16:15 〜 16:30
[SVC25-09] Multiple evidence for episodic melt extraction and eruption in large silicic magma systems from Maninjau caldera, Indonesia: Geochemical and paleomagnetic constraints
キーワード：インドネシア、マニンジャウカルデラ、カルデラ形成噴火、低斑晶量流紋岩、古地磁気による時間スケールの制約
Explosive caldera-forming eruptions are catastrophic phenomena that evacuate voluminous silicic magma exceeding tens of km3. In the archetypal model, caldera-forming eruptions occur on very short time scales (commonly weeks to months) from a single, large, and homogeneous magma chamber. However, recent studies provide evidence that some caldera-forming eruptions had intermittent breaks in their activity, with eruptions occurring spasmodically from multiple silicic magma chambers (e.g., Hasegawa et al., 2018; Pitcher et al., 2021; Wilson et al., 2021; Caron et al., 2023). Therefore, it is important to evaluate the relationship between the origin of silicic magma and eruption time scale with a high resolution to better understand the long-term hazard assessments of caldera-forming eruptions.
In this study, we propose a new model of episodically established silicic magma systems of Maninjau caldera (one of the largest calderas in Indonesia) based on new chronological, geochemical, and paleomagnetic data on Maninjau ignimbrites. The Maninjau caldera-forming eruption deposits consist of voluminous ignimbrites (220–250 km3) with no reports of precursory fall deposits. The ignimbrites are composed of three Units (A to C in ascending order) based on the topographic relationships, field distributions, and componentry (Suhendro et al., 2022). This is in good agreement with the 14C dating results as the estimated age for Unit A and Unit C is 52±3 ka (Alloway et al., 2004) and 48±1 ka (newly obtained in this study), respectively.
The juvenile material of Maninjau ignimbrites consists of crystal-poor (~3 %) white pumice and crystal-rich (37–60%) gray pumice. Generally, white pumice dominates in all Units, gray pumice is absent in Unit A, very rare in Unit B, and common (2–6%) in Unit C (Suhendro et al., 2022). White pumice has rhyolitic compositions (71–78 wt.% SiO2bulk), with a tendency of increasing SiO2 content from Unit A to Unit C (Fig. 1). Gray pumice has andesitic to dacitic compositions (62–67 wt.% SiO2bulk) with a high Zr/Hf ratios. Notably, the glass compositions of gray pumice (70–78 wt.% SiO2glass) overlap well with the least-evolved whole-rock compositions of white pumice (Fig. 1). The petrological features of crystal-rich gray pumice and its high Zr/Hf ratio suggest that it was resulting from a crystal accumulation process like a mush zone. The compositional similarity between gray pumice glass chemistry and white pumice whole-rock chemistry (Fig. 1) suggests that interstitial melt in the mush zone was extracted to generate a crystal-poor magma chamber at a shallower level that eventually erupted to form the white pumice. Slight compositional variations of white pumice from Units A to C suggest that conditions like depth, temperature, and water contents of each melt extraction and/or magma chamber may have differed when each unit erupted. Also, our newly obtained 14C age for Unit C (48±1 ka) is hundreds of years younger than that of Unit A (52±3 ka), indicating episodic and distinct extractions and eruptions of silicic magmas of Units A to C. In addition, each unit shows significantly different paleomagnetic directions beyond the 95% confidence limits (Unit A: Dm=4.6°, Im=-14.0°, α95=6.5°; Unit B: Dm=352.1°, Im=-15.0°, α95=1.0°; Unit C: Dm=359.5°, Im=4.9°, α95=3.8°) that strongly support remarkable time gaps between the three units.
Reference
Hasegawa et al. (2018) J.Geogr.(Chigaku Zasshi), 127, 273-288; Wilson et al. (2021) Nat. Rev. Earth Environ., 2, 610–627; Pitcher et al. (2021) J. Petrol., 62, 1–30; Caron et al. (2023) Sci. Rep., 13, 11575; Suhendro et al. (2022) JVGR, 431, 107643; Alloway et al. (2004) EPSL, 227, 121–133.
In this study, we propose a new model of episodically established silicic magma systems of Maninjau caldera (one of the largest calderas in Indonesia) based on new chronological, geochemical, and paleomagnetic data on Maninjau ignimbrites. The Maninjau caldera-forming eruption deposits consist of voluminous ignimbrites (220–250 km3) with no reports of precursory fall deposits. The ignimbrites are composed of three Units (A to C in ascending order) based on the topographic relationships, field distributions, and componentry (Suhendro et al., 2022). This is in good agreement with the 14C dating results as the estimated age for Unit A and Unit C is 52±3 ka (Alloway et al., 2004) and 48±1 ka (newly obtained in this study), respectively.
The juvenile material of Maninjau ignimbrites consists of crystal-poor (~3 %) white pumice and crystal-rich (37–60%) gray pumice. Generally, white pumice dominates in all Units, gray pumice is absent in Unit A, very rare in Unit B, and common (2–6%) in Unit C (Suhendro et al., 2022). White pumice has rhyolitic compositions (71–78 wt.% SiO2bulk), with a tendency of increasing SiO2 content from Unit A to Unit C (Fig. 1). Gray pumice has andesitic to dacitic compositions (62–67 wt.% SiO2bulk) with a high Zr/Hf ratios. Notably, the glass compositions of gray pumice (70–78 wt.% SiO2glass) overlap well with the least-evolved whole-rock compositions of white pumice (Fig. 1). The petrological features of crystal-rich gray pumice and its high Zr/Hf ratio suggest that it was resulting from a crystal accumulation process like a mush zone. The compositional similarity between gray pumice glass chemistry and white pumice whole-rock chemistry (Fig. 1) suggests that interstitial melt in the mush zone was extracted to generate a crystal-poor magma chamber at a shallower level that eventually erupted to form the white pumice. Slight compositional variations of white pumice from Units A to C suggest that conditions like depth, temperature, and water contents of each melt extraction and/or magma chamber may have differed when each unit erupted. Also, our newly obtained 14C age for Unit C (48±1 ka) is hundreds of years younger than that of Unit A (52±3 ka), indicating episodic and distinct extractions and eruptions of silicic magmas of Units A to C. In addition, each unit shows significantly different paleomagnetic directions beyond the 95% confidence limits (Unit A: Dm=4.6°, Im=-14.0°, α95=6.5°; Unit B: Dm=352.1°, Im=-15.0°, α95=1.0°; Unit C: Dm=359.5°, Im=4.9°, α95=3.8°) that strongly support remarkable time gaps between the three units.
Reference
Hasegawa et al. (2018) J.Geogr.(Chigaku Zasshi), 127, 273-288; Wilson et al. (2021) Nat. Rev. Earth Environ., 2, 610–627; Pitcher et al. (2021) J. Petrol., 62, 1–30; Caron et al. (2023) Sci. Rep., 13, 11575; Suhendro et al. (2022) JVGR, 431, 107643; Alloway et al. (2004) EPSL, 227, 121–133.