the Cretaceous. a) A-type granitoids (cyan squares), I-type granitoids (blue squares), adakitic rocks (extrusive: red circles; intrusive: red squares), high-Mg basalts such as picrites (large green squares), adakites (red triangles), and basin-and-range-type fault basins (gray shaded areas) are compiled from the references in the text. The numbers indicate the representative ages of the igneous rocks and basin development. Major tectonic lines are from Ren et al. (2002) and Xu (2001). The red dashed rectangle labeled ‘abcd’ corresponds to the top surface of the numerical model domain used for this study, shown in Fig. 2a and b. b and c) Trench-normal convergence rate and age of the subducting slab in 5-Myr increments are interpolated using piecewise polynomials, respectively. Slab age linearly decreases with time because of convergence of the mid-ocean ridge separating the Izanagi and Pacific oceanic plates. The inverted red triangle indicates the peak magmatism of the adakites and I-type granitoids in ancient Kyushu, southwestern Japan. d, e and f) Snapshots at 120, 90 and 60 Ma, respectively, based on the plate reconstruction models (Gurnis et al., 2012; Sdrolias and Müller, 2006) with a conjectured core location of the intracontinental mantle plume and channel-like flow of the mantle plume generated by corner flow. The intracontinental mantle plume and channel-like flow of the mantle plume apparently migrated from southwest to northeast because of the opposite migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. The numbers on the colored isochrons represent the ages of the oceanic plate. The black dashed line in f indicates the mid-ocean ridge separating the Izanagi and Pacific oceanic plates. The thin red line indicates the accumulated relative motion of the Izanagi oceanic plate with respect to the East Asian continents. Figures from b to f are modified from Lee and Ryu (2015). (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

of magmatism in eastern China (Imaoka et al., 2014; Menzies et al., 2007; Ren et al., 2002; Wang et al., 2006b). Along with the intracontinental magmatism, the southwest-to-northeast migration of the subducting mid-ocean ridge (Isozaki et al., 2010; Maruyama et al., 1997) separating the Izanagi and Pacific oceanic plates resulted in extensive slab melting that generated the adakites and I-type granitoids in southern Korea (Wee et al., 2006) and southwestern

Japan (Iida et al., 2015; Imaoka et al., 2014; Kinoshita, 1995, 2002; Kutsukake, 2002). Although the abovementioned hypotheses are promising, several issues remain unresolved. First, the picrites in eastern China are attributed to the delamination of the Archean mantle lithosphere of the eastern part of the North China Craton but the passive (adiabatic) upwelling of deep asthenospheric mantle could not

208

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

have generated delamination (Gao et al., 2008). Instead, the delamination of the craton may have resulted from thermal erosion (Burov et al., 2007; Deng et al., 2004; Jahn et al., 1999; Sobolev et al., 2011) or refertilization via diking (Foley, 2008; Lee et al., 2011) by an underlying mantle plume. Second, the intracontinental adakitic rocks may not be correlated with ridge subduction because the locations of these adakitic rocks are too far (>1,000 km) from the ancient trench (Okada, 1999). In addition, the geochemistry of the intracontinental adakitic rocks (Castillo, 2012; Wang et al., 2006b; Xu et al., 2002) shows higher 87 Sr/86 Sr and lower 143 Nd/144 Nd values than those of typical adakites resulting from partial melting of subducted oceanic crust in southern Korea and southwestern Japan; the intracontinental adakitic rocks were likely generated by partial melting of the thickened or delaminated lower continental crust, which is not relevant to ridge subduction. Third, recent and considerably improved plate reconstruction models (Gurnis et al., 2012; Sdrolias and Müller, 2006) show a lack of ridge subduction in East Asia throughout the entire Cretaceous (Fig. 1d–f). The old and rapidly subducting Izanagi oceanic plate (Fig. 1b and c) could not have yielded slab melting and does not explain the southwest-to-northeast migration of the adakites and I-type granitoids in southern Korea (Wee et al., 2006) and southwestern Japan (Iida et al., 2015; Kinoshita, 1995, 2002; Kutsukake, 2002). Other processes for generating adakites, including a cold plume randomly generated via the Taylor instability (Gerya and Yuen, 2003), flat subduction (Gutscher et al., 2000), slab tear (Hu and Liu, 2016) and/or magma fractionation under the ancient arc (Richards and Kerrich, 2007) cannot be correlated with the systematic southwest-to-northeast migration of the adakites and I-type granitoids in southern Korea and southwestern Japan unless further evidence of the same directional migration of the processes is revealed. Thus, a new and alternative explanation should be consistent with the intracontinental and arc magmatism, tectonic history, and plate reconstruction models. Both the intracontinental (partial melting of the delaminated lower continental crust in eastern China, e.g., picrites) and arc magmatism (partial melting of the subducted old oceanic crust in southern Korea and southwestern Japan, e.g., adakites) likely require anomalous heating. The exclusion of the passive upwelling of deep asthenospheric mantle and ridge subduction prompts the question of what mechanism could produce both intracontinental and arc magmatism. As an explanation, an intracontinental mantle plume in eastern China has been suggested to explain both the intracontinental and arc magmatism (Lee and Ryu, 2015). In this scenario, interactions between the intracontinental mantle plume and overlying continental blocks (eastern part of the North China Craton) were responsible for the intracontinental magmatism in eastern China, and the mantle plume that was dragged into the mantle wedge by corner flow, or channel-like flow of the mantle plume, was responsible for the arc magmatism in southern Korea and southwestern Japan. In addition, they suggested that the East Asian continental blocks overlying the mantle plume migrated from northeast to southwest at a rate of ∼3 cm/y with respect to the Izanagi oceanic plate during the Cretaceous, resulting in the apparent southwest-to-northeast migration of the intracontinental mantle plume and channel-like flow of the mantle plume in the mantle wedge (Fig. 1d–f). The apparent southwest-to-northeast migration of the channel-like flow in the mantle wedge can be correlated with the southwest-to-northeast migration of the adakites and I-type granitoids along southwestern Japan from ∼120 to 80 Ma (Kinoshita, 2002). Although the previous study correlated the intracontinental mantle plume and channel-like flow of the mantle plume in the mantle wedge with intracontinental and arc magmatism, it did not quantitatively evaluate whether the corner flow of the mantle wedge caused the intracontinental mantle plume to be dragged

into the mantle wedge and the channel-like flow of the mantle plume in the mantle wedge resulted in partial melting of the subducted oceanic crust, which is required for the temporal occurrences of the adakites and I-type granitoids in the ancient arc, e.g., southern Korea (Wee et al., 2006) and southwestern Japan (Kinoshita, 1995, 2002). In this study, three-dimensional numerical models were formulated to evaluate the behavior of the intracontinental mantle plume that was dragged into the mantle wedge by corner flow with the migration of the overlying continental blocks. From the model calculations, we show that the channel-like flow of the mantle plume in the mantle wedge simultaneously migrated from southwest to northeast and resulted in the temporal occurrences of the adakites and I-type granitoids in the ancient arc. Then, we discuss how the intracontinental mantle plume during the Cretaceous was correlated with the various types of intracontinental magmatic and tectonic activities in eastern China. 2. Numerical model Simulating the intracontinental mantle plume and the northeastto-southwest migration of the East Asian continental blocks in the three-dimensional model shows that the intracontinental mantle plume was dragged into the mantle wedge as a channel-like flow, although the channel-like flow was slightly deflected toward southwest (∼110 km) because of viscous drag of the migrating continental blocks (see the appendix). Here, we focus on the interaction between the channel-like flow of the mantle plume in the mantle wedge and the subducting slab, which is correlated with the temporal and spatial occurrences of adakites and I-type granitoids from southwest to northeast along southern Korea and southwestern Japan. In the three-dimensional numerical kinematic-dynamic subduction experiments (Fig. 2), the behavior of the mantle wedge is dominated by the kinematically subducted slab, and thus, mantle buoyancy is neglected in the governing equations for the incompressible Boussinesq approximation (Kneller and van Keken, 2008; Lee and King, 2010; Lee and Lim, 2014; Wada and Wang, 2009):

