Distinct compositional thin layers at mid-mantle depths beneath northeast China revealed by the USArray
Highlights
► USArray recorded anomalous seismic arrivals from two deep quakes in NE China. ► The source structures are seismically distinct thin layers in the lower mantle. ► The thin layers could be subducted oceanic crust. ► Oceanic crust may be ubiquitously present in the lower mantle of subduction zones.
Introduction
Many seismic studies have revealed that the Earth's lower mantle is featured by two large lower velocity structures beneath Africa and central Pacific, known as the Africa and Pacific Large-Low-Shear-Velocity Provinces (LLSVPs), surrounded by seismically fast anomalies that are interpreted as the subducted slabs in later Mesozoic and Cenozoic (Dziewonski, 1984, Su et al., 1994, Grand et al., 1997, van der Hilst et al., 1997, Ritsema et al., 1999). The origin and evolution of the two LLSVPs, however, have remained poorly understood (e.g., Ohta et al., 2008, Schuberth et al., 2009, Zhang et al., 2010, Dziewonski et al., 2010, Nakagawa and Tackley, 2010). There are speculations that the LLSVPs are piles of dense mid-ocean-ridge basalt (MORB) accumulated above the core–mantle boundary (CMB) (Ohta et al., 2008, Nakagawa and Tackley, 2010). Mapping subducted oceanic crust in the lower mantle thus is crucial to understanding whether oceanic crust can sink to the CMB to form the two large piles.
Kaneshima and Helffrich, 1998, Kaneshima and Helffrich, 1999 observed strong coda waves from teleseismic recordings in western US from deep earthquakes occurring in Mariana and interpreted them as the S to P converted waves from scatterers in the mid-lower mantle depths. The scatterers appeared to be thin lower velocity layers with high dipping angles, which led them to speculate that the scatterers may be former oceanic crust subducted into the lower mantle. In this paper, we define a volumetric anomaly with an arbitrary shape as a scatterer, and a thin layer with distinct velocity/density as a reflector. As such a reflector is a special type of the broadly defined scatterers. Niu et al. (2003) found a strong seismic reflector at ∼1100 km deep beneath the Mariana subduction zone. The reflector is a ∼12 km thick layer with an S-wave velocity of 2–6% lower and a density 2–9% higher than those of the surrounding mantle. The difference in P-wave velocity is rather trivial (<1%). Due to the lack of data on the elasticity of oceanic crust at lower mantle condition at the time, it was not conclusive whether the observed reflector is a piece of subducted oceanic crust or not. In addition to these studies, several recent studies also suggested these reflectors exist over a wide range of depth (∼800–1850 km) beneath various subduction zones surrounding the Pacific (e.g., Kaneshima and Helffrich, 2003, Kaneshima and Helffrich, 2009, Kaneshima and Helffrich, 2010).
Besides the lower mantle reflectors, there were also reports of the existence of sharp discontinuities at depths from 900 to 1100 km in the middle mantle (Kawakatsu and Niu, 1994, Niu and Kawakatsu, 1997, Vinnik et al., 1998, Vanacore et al., 2006). Possible causes of these mid-mantle discontinuities, however, have remained poorly understood. It is unclear whether these discontinuities are related to the seismic reflectors as mentioned above. In this study, we analyze an unusual later arrival recorded by the USArray from two deep earthquakes occurring in NE China. We employ seismic imaging techniques to locate the source structures and utilize forward waveform modeling to estimate the seismic properties of these structures. With the newly available mineral physics data, our goal is to better constrain the nature of the lower mantle reflectors, as well as to understand whether the two types of seismic structures being reported previously are caused by resolution ambiguity in the short-period data.
Section snippets
Data and analysis
The seismic data used in this study are recorded by the transcontinental USArray (Fig. 1), which consists of 400+ transportable broadband seismic stations. Although the primary goal of the USArray deployment is to obtain high-resolution seismic images of the North American lithosphere, with an aperture of more than 2000 by 1000 km, the USArray is also ideally suited for detecting weak signals associated with small-scale 3D structures within Earth's deep interior around the globe. We use
Results and discussion
The vespagram of the 05/19/2008 event is shown in Fig. 3a. The ∼40 s later arrival exhibits a negative relative slowness of 0.21 s/deg. with respect to the P wave, suggesting that it is indeed an S to P converted phase from an anomalous structure below the source. The vespagram also shows another S to P conversion phase at 12.5 s with a slightly negative slowness, −0.03 s/deg. (Fig. 3a). We use iasp91 velocity model and the relocated focal depth of 520 km, which was determined by Wang and Niu (2010)
Conclusion
We observe an unexpected later arrival at the USArray records of two deep earthquakes beneath NE China. Using array processing techniques, we find that the arrival is best explained by an S to P conversion from a dipping structure at about ∼1000 km. While synthetic tests suggest either a discontinuity or a thin low velocity layer can explain the observed short-period data, we argue that the latter is the more physically likely source for the S to P conversion, as the required seismic properties
Acknowledgments
We thank IRIS and EarthScope for providing the data used in this study, Steve Grand, George Helffrich for helpful discussion, and two anonymous reviewers for their constructive comments and suggestions. This research is supported by the NSF grant EAR-0748455.
