4:30 PM - 5:00 PM
[SCG62-15] Dynamics of the lithosphere-asthenosphere boundary decoded from the mantle xenoliths from Ichinomegata maar, NE Japan
★Invited Papers
Keywords:mantle xenolith, lithosphere-asthenosphere boudnary, rheology, strain rate, viscosity structure
The lithosphere-asthenosphere boundary (LAB) represents the "bottom" of plates in the framework of plate tectonics, marking a viscosity contrast in the upper mantle. To better understand the dynamics operating within the LAB, information on rheology and deformation history extracted from mantle xenoliths, which are fragments of mantle material transported by magma, is crucial. This study aimed to reconstruct the viscosity contrast across the LAB in subduction zone settings using eight spinel peridotite xenoliths from the Ichinomegata maar, Oga Peninsula, northeast Japan. From the Ichinomegata xenoliths, the following petrologic information was extracted: ① temperature, ② pressure (depth), ③ mineral composition, ④ chemical composition (water content), ⑤ deformation microstructure, and ⑥ deformation history. These parameters were integrated and applied to the experimentally determined olivine flow law to reconstruct the structure and dynamics of the LAB.
For geothermobarometry of the spinel peridotites, thermal history was decoded from the chemical zoning of constituent minerals to determine the equilibrium chemical compositions just before xenolith extraction. Using Ca-in-OPX and two-pyroxene geothermobarometers, a geotherm of 0.7-1.6 GPa, 28-54 km, and 832-1084 °C was reconstructed. Based on the estimated depth and petrologic information, a layered structure in the mantle was identified: the shallow layer (28-32 km) exhibited low temperature, coarse-grained (0.415-0.814 mm, olivine φ), equigranular texture, and hydrous minerals (amphibole and fluid inclusions), while the deeper layer (41-55 km) showed high temperature, fine-grained (0.147-0.293 mm), porphyroclastic texture, and interstitial melt (hydrous glass).
The water contents of olivine, orthopyroxene, and clinopyroxene were estimated using FTIR and SIMS, yielding 18-33 ppm, 205-468 ppm, and 446-787 ppm, respectively. The distribution coefficients (DOPX/OL=14±3, DCPX/OPX=2.1±0.3) were consistent with experimentally determined values. The estimated water content did not correlate with the layered mantle structure and showed no depth variation. Water diffusion profiles revealed dehydration within ~200 μm from the mineral edges, with diffusion times estimated to be between 20 seconds and 48 minutes. This time scale is extremely short compared to the xenolith transportation time by magma (1-68 days), suggesting that water information at mantle depths is well-preserved in the core of minerals.
The lattice-preferred orientation (LPO) of olivine showed depth variation: the shallow layer (28-32 km) mainly exhibited A-type LPO, while the deeper layer (41-55 km) showed mainly D-type and AG-type LPO. Grain average misorientation (GAM) indicated no grain size dependence in xenoliths from the shallower layer, suggesting uniform deformation in the shallow mantle. In contrast, GAM showed a grain size dependence in xenoliths from the deeper layer, indicating that the deformation mechanism in the shallow mantle was dislocation creep, and dynamic recrystallization in the deeper mantle.
Xenoliths from the deeper layer contained peculiar deformation microstructures ("spinel blebs") which originated from the exsolution spinel lamellae in pyroxene and formed via grain boundary migration. This microstructure was utilized to decode the deformation history in the framework of thermal history. The deformation microstructures in the deeper layer were established during heating (~12 kyrs) before xenolith extraction.
Based on the above petrologic information, the strain rate and viscosity structures beneath Ichinomegata were reconstructed using the experimentally determined flow law of hydrated olivine polycrystals. The strain rate in the shallow mantle (28-32 km) was estimated to be 10-16 to 10-12 s-1, while in the deeper mantle (41-55 km), it was 10-12 to 10-10 s-1. Viscosity in the shallow mantle was estimated to be 1019 to 1023 Pa·s, and it was 1017 to 1019 Pa·s in the deeper mantle, consistent with the viscosity of the asthenosphere estimated by geophysical observations. Therefore, the mantle structure beneath Ichinomegata is interpreted as consisting of the lithospheric mantle (25-40 km), top of the asthenosphere (40-55 km), with the LAB located at ~40 km depth. This mantle structure was dynamically formed through lithospheric thinning ~12 kyrs before the eruption of the Ichinomegata maar.
For geothermobarometry of the spinel peridotites, thermal history was decoded from the chemical zoning of constituent minerals to determine the equilibrium chemical compositions just before xenolith extraction. Using Ca-in-OPX and two-pyroxene geothermobarometers, a geotherm of 0.7-1.6 GPa, 28-54 km, and 832-1084 °C was reconstructed. Based on the estimated depth and petrologic information, a layered structure in the mantle was identified: the shallow layer (28-32 km) exhibited low temperature, coarse-grained (0.415-0.814 mm, olivine φ), equigranular texture, and hydrous minerals (amphibole and fluid inclusions), while the deeper layer (41-55 km) showed high temperature, fine-grained (0.147-0.293 mm), porphyroclastic texture, and interstitial melt (hydrous glass).
The water contents of olivine, orthopyroxene, and clinopyroxene were estimated using FTIR and SIMS, yielding 18-33 ppm, 205-468 ppm, and 446-787 ppm, respectively. The distribution coefficients (DOPX/OL=14±3, DCPX/OPX=2.1±0.3) were consistent with experimentally determined values. The estimated water content did not correlate with the layered mantle structure and showed no depth variation. Water diffusion profiles revealed dehydration within ~200 μm from the mineral edges, with diffusion times estimated to be between 20 seconds and 48 minutes. This time scale is extremely short compared to the xenolith transportation time by magma (1-68 days), suggesting that water information at mantle depths is well-preserved in the core of minerals.
The lattice-preferred orientation (LPO) of olivine showed depth variation: the shallow layer (28-32 km) mainly exhibited A-type LPO, while the deeper layer (41-55 km) showed mainly D-type and AG-type LPO. Grain average misorientation (GAM) indicated no grain size dependence in xenoliths from the shallower layer, suggesting uniform deformation in the shallow mantle. In contrast, GAM showed a grain size dependence in xenoliths from the deeper layer, indicating that the deformation mechanism in the shallow mantle was dislocation creep, and dynamic recrystallization in the deeper mantle.
Xenoliths from the deeper layer contained peculiar deformation microstructures ("spinel blebs") which originated from the exsolution spinel lamellae in pyroxene and formed via grain boundary migration. This microstructure was utilized to decode the deformation history in the framework of thermal history. The deformation microstructures in the deeper layer were established during heating (~12 kyrs) before xenolith extraction.
Based on the above petrologic information, the strain rate and viscosity structures beneath Ichinomegata were reconstructed using the experimentally determined flow law of hydrated olivine polycrystals. The strain rate in the shallow mantle (28-32 km) was estimated to be 10-16 to 10-12 s-1, while in the deeper mantle (41-55 km), it was 10-12 to 10-10 s-1. Viscosity in the shallow mantle was estimated to be 1019 to 1023 Pa·s, and it was 1017 to 1019 Pa·s in the deeper mantle, consistent with the viscosity of the asthenosphere estimated by geophysical observations. Therefore, the mantle structure beneath Ichinomegata is interpreted as consisting of the lithospheric mantle (25-40 km), top of the asthenosphere (40-55 km), with the LAB located at ~40 km depth. This mantle structure was dynamically formed through lithospheric thinning ~12 kyrs before the eruption of the Ichinomegata maar.