 0=∇ ·ν

continuity equation,

momentum equation, 0 = −∇ P + ∇ · τ¯ DT ρc C p = ∇ · (k∇ T ) energy equation, Dt

(1) (2) (3)

 is the velocity (m/s), P is the pressure (Pa), τ¯ is the dewhere ν viatoric stress tensor (Pa), ρc is the density (kg m−3 ), C p is the specific heat capacity at constant pressure (J kg−1 K−1 ), T is the temperature (K), t is the time (s), and k is the thermal conductivity (W m−1 K−1 ). As in the appendix, the radiogenic heat production in the mantle and lithosphere is neglected. The model domain is constructed as a rectangular box of 400 × 600 × 200 km (x-× z-× y extent). For simplicity, the threedimensional curving geometry of the subducting slab is prescribed using a part of a cylinder with a radius and height (z-direction) of 425 and 600 km, respectively (Fig. 2a). Because the slab dip is fixed over time, temporal variation in the slab dip by dynamic slab buckling or trench migration (Lee and King, 2011) is not considered. The thermal contribution of the sub-slab mantle to the mantle wedge is minor because the cold subducting slab effectively hinders heat transfer from the sub-slab mantle to the mantle wedge. Thus, we extend the subducting slab to the bottom of the model domain. A 35-km-thick rigid layer is used for the overlying crust. The corner of the mantle wedge may be substantially serpentinized by water from the dehydrated subducting slab (Blakely et al., 2005; Bostock et al., 2002). Thus, the less dense and viscous serpentinized mantle wedge (Hilairet et al., 2007) may be decoupled from the subducting slab and may not be involved in the corner flow of

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

209

Fig. 2. Model geometry of the numerical subduction model with implementation of the apparent southwest-to-northeast migration of the channel-like flow of the mantle plume. a) The top surface of the model domain (abcd) represents the modeled region in Fig. 1a. The thick black arrow indicates the direction of subduction. The red dashed lines are the boundaries of the subducting slab, mantle wedge and overlying rigid crust. The temperature anomaly on the back-arc wall boundary results in the channel-like flow of the mantle plume in the mantle wedge, and migrates from southwest to northeast at a rate of 3 cm/y. b) Locations of the temperature anomaly on the back-arc wall boundary (dcef plane) at 115, 110, 105, 100, and 95 Ma. The details of the temperature anomaly in the dashed red plane and circle are shown in c. c) Net temperature anomaly on the back-arc wall boundary at 115, 110, 105, 100, and 95 Ma. The half-thickness and half-width are defined as the distances from the center of the temperature anomaly when the temperature is reduced to 36.79% of the peak temperature of the temperature anomaly (96.76 ◦ C for the peak temperature of 263 ◦ C) (Lee and Lim, 2016). The red dashed circle indicates the temperature contour of 96.76 ◦ C when both the half-thickness and half-width are 20 km. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

210

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

the mantle wedge, consistent with the low heat flow observations in the fore-arc (Wada and Wang, 2009). To consider the mantle wedge decoupled by serpentinization in our models, we extend the overlying rigid crust to the subducting slab surface at a depth of 70 km (Currie et al., 2004; Wada and Wang, 2009). A no-slip boundary condition is applied to the bottom and vertical wall of the rigid crust (green planes in Fig. 2a and b). A stress-free boundary condition allows for dynamic mantle inflow/outflow into/from the mantle wedge through the back-arc and bottom wall boundaries of the mantle wedge (blue planes in Fig. 2a and b). No mantle inflow/outflow is allowed through the southwestern and northeastern sidewalls of the model domain, consistent with the use of the trench-normal component of the convergence vector of the subducting slab. Previous studies show that the time-evolving convergence rate and age of the subducting slab significantly affect the thermal structures of the subducting slab (Kim and Lee, 2014; Lee and Lim, 2014). Thus, we consider the time-evolving subduction parameters estimated from the plate reconstruction models (Sdrolias and Müller, 2006) depicted by the GPlates software (Gurnis et al., 2012). The behavior of the mantle wedge is mostly governed by the trench-normal component of the convergence velocity (Bengtson and van Keken, 2012; Honda and Yoshida, 2005); thus, we considered only the trench-normal convergence of the subducting slab. The convergence rate and age of the subducting slab for 5-Myr increments are interpolated using piecewise polynomials (Fig. 1b and c). The thermal structure of the subducting slab is approximated by the half-space cooling model (Turcotte and Schubert, 2002) with a mantle potential temperature of 1350 ◦ C, and the thermal structure is prescribed as the temperature boundary condition on the trench-side wall boundary (the opposite wall of the back-arc wall boundary). The initial temperature distribution of the model domain is calculated using the half-space cooling model relevant to a 50-Myr-old oceanic plate. Owing to the incompressible model approximation, the mantle adiabat of 0.35 ◦ C/km is not considered in the governing equations but is added to the calculated temperatures. Temperatures on the back-arc wall boundary use the initial temperature boundary condition down to a depth of 160 km (fixed T zone in Fig. 2a and b), where the mantle inflow/outflow transition occurs. No temperature boundary condition is prescribed when depth is deeper than 160 km (‘not fixed T’ zone in Fig. 2a and b). The sidewalls of the model domain are insulated. The top surface temperature is fixed at 0 ◦ C. As shown in the appendix, the intracontinental mantle plume was dragged into the mantle wedge as a channel-like flow, although the channel-like flow was slightly deflected toward the southwest (∼110 km) because of viscous drag of the overlying continental blocks. To approximate the channel-like flow in the mantle wedge, we prescribe a temperature anomaly on the backarc wall boundary (Lee and Lim, 2016). The temperature anomaly that results in a channel-like flow in the mantle wedge is expressed as a modified normal distribution function that is a function of depth from the top surface (y) and distance from the southwestern sidewall (z) (Fig. 2b and c),

mantle plume; and y p and z p are the depth and distance of the peak temperature of the channel-like flow of the mantle plume from the top surface (km) and southwestern sidewall (km), respectively. Standard deviations of the modified normal distribution function, expressed as σt and σ w , are appointed as the halfthickness (km) and half-width (km) of the temperature anomaly, respectively (Fig. 2c). Previous evaluation of the plate motion of the East Asian continental blocks and Izanagi oceanic plate from 120 to 80 Ma shows that the East Asian continental blocks migrated from northeast to southwest at a rate of ∼3 cm/y with respect to the Izanagi oceanic plate (Lee and Ryu, 2015). Because mantle plumes can be stationary for millions of years (e.g., Hawaiian mantle plume), the intracontinental mantle plume and channel-like flow of the mantle plume apparently migrated from southwest to northeast with respect to the overlying continental blocks, although the trenchward end of the channel-like flow was deflected by ∼110 km (Fig. A2). Thus, we model the southwest-to-northeast migration of the channel-like flow of the mantle plume by fixing the overlying continental blocks, changing the reference frame from the mantle plume to the overlying continental blocks. The southwestto-northeast migration of the channel-like flow of the mantle plume should match the peak magmatism of the adakites and Itype granitoids in the ancient Kyushu arc, southwestern Japan at ∼105 Ma (Fig. 1a). This can be modeled by migrating the temperature anomaly (z p ) from southwest to northeast, defined as,

