*Yuki Suzuki1, HOKANISHI Natsumi2, Fukashi Maeno2, ATSUSHI YASUDA2
(1.Department of Earth Sciences, Faculty of Education and Integrated Arts of Sciences, Waseda University, 2.Earthquake Research Institute, University of Tokyo)
Keywords:Pre-eruption process, Shinmoe-dake, Sub-Plinian eruption, Magnetite, Multi-stage magma mixing, Magma mingling
The Kyoho eruption is larger in scale than recent eruptions including 2011 eruption. An outcrop having all units (SmKP-1-7) related to the sub-Plinian phase was examined. The volcanic ash units (SmKP-1 and SmKP-3) were excluded. The subunits were defined by component size and type. Pumice with homogeneous color was defined as brown/gray pumice or white pumice. White pumice appears in negligible amount. Banded pumices are those with white parts exceeding 10%. In SmKP-2, SmKP-4-5, and SmKP-6, the proportion of banded pumice was higher in upper horizons. The bulk density of pumice decreases toward upper horizon of each, indicating upper horizon is the eruption peak. Petrological studies revealed two common endmember magmas were involved throughout the eruption. Whole-rock analysis was performed using the RIGAKU ZSX Primus II at ERI, and acquisition of BSE image and phenocryst analysis were performed using the JXA-8530FPlus and JXA-8800R at ERI. The whole-rock SiO2 content was 57.3~59.3wt.% for brown/gray pumice, and 61.1wt.% for white pumice. The white pumice represents either the low-temperature endmember itself or mixing product. The decreased contribution of high-temperature endmember in upper part of each unit (SmKP-2, SmKP-4-5, and SmKP-6) was indicated not only by the increased ratio of banded pumice but also increased SiO2 content of brown/gray pumice. The groundmass of brown/gray pumice contains small high-SiO2 parts (resembling that of white pumice) which were formed by magma mingling. The ratio of high-SiO2 parts in pumice is higher in the upper horizon, which is linked to increase in SiO2 content of brown/gray pumice. During the eruption preparation process, the viscosity of the mush magma (low-temperature endmember) decreased due to mixing with the high-temperature magma. This process was examined by MgO profile of magnetite from the low-temperature endmember. SmKP-2 was examined. Line analysis of about 10 magnetites was performed for gray/brown pumices from the Lower and Upper subunits, respectively. Magnetite has reverse zoning or flat profile. The MgO content in rim roughly reflects the composition of the groundmass in which the crystal is located. In the Upper pumice with more heterogeneity in the groundmass, two types of magnetites (low and high MgO contents in the rims) coexist. The core of the magnetite with reverse zoning is relatively homogeneous, and the MgO content of the core is higher than that expected for magnetite in endmember magma. This core had formed in a prior mixing event before the magma mixing event that left the reversed zoning in the magnetite. Enough time after the prior magma mixing event once homogenized the magnetite. After the two magma mixing events described above, the magma reservoir came to have parts where mixing with high-temperature magma was prominent (corresponding to the gray/brown parts of pumice) and parts not prominent (corresponding to the white parts of pumice). Then two parts underwent magma mingling in the conduit. When the magma moved to the conduit, the parts with lower viscosity moved preferentially. Therefore, the contribution of high-temperature magma decreased with time. We investigated the time from the magma mixing event immediately preceding the eruption to the eruption. We estimated the time reproducing actual MgO profile, assuming crystals with core composition had overgrowth with rim composition upon the mixing and that element diffusion occurred between the core and the rim. The time is concentrated in the range of several hours to about 30 hours for both the Lower and Upper subunits and does not change with MgO content of the magnetite rim. It is likely that the magma mixing immediately prior to the eruption occurred at roughly the same time for all magmas, regardless of the composition of the magma in the mush-like magma reservoir and the timing of the final eruption and deposition.