日本地球惑星科学連合2023年大会

講演情報

[J] 口頭発表

セッション記号 M (領域外・複数領域) » M-IS ジョイント

[M-IS13] 地質学のいま

2023年5月24日(水) 13:45 〜 15:15 展示場特設会場 (2) (幕張メッセ国際展示場)

コンビーナ:辻森 樹(東北大学)、小宮 剛(東京大学大学院総合文化研究科広域科学専攻)、山口 飛鳥(東京大学大気海洋研究所)、尾上 哲治(九州大学 大学院理学研究院 地球惑星科学部門)、座長:河上 哲生(京都大学大学院理学研究科)、辻森 樹(東北大学)

14:00 〜 14:15

[MIS13-13] Quantifying metamorphic elemental transfer history for fluid-rock interaction by combining machine learning and mineral records

★Invited Papers

*松野 哲士1宇野 正起1岡本 敦1 (1.東北大学)


In the subduction zone, the fluid flow within the plate boundary induces earthquakes, promotes volcanism, and drives global elemental cycling. The direct evidence of fluid flow remains in chemical alteration of metamorphic rock originated from plate boundaries. The bulk compositional analysis of metamorphic rocks have revealed the qualitative element trends with metamorohic grades (Bebout 2007); however, the bulk composition is the summation of the initial composition and chemical alteration of several fluid-rock interactions, and there is difficulty to setarating total element transfer into that of individual fluid-rock interaction. The recent methodological advances with geochemical machine learning (Matsuno et al. 2022) enable us to estimate initial composition and quantify element transfer for each sample. In addition, the mineral assemblage and chemistry provide the historical records for individual fluid-rock interactions: there is a possibility of quantifying element transfer for individual fluid-rock interactions. In this study, we establish the combination methodology with machine learning and mineral records and perform the case study to reveal fluid flow within the plate boundary.

To understand the effects of lithological boundaries on fluid flow at subduction zones, we investigated a mafic-pelitic shist boundary exposed in the Asemigawa area, Central Shikoku, the Sanbagawa metamorphic belt. The sampling point is Karagoshi outcrop (Latitude:33.7950, Longitude:133.5516) located within the chlorite zone, where the peak metamorphic P-T condition is estimated as 300°C and 0.55-0.65GPa (Kouketsu et al. 2021; Enami et al. 1994). We performed traverse sampling (~10m) from the mafic-pelitic shist boundary to mafic shist and analyzed major/trace element concentrations, mineral assemblage, and mineral chemistry. For each sample, the amount of element transfer is estimated with Protolith Reconstruction Models for metabasalt (PRMs) based on machine learning, which assumes immobile elements, input their concentrations in metabasalt, and output igneous protolith element concentration. We assumed Th, Nb, Zr, and Ti as immobile elements and obtained the amount of element transfer for Rb, Ba, U, K, La, Ce, Pb, Sr, P, Nd, Y, Yb, and Lu. As the index of hydration reaction of retrograde metamorphism, we adapt the reaction progress of Yhydration, originally defined in Okamoto et al. (2005), which is the ratio of the area of actinolite to the total area of amphibole.

The typical mineral assemblage in the mafic shist is chlorite, epidote, amphibole, plagioclase, quartz, and calcite, varying continuously from chlorite rich, actinolite rich, and winchite and actinolite rich as moved away from the boundary, indicating increasing hydration towards the mafic-pelitic boundary. The mafic shist represents the N-MORB like trace element composition and have positive anomaly in Rb, Ba, K, Pb, and Sr relative to N-MORB. The concentration profile of Li, Rb, and Ba continuously enriched 4-1.5 times compared to samples away from boundaries, which is >5m away from the mafic-pelitic shist boundary. The amount of element transfer estimated by PRMs represents more than 10 times the enrichment in Rb and Ba for all samples. The enrichments of Rb and Ba also continuously increase at the mafic-pelitic shist boundary. These results indicate that element enrichment and hydration reaction are continuous increasing towards the mafic-pelitic boundary, which represents the concentrated fluid flow along the lithological boundary. We will obtain the continuous mineral mode and Yhydration for the series of samples and conduct the quantitative comparison of element transfer and mineral records, and finally quantify the element transfer during the hydration reaction.

References:

Bebout, Gray E. 2007. EPSL.

Kouketsu et al. 2021. J. Metamorph. Geol.

Enami et al. 1994. CMP.

Matsuno et al. 2022. Sci.Rep.

Okamoto & Toriumi. 2005. J. Metamorph. Geol.