T ( y , z)

where A is the prefactor, d g is the grain size, m is the grain size exponent, E is the activation energy, V is the activation volume, R is the gas constant, T is the mantle temperature, and T adia is the mantle adiabat. All of the relevant parameters are enumerated in Table 1. To consider the hydrated mantle wedge (i.e., wet olivine) that results when water escapes from the dehydrated subducting slab, the calculated dry olivine rheology is multiplied by a viscosity reduction factor of 1/20 and extended to a depth of 100 km (Honda and Saito, 2003). To consider the gradual termination of slab dehydration with depth (Hyndman and Peacock, 2003;



   ( y − y p )2 ( z − z p )2 − = T mantle ( y , z) + T plume exp − exp (σt )2 (σ w )2 temperature boundary condition on the back-arc wall boundary

(4)

where T mantle is the mantle temperature calculated from the halfspace cooling model relevant to a 50-Myr-old oceanic plate; T plume is the peak temperature (263 ◦ C) of the channel-like flow of the

z p = v channel (t peak,0 km − t ), location of the temperature anomaly

(5)

where v channel is the apparent southwest-to-northeast migration rate of the channel-like flow (3 cm/y), and t peak,0 km is the peak time (115 Ma) of the arc magmatism of the adakites and I-type granitoids located at 0 km from the southwestern wall boundary. With time, the temperature anomaly at the back-arc wall boundary migrates from southwest to northeast at a rate of 3 cm/y (Fig. 2b and c). For example, z p is 300 km when t is 105 Ma; thus, the peak temperature of the channel-like flow of the mantle plume on the back-arc wall boundary exists at 300 km from the southwestern sidewall boundary and coincides with the peak magmatism in the ancient Kyushu arc at ∼105 Ma. The term y p is fixed at 120 km; no depth change of the temperature anomaly occurs. For the size of the temperature anomaly on the back-arc wall boundary, we vary the half-thicknesses (σt ) using 20, 30 and 40 km with a fixed half-width (σ w ) of 40 km, values that are similar to those used for plume implementation in previous studies (Moore et al., 1998; Ribe and Christensen, 1994). For mantle rheology, we use the diffusion creep of dry olivine (Karato and Wu, 1993) because diffusion creep rheology can approximate mantle flow and thermal structures in kinematicdynamic subduction experiments if the depth is shallow, as used in previous studies (Honda and Saito, 2003; Lee and King, 2009):

η = A −1 dmg exp



E + PV



R ( T + T adia )

diffusion creep for mantle rheology,

(6)

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

Table 1 Model and rheological parameters.

ρc (reference density, kg m3 )

C p (specific heat capacity, J kg−1 K−1 ) k (thermal conductivity, W m−1 K−1 ) A (prefactor, m2.5 Pa−1 s−1 ) d g (grain size, m) m (grain size exponent, ·) E (activation energy, J mol−1 ) V (activation volume, m3 mol−1 ) R (gas constant, J mol−1 K−1 )

3300 1200 3.96 6.1 × 10−19 1.0 × 10−3 2.5 300 × 103 6.0 × 10−6 8.314

Schmidt and Poli, 1998), we linearly increase the viscosity reduction factor from 1/20 (wet olivine) to 1 (dry olivine) for the mantle above the surface of the subducting slab from 100-km to 150-km depths (Kim and Lee, 2014; Lee and Lim, 2014). The governing equations are non-dimensionalized using the reference values indicated in Table 1, following the approach of King et al. (2010). We use the finite element package COMSOL Multiphysics© (www.comsol.com) to numerically solve the governing equations. The whole model domains consist of 445,860 elements including prism, hexahedral, triangular and quadrilateral elements. To improve numerical accuracy, the elements along the boundary of the mantle wedge and subducting slab are refined (Fig. 2a and b). The time-dependent direct and segregated solver used in the appendix calculates the Stokes and energy equations from 170 to 60 Ma. An initial 30-Myr model run was conducted to minimize the effects of the initial conditions on the thermal and flow structures of the model. Because the age of the subducting slab evolves over time, the temperature boundary condition at the trench-side vertical wall boundary evolves at each time step. The convergence rate of the subducting slab is imposed on the subducting slab domain and is updated at each time step to generate the time-evolving convergence of the incoming Izanagi oceanic plate at the trench. 3. Results First, we evaluate our model calculations without the channellike flow of the mantle plume (Fig. 3). The three-dimensional model calculations show negligible variations in thermal and flow structures along the trench direction (Fig. 3a and b). Viscous coupling between the subducting slab and the overlying mantle wedge results in all the velocity vectors of the mantle being normal to the trench, as shown in the dipping map view of the mantle wedge (Fig. 3b). No significant temperature increases in the mantle wedge are observed (Fig. 3b, c and d). To evaluate the occurrences of adakites, we assume that partial melting of the subducted oceanic crust occurs at a depth of 100 km, consistent with the average depth of the slab surface under arc volcanoes (Syracuse and Abers, 2006) and the deepest depth of partial melting of the subducted oceanic crust (Mibe et al., 2011). For the solidus of the subducted oceanic crust, we use a value of 730 ◦ C at a depth of 100 km, constrained from laboratory experiments on hydrated basalt (Schmidt and Poli, 1998). If the slab surface temperature is higher than 730 ◦ C, we assume that partial melting of the subducted oceanic crust occurs; partial melting is determined by post-processing and does not affect the model calculations. The slab surface temperatures under the ancient Kyushu arc at 300 km from the southwestern sidewall do not yield any partial melting at 105 Ma, which is inconsistent with the occurrences of adakites (Fig. 3e). Next, we consider the channel-like flow of the mantle plume using three different half-thicknesses of the temperature anomaly on the back-arc wall boundary: 20, 30 and 40 km. Fig. 4 shows a snapshot of the model calculations at 105 Ma from the experiment using a half-thickness of 40 km. Owing to the southwest-tonortheast migration of the temperature anomaly on the back-arc wall boundary, the channel-like flow of the mantle plume in the