References (40)
- et al.
Mantle anchor structure: an argument for bottom up tectonics
Earth Planet Sci. Lett.
(2010) - et al.
Thermomechanical modelling of slab detachment
Earth Planet. Sci. Lett.
(2004) - et al.
Lower mantle scattering profiles and fabric below Pacific subduction zones
Earth Planet Sci. Lett.
(2009) - et al.
Small scale heterogeneity in the mid-lower mantle beneath the circum-Pacific area
Phys. Earth Planet. Inter.
(2010) - et al.
Receiver functions in northeast China—implications for slab penetration into the lower mantle in northwest Pacific subduction zone
Earth Planet. Sci. Lett.
(2003) - et al.
Seismic constraints on the depth and composition of the mantle keel beneath the Kaapvaal craton
Earth Planet. Sci. Lett.
(2004) - et al.
Phase transition in pyrolite and MORB at lowermost mantle convections: implications for a MORB-rich pile above the core-mantle boundary
Earth Planet. Sci. Lett.
(2008) Elasticity of subducted basaltic crust at the lower mantle pressures: insights on the nature of deep mantle heterogeneity
Phys. Earth Planet. Inter.
(2011)- et al.
Grain growth kinetics of ringwoodite and its implication for rheology of the subducting slab
Earth Planet. Sci. Lett.
(2005) - et al.
Quantitative Seismology: Theory and Methods
(1980)
Mapping the lower mantle: determination of lateral heterogeneity in P-velocity up to degree and order 6
J. Geophys. Res.
180 million years of subduction
GSA Today
Stagnant slabs in the upper and lower mantle transition zone
Geophys. Rev.
Receiver functions in the western United States, with implications for upper mantle structure and dynamics
J. Geophys. Res.
Global seismic tomography: a snapshot of convection in the Earth
GSA Today
Mantle shear-wave tomography and the fate of subducted slabs
Philosophical Transactions - Royal Society, Mathematical, Physical and Engineering Sciences
Crustal reflections of the plane P and SV waves
J. Geophys. Res.
High-resolution mantle tomography of China and surrounding regions
J. Geophys. Res.
Detection of lower mantle scatterers northeast of the Mariana subduction zone using short-period array data
J. Geophys. Res.
Dipping lower-velocity layer in the mid-lower mantle: evidence for geochemical heterogeneity
Science
Cited by (46)
Mid-mantle seismic scatterers beneath the Samoan hotspot
2023, Physics of the Earth and Planetary InteriorsHigh thermal conductivity of stishovite promotes rapid warming of a sinking slab in Earth's mantle
2022, Earth and Planetary Science LettersCitation Excerpt :Additionally, stishovite undergoes a ferro-elastic post-stishovite structural transition from a tetragonal (rutile, P42/mnm) to orthorhombic (CaCl2-type, Pnnm) structure at ∼50–60 GPa (Andrault et al., 1998; Bolfan-Casanova et al., 2009; Hirose et al., 2005; Kingma et al., 1995). Such a phase transition occurs with a pronounced shear-wave velocity reduction anomaly (Carpenter et al., 2000; Yang and Wu, 2014; Zhang et al., 2021a) that has been used to explain the occurrence of regional seismic anomalies in some scatterers observed along subduction zones in the shallow to mid lower mantle (Kaminsky, 2017; Kaneshima, 2019; Kaneshima and Helffrich, 1999; Niu, 2014; Nomura et al., 2010), such as the Tonga subduction zone. However, due to experimental and computational challenges, very little is known about stishovite's thermal conductivity at relevant high P–T conditions of the mantle, and the literature results are not consistent with each other (e.g., (Aramberri et al., 2017; Osako and Kobayashi, 1979; Yukutake and Shimada, 1978)).
Southeastward dipping mid-mantle heterogeneities beneath the sea of Okhotsk
2021, Earth and Planetary Science LettersThe dimensions of scatterers in the lower mantle using USArray recordings of S-wave to P-wave conversions
2020, Physics of the Earth and Planetary InteriorsCaCO<inf>3</inf> phase diagram studied with Raman spectroscopy at pressures up to 50 GPa and high temperatures and DFT modeling
2018, Physics of the Earth and Planetary InteriorsSeismic scatterers in the mid-lower mantle beneath Tonga-Fiji
2018, Physics of the Earth and Planetary InteriorsCitation Excerpt :For these pairs the amplitudes of the S-to-P waves are usually 3 to 10 % of direct P except for a few cases exceeding 20 % (Tables S1 to S17). The amplitudes are comparable to what were observed for Mariana (Kaneshima and Helffrich, 1999) and for Japan Sea (Niu, 2014), so that the elastic anomalies of the scatterers likely exceed 5 %. We find 34 event-array pairs that meet the criterion.