2:30 PM - 2:45 PM
[SCG54-16] A new hypothesis of the water transportation to the mantle transition zone obtained by receiver function imaging
★Invited Papers
Keywords:Pacific plate, serpentinite layer, subduction
Geophysical observations have revealed that large amounts of water exist in the mantle transition zone beneath subduction zones like Northeastern Japan (e.g., Kelbert et al., 2009). While this water is thought to be transported by subducting plates, its transport mechanisms are not well understood. Most water stored in the subducting plate is released by dehydration reaction of the oceanic crust. The dehydration depth depends on the temperature of the slab, with colder slabs able to transport water to greater depth. However, even in colder plates like those beneath Northeastern Japan, major dehydration reactions are thought to complete by ~95 km (e.g., Katayama et al., 2010). Meanwhile, water released from the oceanic crust reacts with peridotite in the mantle wedge to form serpentinite layer just above the slab surface. This serpentinite layer subducts with the slab and becomes a source of water transport to depths exceeding 200 km (e.g., Iwamori 2000). Furthermore, when the slab maintains low temperatures, serpentinite can transform into Phase A, enabling water transport to even greater depths (Iwamori 2000; Komabayashi et al., 2005). However, very few studies have directly detected these hydrous layers above the slab, thus the spatial distribution of hydrous layers poorly understood (Kawakatsu and Watada 2007; Tonegawa et al., 2008). Therefore, in this study, we developed a new receiver function imaging technique specialized for analyzing dipping structures to comprehensively investigate the spatial distribution of hydrous layers above the slab in the Northeastern Japan subduction zone.
In this study, we used seismic waveforms from 1086 earthquakes with magnitudes greater than 5.5 and epicentral distances between 30º-90º that occurred from April 2005 to March 2023. First, we applied an instrumental response correction (Maeda et al., 2011) to waveform data and calculated receiver functions with a water level of 0.001 in a frequency range of 0.1–0.5 Hz with high signal-to-noise ratio. The obtained receiver functions were migrated to cross-sections using the IASP91 1D velocity model (Kennett and Engdahl, 1991). During this process, we considered the geometry of the surface of the Pacific plate (Nakajima et al., 2009) and mathematically evaluated the effects of refraction at the slab surface. Furthermore, we developed and applied a new amplitude-correction method. Finally, we established 17 profiles across the Northeastern Japan subduction zone and generated cross-sectional images using a new stacking technique that accounts for slab geometry.
Our analysis revealed three distinct velocity boundaries corresponding, respectively, to the slab surface, oceanic Moho, and the bottom of the hydrous layer in almost all profiles. The velocity contrast at the slab surface and oceanic Moho disappears at depths of 90–160 km, which reflects the depth of eclogite phase transformation and dehydration reaction of the oceanic crust. The depth of this eclogite phase transformation shows regional variation, occurring ~30 km deeper beneath the Kanto region compared to the Tohoku region. This distribution is consistent with the model in which the Pacific slab maintains lower temperatures due to the overlapped Philippine Sea slab (e.g., Hasegawa et al., 2007). Furthermore, our results show that the hydrous layer is continuously distributed to the mantle transition zone. This indicates that the mantle wedge serves as a major pathway for water transport to the mantle transition zone. Meanwhile, a velocity contrast interpreted as the upper and lower surface of the metastable olivine wedge was imaged within the slab at depth of 300–400 km. This finding suggests that the slab mantle is essentially dry and the water transport by the slab mantle to the mantle transition zone is minor in this region.
In this study, we used seismic waveforms from 1086 earthquakes with magnitudes greater than 5.5 and epicentral distances between 30º-90º that occurred from April 2005 to March 2023. First, we applied an instrumental response correction (Maeda et al., 2011) to waveform data and calculated receiver functions with a water level of 0.001 in a frequency range of 0.1–0.5 Hz with high signal-to-noise ratio. The obtained receiver functions were migrated to cross-sections using the IASP91 1D velocity model (Kennett and Engdahl, 1991). During this process, we considered the geometry of the surface of the Pacific plate (Nakajima et al., 2009) and mathematically evaluated the effects of refraction at the slab surface. Furthermore, we developed and applied a new amplitude-correction method. Finally, we established 17 profiles across the Northeastern Japan subduction zone and generated cross-sectional images using a new stacking technique that accounts for slab geometry.
Our analysis revealed three distinct velocity boundaries corresponding, respectively, to the slab surface, oceanic Moho, and the bottom of the hydrous layer in almost all profiles. The velocity contrast at the slab surface and oceanic Moho disappears at depths of 90–160 km, which reflects the depth of eclogite phase transformation and dehydration reaction of the oceanic crust. The depth of this eclogite phase transformation shows regional variation, occurring ~30 km deeper beneath the Kanto region compared to the Tohoku region. This distribution is consistent with the model in which the Pacific slab maintains lower temperatures due to the overlapped Philippine Sea slab (e.g., Hasegawa et al., 2007). Furthermore, our results show that the hydrous layer is continuously distributed to the mantle transition zone. This indicates that the mantle wedge serves as a major pathway for water transport to the mantle transition zone. Meanwhile, a velocity contrast interpreted as the upper and lower surface of the metastable olivine wedge was imaged within the slab at depth of 300–400 km. This finding suggests that the slab mantle is essentially dry and the water transport by the slab mantle to the mantle transition zone is minor in this region.