211

mantle wedge simultaneously migrates from southwest to northeast with time (Fig. 5). The channel-like flow of the mantle plume increases the temperatures of the mantle wedge (Fig. 4b). Along the channel-like flow of the mantle plume, mantle flow oblique to the trench develops under the fore-arc and arc. The oblique mantle flow might be caused by the decreased mantle viscosity of the channel-like flow, resulting in a mantle plume that is vigorously dragged into the mantle wedge (Lee and Lim, 2016). At 105 Ma, the hottest channel-like flow of the mantle plume is observed at ∼270 km from the southwestern sidewall, which is displaced by ∼30 km behind the location of the temperature anomaly on the back-arc wall boundary. This observation indicates that the channel-like flow of the mantle plume requires ∼1 Ma to reach the corner of the mantle wedge from the back-arc wall boundary. The trench-normal cross-sections show that the highest temperature of the mantle wedge under the fore-arc is observed at ∼270 km from the southwestern sidewall (Fig. 4c1) in comparison with the temperature anomaly on the back-arc wall boundary at 300 km from the southwestern sidewall (Fig. 4c2). A comparison of Fig. 4c2 and 4c3 reveals that the channel-like flow of the mantle plume increases the temperatures of the mantle wedge under the fore-arc and arc by more than 150 ◦ C. Owing to the increased mantle temperatures (Fig. 4d), the slab surface shows pulse-like increases in temperature, resulting in partial melting of the subducted oceanic crust from ∼104.5 to 102.5 Ma, which is consistent with the peak magmatism of adakites and I-type granitoids in the ancient Kyushu arc (Fig. 4e). We also conducted experiments using half-thicknesses of 20 and 30 km for the temperature anomaly, but these conditions yield lower slab-surface temperatures than the solidus of the hydrated oceanic crust (basalt); thus, no adakites are generated. The half-thickness and half-width of 40 km are the minimum sizes of the temperature anomaly on the back-arc wall boundary required to produce the temporal pattern (pulse-like) of partial melting of the subducted oceanic crust. To quantitatively evaluate the southwest-to-northeast migration of the occurrences of adakites and I-type granitoids in southern Korea and southwestern Japan, we analyze the apparent southwest-to-northeast migration of the channel-like flow of the mantle plume from 115 to 95 Ma related to the northeast-tosouthwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. As noted above, we evaluate the experiment using a half-thickness of 40 km for the temperature anomaly on the back-arc wall boundary. Fig. 6a shows the temperature contours of the inflowing and outflowing channel-like flow of the mantle plume (circular contours and distorted halfelliptical contours on the back-arc wall boundary) established by corner flow. As observed in Figs. 4 and 5, the outflowing channellike flow of the mantle plume is displaced in the mantle wedge as the temperature anomaly on the back-arc wall boundary migrates from southwest to northeast at a rate of 3 cm/y, which reflects the opposite migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. This displaced behavior is also confirmed in the dipping map view: the channel-like flow of the mantle plume in the mantle wedge is displaced to the southwest by ∼30 km (Fig. 6b). The temperature contours at 95 Ma are different from the other contours because of the reflected northeastern sidewall of the model domain. The channel-like flow of the mantle plume results in pulse-like increases in temperature higher than 150 ◦ C in the mantle wedge that migrate from southwest to northeast with time (Fig. 6c). Increases in the temperatures of the mantle wedge result in increases in slab-surface temperatures; the slab surface shows that pulse-like increases in temperature migrated from southwest to northeast with time (Fig. 6d). As noted above, we assume that partial melting of the subducted oceanic crust occurs when the slab-surface temperature is higher than 730 ◦ C at a depth of 100 km. This condition indicates that the

212

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

Fig. 3. Thermal (◦ C) and flow (cm/y) structures of our numerical model without a channel-like flow of the mantle plume in the mantle wedge; the snapshot at 105 Ma is shown. a) Temperature distributions of the subducting slab, mantle wedge and overlying rigid crust. The red dashed lines represent the boundaries of the subducting slab, mantle wedge and rigid crust. The temperature contours are depicted every 150 ◦ C. The white dashed lines indicate the dipping map view, trench-normal cross-section, and trench-parallel cross-section shown in b, c and d, respectively. b) Thermal and flow structures of the mantle wedge on the dipping map view; the dipping plane corresponds to b1b2b3b4 in a, and b4 is hidden. Temperature contours are depicted every 100 ◦ C. No lateral variations in the thermal and flow structures occur. c) Trench-normal cross-section of the model domain at 300 km from the southwestern sidewall. Temperature contours are depicted every 100 ◦ C. The thermal and flow structures of the model calculations show typical corner flow. d) Trench-parallel cross-section under the arc. As noted above, no lateral variations in the thermal and flow structures occur. e) Slab-surface temperature at a depth of 100 km under the ancient Kyushu arc. The red dashed line corresponds to the solidus of hydrated oceanic basalt (Schmidt and Poli, 1998). No partial melting of the subducted oceanic crust occurs, which is inconsistent with the occurrences of adakites and I-type granitoids. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

213

Fig. 4. Same as in Fig. 3 except for the presence of a channel-like flow of the mantle plume in the mantle wedge; the snapshot at 105 Ma is shown. a) The inflowing and outflowing channel-like flows of the mantle plume are shown on the back-arc wall boundary as a dark red circle (intermediate depth) and a distorted half-ellipse (deep depth), respectively. b) Owing to the channel-like flow of the mantle plume, the temperatures in the mantle wedge are increased by up to ∼150 ◦ C. c) Trench-normal cross-sections show that the hottest channel-like flow of the mantle plume is observed at ∼270 km from the southwestern sidewall (c1), which is displaced ∼30 km from the location of the channel-like flow of the mantle plume at 300 km from the southwestern sidewall (c2). d) Trench-parallel cross-section under the ancient Kyushu arc shows that the temperatures of the channel-like flow of the mantle plume in the mantle wedge under the arc reach up to 1400 ◦ C, ∼150 ◦ C higher than the ambient mantle. e) Slab surface at a depth of 100 km under the ancient Kyushu arc has higher temperatures than the solidus of hydrated oceanic crust, resulting in adakites and I-type granitoids. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

214

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

Fig. 5. Thermal structures of the channel-like flow of the mantle plume in the mantle wedge; the snapshots at 115, 110, 105, 100 and 95 Ma are shown. Temperature contours are depicted at 1450 (dark red), 1500 (red), 1550 (yellow) and 1600 ◦ C (white) and the temperatures below 1450 ◦ C remain blank. The transparent planes in the model domain represent the boundaries of the subducting slab, mantle wedge and rigid crust shown in Fig. 2a. The channel-like flow of the mantle plume migrates from southwest to northeast because of the migration of the temperature anomaly on the back-arc wall boundary, modeled for the northeast-to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

width of the partial melting zone remains at ∼55 km and that the melting zone migrates at a rate of 3 cm/y from southwest to northeast, consistent with the migration of the adakites and I-type granitoids along the ancient arc. 4. Discussion Our model experiments show that an intracontinental mantle plume beneath eastern China was dragged into the mantle wedge as a channel-like flow with simultaneous northeast-to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. This can explain the partial melting of the subducted oceanic crust under the arc as well as the southwest-to-

northeast migration in the occurrences of the adakites and I-type granitoids in southern Korea and southwestern Japan. If an intracontinental mantle plume existed, it should be correlated with the intracontinental magmatism (Fig. 1a). The adakitic rocks and Atype granitoids (∼136–125 Ma in the Dabie-Sulu collision zone and nearby areas (Wang et al., 2006b)) and picrites (∼119 Ma in western Shandong (Gao et al., 2008)) can be correlated with the partial melting of delaminated lower continental crust. The delamination (foundering) of the eastern part of the North China Craton actively occurred during the Cretaceous (Ma et al., 2016; Menzies et al., 2007; Ouyang et al., 2015) and may have resulted from thermal erosion (Burov et al., 2007; Deng et al., 2004; Jahn et al., 1999; Sobolev et al., 2011) or refertilization via dik-

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

215

Fig. 6. Temperature distributions of the mantle wedge and slab surface from the experiment including the channel-like flow of the mantle plume in the mantle wedge; the snapshots at 115, 110, 105, 100, and 95 Ma are depicted together. a) Temperature contours are depicted every 50 ◦ C on the back-arc wall boundary (dcef plane); the temperatures lower than 1450 ◦ C remain blank. The white contour corresponds to 1550 ◦ C. b) Temperature contours of the channel-like flow of the mantle plume in the mantle wedge are depicted every 50 ◦ C on the dipping map view (b1b2b3b4 plane); the temperatures lower than 1400 ◦ C remain blank. The white contour corresponds to 1550 ◦ C. The red dashed line corresponds to the mantle temperature at a depth of 80 km on the dipping plane shown in c. c) Temperatures of the channel-like flow of the mantle plume in the mantle wedge at a depth of 80 km. Owing to the apparent southwest-to-northeast migration of the channel-like flow of the mantle plume, the increased mantle temperatures under the arc migrate from southwest to northeast. d) Slab-surface temperatures below the overlying mantle wedge are increased by the plume-slab interactions; the temperatures lower than 730 ◦ C remain blank. Temperature contours are depicted every 20 ◦ C. All of the slab-surface temperatures higher than 730 ◦ C are overlapped on the slab surface. The red dashed line corresponds to a depth of 100 km on the slab surface; the surface region (width: ∼55 km) shallower than a depth of 100 km is considered to be molten, resulting in adakites and I-type granitoids in the ancient arc. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

ing (Foley, 2008; Lee et al., 2011) by the intracontinental mantle plume. Thus, an intracontinental mantle plume beneath eastern China can explain the intracontinental magmatism there (Fig. 1a). Hence, passive upwelling of the deep asthenospheric mantle or ridge subduction, which are inconsistent with the delamination and recent plate reconstruction models, are not required. Based on the integrated analyses and model calculations, we speculate a new tectonic model of East Asia during the Cretaceous (Fig. 7b). An intracontinental mantle plume was established in the Early Cretaceous (∼140 Ma) or even earlier under the Bohaiwan Basin (∼145 to 130 Ma) and yielded adakitic rocks and A-type granitoids (∼136–125 Ma in the Dabie-Sulu collision zone and vicinity (Wang et al., 2006b)) by the partial melting of delaminated lower continental crust, and picrites (∼119 Ma in western Shandong (Gao et al., 2008)). Owing to the northeast-

to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate at a rate of ∼3 cm/y, the intracontinental mantle plume apparently migrated to the northeast with the continuous delamination of the lower continental crust, resulting in adakitic rocks (∼125 Ma in Liaoning (Wu et al., 2005)), high-Mg basalts (∼125 Ma in Liaoning (Gao et al., 2008)), A-type granitoids (∼125 Ma in Liaoning and Jilin (Wu et al., 2005, 2002) and ∼120 to 110 Ma in Korean Peninsula (Kim et al., 2016)), and the Songliao Basin (∼130 to 110 Ma (Ren et al., 2002; Zhang et al., 2011)). As shown in the appendix, the channel-like flow from the intracontinental mantle plume (red stars in Fig. 7b) reaches the subduction zone in ∼15 Ma. This is fairly consistent with the time differences between the peak magmatism in eastern China and in southwestern Japan, e.g., the intracontinental magmatism in Shandong and vicinity at ∼125 Ma and arc magmatism in

216

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

Fig. 7. Schematic diagram of the plume-continent and plume-slab interactions in China, Korea and Japan and the Cretaceous tectonic model. a) Three-dimensional schematic diagram of the plume-continent and plume-slab interactions in East Asia. The intracontinental mantle plume resulted in the adakitic rocks, picrites, A-type granitoids, and basin-and-range-type fault basins in eastern China. The channellike flow of the mantle plume that was dragged into the mantle wedge by corner flow resulted in the partial melting of the subducted oceanic crust, generating the adakites and I-type granitoids in southern Korea and southwestern Japan. The rate of the northeast-to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate was 3 cm/y. Distributions of the magmatism and basin are from Fig. 1a. b) Cretaceous tectonic model deduced from our numerical model experiments and depicted on the plate reconstruction models from the GPlate at 90 Ma. Owing to the northeast-to-southwest migration of the East Asian continental blocks with respect to the Izanagi oceanic plate (3 cm/y), the plume-continent and plume-slab interactions apparently migrated from southwest to northeast, resulting in the migration of the intracontinental and arc magmatism in the same direction. The dashed red contours correspond to the channel-like flow of the mantle plume that was dragged into the mantle wedge by corner flow. The red stars correspond to the core of the ascending mantle plume. The red and yellow numbers indicate the times for the locations of the intracontinental mantle plume in eastern China and peaks of the adakites and I-type granitoids in southwestern Japan (Ma), respectively. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

the ancient Kyushu arc at ∼110 Ma (Fig. 7b). Thus, the Cretaceous intracontinental and arc magmatism and basins were generated by the plume-continent and plume-slab interactions through time. As discussed above, our numerical model calculations are promising for discerning the temporal history of intracontinental and arc magmatism in East Asia during the Cretaceous. However, several caveats must be discussed. First, our model calculations assume an intracontinental mantle plume in the back-arc. Although intraplate mantle plumes in the back-arc are rare at present (French and Romanowicz, 2015; Hassan et al., 2015), previous studies (Heron et al., 2015; Steinberger and Torsvik, 2012; Torsvik et al., 2008) show that some large igneous provinces (LIPs) were established in the back-arc settings (e.g., Skagerrak and

Siberian Traps). Numerical modeling studies (Heron et al., 2015; Steinberger and Torsvik, 2012) show that an intraplate mantle plume can be triggered by a subducted slab, which implies that the intracontinental mantle plume in eastern China might have been a consequence of the subducted paleo-oceanic plate. If an intracontinental mantle plume was generated by the subducted slab, corner flow drove the mantle plume into the mantle wedge (see the appendix; Kincaid et al., 2013) and resulted in plume-slab interaction. Thus, it is plausible that an intracontinental mantle plume was dragged into the mantle wedge in East Asia during the Cretaceous. Second, mantle buoyancy is neglected in our model experiments which focus on the subduction zone. Although the behavior of the channel-like flow of the mantle plume in the mantle wedge may be mostly controlled by corner flow, the lower density of the hot mantle plume results in upward flow of the plume as it is dragged into the mantle wedge (Morishige and Honda, 2011). Therefore, the vigor of the channel-like flow of the mantle plume in the corner flow may be weakened, and increases in the slabsurface temperatures would be lower than those from our model calculations. Along with the upward flow, small-scale plumes (Zhu et al., 2011) or cold plumes (Gerya and Yuen, 2003) from the slab surface may disturb the channel-like flow of the mantle plume. In contrast, the recycled eclogitized oceanic crust in the mantle plume may reduce its buoyancy (Sobolev et al., 2011) and weaken its upward flow, which may yield behavior of the channel-like flow of the mantle plume similar to that observed in our model calculations. The presence of recycled oceanic crust in the intracontinental mantle plume in eastern China should be verified by geochemical/petrological studies. Third, the lack of flood basalts and topographic swells, the primary reason that an intracontinental mantle plume has not been accepted among scientists, can be explained by considering recent studies indicating that such features are not mandatory surface expressions of an intracontinental mantle plume (Ali et al., 2010; Betts et al., 2007; Burov et al., 2007). Instead, basin-and-rangetype fault basins can occur as the surface expression of an intracontinental mantle plume (Burov et al., 2007; Gö˘güs, ¸ 2015). The Bohaiwan and Songliao Basins (Okada, 1999; Ren et al., 2002; Zhang et al., 2011), which developed from ∼145 to 130 Ma and from ∼130 to 110 Ma, respectively, can be interpreted as the surface expressions of the intracontinental mantle plume (Figs. 1a and 7a). However, further geochemical evidence of the intracontinental mantle plume must be found, although crustal contamination or fractionation in the molten magma may overwhelm the geochemical characteristics of the mantle plume (Ernst and Buchan, 2003), in contrast to intraoceanic mantle plumes such as Hawaii and Samoa (Konter and Jackson, 2012; Ren et al., 2005). Fourth, numerous magmatic activities occurred in the Mongolia and the Great Xing’an Range (Ren et al., 2002; Wang et al., 2006a) during the Late Jurassic (∼160 to 140 Ma) are not considered in our study because we focus on the Cretaceous magmatism. For the genesis of the Late Jurassic magmatism, eastward propagation of the delamination of the North China Craton, activated by the collision between North China and Siberia at ∼160 Ma, and resultant passive upwelling of the asthenospheric mantle have been suggested (Wang et al., 2006a). Although the eastward propagation of delamination may explain the magmatism in North China including that in the Songliao Basin (∼130 to 110 Ma), it is not consistent with the southwest-to-northeast migration of the magmatism distributed in eastern China and southwestern Japan. A recent seismic tomographic study indicates that the Songliao Basin was created by a Cretaceous mantle plume (He and Santosh, 2016), in contrast to the hypothesis based on the eastward migration of delamination. Thus, further studies are required to decipher the

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

genesis of the extensive magmatism in Mongolia and the Great Xing’an Range. Fifth, as far as we know, adakites and I-type granitoids are found only in southern Korea and southwestern Japan, which were the Cretaceous subduction zones. If the intracontinental mantle plume existed under the Bohaiwan Basin at ∼140 Ma, the channellike flow of the mantle plume in the mantle wedge by corner flow may have reached the ancient arc at ∼125 Ma; partial melting of the subducted oceanic crust by the hot channel-like flow may have resulted in adakites and I-type granitoids in the present East China Sea (Fig. 7b). This implies that the southwest-to-northeast migration of the A-type granitoids and adakites might have initiated in the East China Sea. Southwest-to-northeast migration of the adakites and I-type granitoids was terminated in central Japan at ∼80 Ma (Fig. 1a) which implies that the channel-like flow of the mantle plume may have ceased at ∼95 Ma in northeastern China or Far East Russia. Further petrological and geochemical studies in northeastern China and Far East Russia, as well as coring expeditions of the International Ocean Discovery Program (IODP) in the East China Sea would be useful for evaluating the initiation and termination of arc magmatism and intracontinental mantle plume. Last, the motion of the Izanagi oceanic plate utilized in our study is constrained from the assumption that the Pacific hotspots had been stationary during the Cretaceous; it contrasts with the drift of the Hawaiian hotspot during the Cretaceous (Tarduno et al., 2003). In addition, there is an uncertainty for constraining the motion of the Izanagi-Pacific spreading ridge because the Izanagi oceanic plate was completely subducted in the Early Cenozoic; the motion of the Izanagi oceanic plate during the Cretaceous could be different than what is used in our study. However, no direction change of the Izanagi-Pacific spreading ridge since ∼120 Ma (Whittaker et al., 2007), high geothermal gradients of the Eocene Shimanto belt in southwest Japan (Lewis et al., 2000; Sakaguchi, 1996) and cessation of the granitic plutonism in southern Korea at ∼50 Ma are consistent with the subparallel subduction of the Izanagi-Pacific spreading ridge along the margin of East Asia in the Early Cenozoic (∼60 to 50 Ma). Although the trenchnormal subduction of the Izanagi-Pacific spreading ridge during the Cretaceous, numerously used in previous studies, cannot be completely ruled out, the plate reconstruction models used in this study would be more plausible. 5. Conclusions Three-dimensional numerical models were conducted to explain the southwest-to-northeast migration of the adakites and I-type granitoids in southern Korea and southwestern Japan during the Cretaceous by excluding ridge subduction which is inconsistent with recent plate reconstruction models. For the genesis of the adakites and I-type granitoids, an intracontinental mantle plume that had been dragged into the mantle wedge as a channel-like flow, and the channel-like flow of the mantle plume resulted in the partial melting of the subducted oceanic crust. If an intracontinental mantle plume existed in eastern China, the southwest-tonortheast migration of the adakites and I-type granitoids in southern Korean and southwestern Japan could be explained by the apparent southwest-to-northeast migration of the intracontinental mantle plume resulting from the opposite migration of the East Asian continental blocks with respect to the Izanagi oceanic plate. Our numerical model calculations imply that an intracontinental mantle plume existed in eastern China during the Cretaceous, which is well correlated with the intracontinental magmatic and tectonic activities such as A-type granitoids, picrites, adakitic rocks, basin-and-range type fault basins and the thinning of the eastern part of the North China Craton during the Cretaceous. Although the ridge subduction during the Cretaceous cannot be completely

217

ruled out, the mantle plume and its interaction with the overlying continent and subducting slab through time plausibly explain the Cretaceous tectonic history of East Asia. Acknowledgements We thank the editors, Prof. Frederic Moynier and Dr. Christophe Sotin, and an anonymous reviewer for their careful reviews and comments, which substantially improved our manuscript. This study is supported by the Korea Meteorological Administration Research and Development Program (KMI2017-9090 for I.-C. Ryu) and the National Research Foundation of Korea (2016R1D1A1B039 30906 for C. Lee). Appendix A. Supplementary material Supplementary material related to this article can be found online at https://doi.org/10.1016/j.epsl.2017.09.032. References Ali, J.R., Fitton, J.G., Herzberg, C., 2010. Emeishan large igneous province (SW China) and the mantle-plume up-doming hypothesis. J. Geol. Soc. 167, 953–959. Bengtson, A.K., van Keken, P.E., 2012. Three-dimensional thermal structure of subduction zones: effects of obliquity and curvature. Solid Earth 3, 365–373. Betts, P.G., Giles, D., Schaefer, B.F., Mark, G., 2007. 1600–1500 Ma hotspot track in eastern Australia: implications for Mesoproterozoic continental reconstructions. Terra Nova 19, 496–501. Blakely, R.J., Brocher, T.M., Wells, R.E., 2005. Subduction-zone magnetic anomalies and implications for hydrated forearc mantle. Geology 33, 445–448. Bostock, M.G., Hyndman, R.D., Rondenay, S., Peacock, S.M., 2002. An inverted continental Moho and serpentinization of the forearc mantle. Nature 417, 536–538. Burov, E., Guillou-Frottier, L., d’Acremont, E., Le Pourhiet, L., Cloetingh, S., 2007. Plume head–lithosphere interactions near intra-continental plate boundaries. Tectonophysics 434, 15–38. Castillo, P.R., 2012. Adakite petrogenesis. Lithos 134–135, 304–316. Currie, C.A., Wang, K., Hyndman, R.D., He, J., 2004. The thermal effects of steadystate slab-driven mantle flow above a subducting plate: the Cascadia subduction zone and backarc. Earth Planet. Sci. Lett. 223, 35–48. Deng, J.F., Mo, X.X., Zhao, H.L., Wu, Z.X., Luo, Z.H., Su, S.G., 2004. A new model for the dynamic evolution of Chinese lithosphere: ‘continental roots–plume tectonics’. Earth-Sci. Rev. 65, 223–275. Ernst, R.E., Buchan, K.L., 2003. Recognizing mantle plumes in the geological record. Annu. Rev. Earth Planet. Sci. 31, 469–523. Foley, S.F., 2008. Rejuvenation and erosion of the cratonic lithosphere. Nat. Geosci. 1, 503–510. French, S.W., Romanowicz, B., 2015. Broad plumes rooted at the base of the Earth’s mantle beneath major hotspots. Nature 525, 95–99. Gao, S., Rudnick, R.L., Xu, W.-L., Yuan, H.-L., Liu, Y.-S., Walker, R.J., Puchtel, I.S., Liu, X., Huang, H., Wang, X.-R., Yang, J., 2008. Recycling deep cratonic lithosphere and generation of intraplate magmatism in the North China Craton. Earth Planet. Sci. Lett. 270, 41–53. Gerya, T.V., Yuen, D.A., 2003. Rayleigh–Taylor instabilities from hydration and melting propel ‘cold plumes’ at subduction zones. Earth Planet. Sci. Lett. 212, 47–62. Gö˘güs, ¸ O.H., 2015. Rifting and subsidence following lithospheric removal in continental back arcs. Geology 43, 3–6. Gurnis, M., Turner, M., Zahirovic, S., DiCaprio, L., Spasojevic, S., Müller, R.D., Boyden, J., Seton, M., Manea, V.C., Bower, D.J., 2012. Plate tectonic reconstructions with continuously closing plates. Comput. Geosci. 38, 35–42. Gutscher, M.-A., Maury, R., Eissen, J.-P., Bourdon, E., 2000. Can slab melting be caused by flat subduction? Geology 28, 535–538. Hassan, R., Flament, N., Gurnis, M., Bower, D.J., Müller, D., 2015. Provenance of plumes in global convection models. Geochem. Geophys. Geosyst. 16, 1465–1489. He, C., Santosh, M., 2016. Seismic tomographic evidence for upwelling mantle plume in NE China. Phys. Earth Planet. Inter. 254, 37–45. Heron, P.J., Lowman, J.P., Stein, C., 2015. Influences on the positioning of mantle plumes following supercontinent formation. J. Geophys. Res., Solid Earth 120, 3628–3648. Hilairet, N., Reynard, B., Wang, Y., Daniel, I., Merkel, S., Nishiyama, N., Petitgirard, S., 2007. High-pressure creep of serpentine, interseismic deformation, and initiation of subduction. Science 318, 1910–1913. Honda, S., Saito, M., 2003. Small-scale convection under the back-arc occurring in the low viscosity wedge. Earth Planet. Sci. Lett. 216, 703–715. Honda, S., Yoshida, T., 2005. Effects of oblique subduction on the 3-D pattern of small-scale convection within the mantle wedge. Geophys. Res. Lett. 32, L13307.

218

I.-C. Ryu, C. Lee / Earth and Planetary Science Letters 479 (2017) 206–218

Hu, J., Liu, L., 2016. Abnormal seismological and magmatic processes controlled by the tearing South American flat slabs. Earth Planet. Sci. Lett. 450, 40–51. Hyndman, R.D., Peacock, S.M., 2003. Serpentinization of the forearc mantle. Earth Planet. Sci. Lett. 212, 417–432. Iida, K., Iwamori, H., Orihashi, Y., Park, T., Jwa, Y.-J., Kwon, S.-T., Danhara, T., Iwano, H., 2015. Tectonic reconstruction of batholith formation based on the spatiotemporal distribution of Cretaceous–Paleogene granitic rocks in southwestern Japan. Isl. Arc 24, 205–220. Imaoka, T., Nakashima, K., Kamei, A., Itaya, T., Ohira, T., Nagashima, M., Kono, N., Kiji, M., 2014. Episodic magmatism at 105 Ma in the Kinki district, SW Japan: petrogenesis of Nb-rich lamprophyres and adakites, and geodynamic implications. Lithos 184–187, 105–131. Isozaki, Y., Aoki, K., Nakama, T., Yanai, S., 2010. New insight into a subductionrelated orogen: a reappraisal of the geotectonic framework and evolution of the Japanese islands. Gondwana Res. 18, 82–105. Jahn, B.-m., Wu, F., Lo, C.-H., Tsai, C.-H., 1999. Crust–mantle interaction induced by deep subduction of the continental crust: geochemical and Sr–Nd isotopic evidence from post-collisional mafic–ultramafic intrusions of the northern Dabie complex, central China. Chem. Geol. 157, 119–146. Karato, S.-I., Wu, P., 1993. Rheology of the upper mantle: a synthesis. Science 260, 771–778. Kim, S.W., Kwon, S., Park, S.-I., Lee, C., Cho, D.-L., Lee, H.-J., Ko, K., Kim, S.J., 2016. SHRIMP U–Pb dating and geochemistry of the Cretaceous plutonic rocks in the Korean Peninsula: a new tectonic model of the Cretaceous Korean Peninsula. Lithos 262, 88–106. Kim, Y.-M., Lee, C., 2014. Effect of time-evolving age and convergence rate of the subducting plate on the Cenozoic adakites and boninites. J. Asian Earth Sci. 95, 300–312. Kincaid, C., Druken, K.A., Griffiths, R.W., Stegman, D.R., 2013. Bifurcation of the Yellowstone plume driven by subduction-induced mantle flow. Nat. Geosci. 6, 395–399. King, S.D., Lee, C., Keken, P.E.v., Leng, W., Zhong, S., Tan, E., Tosi, N., Kameyama, M.C., 2010. A community benchmark for 2-D Cartesian compressible convection in the Earth’s mantle. Geophys. J. Int. 180, 73–87. Kinoshita, O., 1995. Migration of igneous activities related to ridge subduction in Southwest Japan and the East Asian continental margin from the Mesozoic to the Paleogene. Tectonophysics 245, 25–35. Kinoshita, O., 2002. Possible manifestations of slab window magmatisms in Cretaceous southwest Japan. Tectonophysics 344, 1–13. Kneller, E.A., van Keken, P.E., 2008. Effect of three-dimensional slab geometry on deformation in the mantle wedge: implications for shear wave anisotropy. Geochem. Geophys. Geosyst. 9, Q01003. Konter, J.G., Jackson, M.G., 2012. Large volumes of rejuvenated volcanism in Samoa: evidence supporting a tectonic influence on late-stage volcanism. Geochem. Geophys. Geosyst. 13, Q0AM04. Kutsukake, T., 2002. Geochemical characteristics and variations of the Ryoke Granitoids, Southwest Japan: petrogenetic implications for the plutonic rocks of a magmatic arc. Gondwana Res. 5, 355–372. Lee, C.-T.A., Luffi, P., Chin, E.J., 2011. Building and destroying continental mantle. Annu. Rev. Earth Planet. Sci. 39, 59–90. Lee, C., King, S.D., 2009. Effect of mantle compressibility on the thermal and flow structures of the subduction zones. Geochem. Geophys. Geosyst. 10. Lee, C., King, S.D., 2010. Why are high-Mg# andesites widespread in the western Aleutians? A numerical model approach. Geology 38, 583–586. Lee, C., King, S.D., 2011. Dynamic buckling of subducting slabs reconciles geological and geophysical observations. Earth Planet. Sci. Lett. 312, 360–370. Lee, C., Lim, C., 2014. Short-term and localized plume-slab interaction explains the genesis of Abukuma adakite in Northeastern Japan. Earth Planet. Sci. Lett. 396, 116–124. Lee, C., Lim, C., 2016. Three-dimensional time-evolving plume–slab interaction for the generation of the Abukuma adakite in Northeast Japan. Gondwana Res. 38, 99–112. Lee, C., Ryu, I.-C., 2015. A new tectonic model for the genesis of Adakitic Arc magmatism in Cretaceous East Asia. In: Morra, G., Yuen, D., King, S., Lee, S.-M., Stein, S. (Eds.), Subduction Dynamics: from Mantle Flow to Mega Disasters. In: AGU Monograph. Wiley, pp. 69–79. Lewis, Byrne, Pasteris, London, Morgan, V., 2000. Early Tertiary fluid flow and pressure–temperature conditions in the Shimanto accretionary complex of south-west Japan: constraints from fluid inclusions. J. Metamorph. Geol. 18, 319–333. Ma, L., Jiang, S.-Y., Hofmann, A.W., Xu, Y.-G., Dai, B.-Z., Hou, M.-L., 2016. Rapid lithospheric thinning of the North China Craton: new evidence from cretaceous mafic dikes in the Jiaodong Peninsula. Chem. Geol. 432, 1–15. Maruyama, S., Isozaki, Y., Kimura, G., Terabayashi, M., 1997. Paleogeographic maps of the Japanese Islands: plate tectonic synthesis from 750 Ma to the present. Isl. Arc 6, 121–142. Menzies, M., Xu, Y., Zhang, H., Fan, W., 2007. Integration of geology, geophysics and geochemistry: a key to understanding the North China Craton. Lithos 96, 1–21. Mibe, K., Kawamoto, T., Matsukage, K.N., Fei, Y., Ono, S., 2011. Slab melting versus slab dehydration in subduction-zone magmatism. Proc. Natl. Acad. Sci. 108, 8177–8182.

Moore, W.B., Schubert, G., Tackley, P., 1998. Three-dimensional simulations of plume–lithosphere interaction at the Hawaiian swell. Science 279, 1008–1011. Morishige, M., Honda, S., 2011. Three-dimensional structure of P-wave anisotropy in the presence of small-scale convection in the mantle wedge. Geochem. Geophys. Geosyst. 12. Okada, H., 1999. Plume-related sedimentary basins in East Asia during the Cretaceous. Palaeogeogr. Palaeoclimatol. Palaeoecol. 150, 1–11. Ouyang, H., Mao, J., Zhou, Z., Su, H., 2015. Late Mesozoic metallogeny and intracontinental magmatism, southern Great Xing’an Range, northeastern China. Gondwana Res. 27, 1153–1172. Ren, J., Tamaki, K., Li, S., Junxia, Z., 2002. Late Mesozoic and Cenozoic rifting and its dynamic setting in Eastern China and adjacent areas. Tectonophysics 344, 175–205. Ren, Z.-Y., Ingle, S., Takahashi, E., Hirano, N., Hirata, T., 2005. The chemical structure of the Hawaiian mantle plume. Nature 436, 837–840. Ribe, N.M., Christensen, U.R., 1994. 3-Dimensional modeling of plume–lithosphere interaction. J. Geophys. Res., Solid Earth 99, 669–682. Richards, J.P., Kerrich, R., 2007. Special paper: Adakite-like rocks: their diverse origins and questionable role in metallogenesis. Econ. Geol. 102, 537–576. Sakaguchi, A., 1996. High paleogeothermal gradient with ridge subduction beneath the Cretaceous Shimanto accretionary prism, southwest Japan. Geology 24, 795–798. Schmidt, M.W., Poli, S., 1998. Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation. Earth Planet. Sci. Lett. 163, 361–379. Sdrolias, M., Müller, R.D., 2006. Controls on back-arc basin formation. Geochem. Geophys. Geosyst. 7. Sobolev, S.V., Sobolev, A.V., Kuzmin, D.V., Krivolutskaya, N.A., Petrunin, A.G., Arndt, N.T., Radko, V.A., Vasiliev, Y.R., 2011. Linking mantle plumes, large igneous provinces and environmental catastrophes. Nature 477, 312–316. Steinberger, B., Torsvik, T.H., 2012. A geodynamic model of plumes from the margins of Large Low Shear Velocity Provinces. Geochem. Geophys. Geosyst. 13. Syracuse, E.M., Abers, G.A., 2006. Global compilation of variations in slab depth beneath arc volcanoes and implications. Geochem. Geophys. Geosyst. 7. Tarduno, J.A., Duncan, R.A., Scholl, D.W., Cottrell, R.D., Steinberger, B., Thordarson, T., Kerr, B.C., Neal, C.R., Frey, F.A., Torii, M., Carvallo, C., 2003. The emperor seamounts: southward motion of the Hawaiian hotspot plume in Earth’s mantle. Science 301, 1064–1069. Torsvik, T.H., Steinberger, B., Cocks, L.R.M., Burke, K., 2008. Longitude: linking Earth’s ancient surface to its deep interior. Earth Planet. Sci. Lett. 276, 273–282. Turcotte, D., Schubert, G., 2002. Geodynamics, 2nd ed. Cambridge Univ. Press, Cambridge. Wada, I., Wang, K., 2009. Common depth of slab-mantle decoupling: reconciling diversity and uniformity of subduction zones. Geochem. Geophys. Geosyst. 10. Wang, F., Zhou, X.-H., Zhang, L.-C., Ying, J.-F., Zhang, Y.-T., Wu, F.-Y., Zhu, R.-X., 2006a. Late Mesozoic volcanism in the Great Xing’an Range (NE China): timing and implications for the dynamic setting of NE Asia. Earth Planet. Sci. Lett. 251, 179–198. Wang, Q., Wyman, D.A., Xu, J.-F., Zhao, Z.-H., Jian, P., Xiong, X.-L., Bao, Z.-W., Li, C.-F., Bai, Z.-H., 2006b. Petrogenesis of Cretaceous adakitic and shoshonitic igneous rocks in the Luzong area, Anhui Province (eastern China): implications for geodynamics and Cu–Au mineralization. Lithos 89, 424–446. Wee, S.-M., Choi, S.-G., Ryu, I.-C., Shin, H.-J., 2006. Geochemical Characteristics of the Cretaceous Jindong Granites in the Southwestern Part of the Gyeongsang Basin, Korea: focussed on Adakitic Signatures. Econ. Environ. Geol. 39, 555–566. Whittaker, J.M., Müller, R.D., Leitchenkov, G., Stagg, H., Sdrolias, M., Gaina, C., Goncharov, A., 2007. Major Australian-Antarctic Plate Reorganization at HawaiianEmperor Bend Time. Science 318, 83–86. Wu, F.-Y., Lin, J.-Q., Wilde, S.A., Zhang, X.O., Yang, J.-H., 2005. Nature and significance of the Early Cretaceous giant igneous event in eastern China. Earth Planet. Sci. Lett. 233, 103–119. Wu, F.-y., Sun, D.-y., Li, H., Jahn, B.-m., Wilde, S., 2002. A-type granites in northeastern China: age and geochemical constraints on their petrogenesis. Chem. Geol. 187, 143–173. Xu, J.-F., Shinjo, R., Defant, M.J., Wang, Q., Rapp, R.P., 2002. Origin of Mesozoic adakitic intrusive rocks in the Ningzhen area of east China: partial melting of delaminated lower continental crust? Geology 30, 1111–1114. Xu, Y.G., 2001. Thermo-tectonic destruction of the Archean lithospheric keel beneath the Sino-Korean craton in China: evidence, timing and mechanism. Physics and chemistry of the Earth, part a: solid Earth and Geodesy 26, 747–757. Zhang, F.Q., Chen, H.L., Yu, X., Dong, C.W., Yang, S.F., Pang, Y.M., Batt, G.E., 2011. Early Cretaceous volcanism in the northern Songliao Basin, NE China, and its geodynamic implication. Gondwana Res. 19, 163–176. Zhu, G., Gerya, T.V., Honda, S., Tackley, P.J., Yuen, D.A., 2011. Influences of the buoyancy of partially molten rock on 3-D plume patterns and melt productivity above retreating slabs. Phys. Earth Planet. Inter. 185, 112–121.