13:45 〜 15:15
[SCG57-P02] Magma formation and water excess beneath Naruko caldera, NE Honshu
キーワード：Caldera, Rhyolite, Melt inclusion, Water, SIMS, Rejuvenation
Naruko volcano is an active volcano in NE Japan that caused two caldera-forming eruptions at 72 ka (Nizaka episode) and 45 ka (Yanagisawa episode). Here we present a detailed petrological and melt inclusion study to constrain the pre-eruptive magma storage conditions and volatile content from both eruption products.
Naruko rhyolite magmas has been stored at temperatures of 777-832 C and 790-838 C for the Nizaka and Yanagisawa episodes, respectively. Both magmas have been stored at a depth of 4.5-7.7 km under the pressure of 116-197 MPa in H2O-saturated conditions. Despite similar whole-rock composition and P-T-H2O conditions, mineral assemblages of both eruptions have completely reverse features. The textural and compositional features of the orthopyroxene and plagioclase phenocrysts in both eruptions strongly suggest various magmatic conditions and repetitive intruding processes. Low-Al Mg-Hornblende occurs as a phenocryst in Yanagisawa tuffs but is absent in Nizaka tuff.
The orthopyroxene and plagioclase phenocrysts of the Nizaka eruption have normal zoning with high-Mg and high-Ca cores, respectively, which indicates that the phenocrysts were in disequilibrium at the early stages of magma evolution. The magma of the first Nizaka eruption was formed during the partial melting of crustal (probably mushy) rocks due to strong heating by underplating primitive magmas at a temperature of ~1000 C. This process resulted in crystallization of high-Ca patchy cores in plagioclase and mantle high-Mg zones in orthopyroxene. After remobilization, newly formed Nizaka magmas re-equilibrated with the original rhyolites, resulting in the crystallization of Low-Ca and Low-Mg broad rims in phenocryst. Re-heating of the magma led to amphibole breakdown during the Nizaka magma generation, which served as a source of halogens in the melt.
The generation of Yanagisawa magma probably took place in the same magmatic system due to the gradual stable heating of residual partly solidified Nizaka magmas. This process was not accompanied by magma mixing, but only heat and volatiles were injected into the original Nizaka mush to form the Yanagisawa magmas. Mineral zoning in Yanagisawa phenocrysts reflects the multiple injections of hotter magma below rhyolites that lead to thermal convection and remobilization of the resident magma. These conditions led to the formation of an association of phenocrysts with broad inverse zoning and did not induce an eruption long enough to generate 10 km3 of chemically homogeneous magma (given that the entire range between the two eruptions is 27 thousand years). In the Yanagisawa magmas, hornblende again became a stable phase, associated with more stable generation conditions (without sporadically intense heating), a slight enrichment of magmas with MgO and an increase in the oxygen fugacity by 0.5-0.8 log units.
We also considered the theoretical scenarios of the Naruko rhyolite formation through batch crystallization of Naruko andesitic basalt (most primitive samples) to estimate potential water excess during fractional crystallization under shallow crustal conditions. One of the Rhyolite-Melts modeling results suggest that the Naruko rhyolites can be obtained by 70% crystallization of andesitic basalt with initial H2O=4.2 wt%, CO2=0.001 wt% at a pressure of 150 MPa. Based on the Rhyolite-Melts results and mass balance calculation, we estimated the mass of water that could be dissolved in the parental magma in the final Naruko rhyolite and the part that was not dissolved in the final rhyolite and probably was separated as a free fluid. We estimate that the parental melt contained 3.86 Gt bulk water content when the final rhyolite dissolved only 1.36 Gt. Thus, approximately 2.5 Gt of water (65 wt%) has not been dissolved in the rhyolitic endmember but became excessive due to pressure limitation of water dissolution in the melt.
Naruko rhyolite magmas has been stored at temperatures of 777-832 C and 790-838 C for the Nizaka and Yanagisawa episodes, respectively. Both magmas have been stored at a depth of 4.5-7.7 km under the pressure of 116-197 MPa in H2O-saturated conditions. Despite similar whole-rock composition and P-T-H2O conditions, mineral assemblages of both eruptions have completely reverse features. The textural and compositional features of the orthopyroxene and plagioclase phenocrysts in both eruptions strongly suggest various magmatic conditions and repetitive intruding processes. Low-Al Mg-Hornblende occurs as a phenocryst in Yanagisawa tuffs but is absent in Nizaka tuff.
The orthopyroxene and plagioclase phenocrysts of the Nizaka eruption have normal zoning with high-Mg and high-Ca cores, respectively, which indicates that the phenocrysts were in disequilibrium at the early stages of magma evolution. The magma of the first Nizaka eruption was formed during the partial melting of crustal (probably mushy) rocks due to strong heating by underplating primitive magmas at a temperature of ~1000 C. This process resulted in crystallization of high-Ca patchy cores in plagioclase and mantle high-Mg zones in orthopyroxene. After remobilization, newly formed Nizaka magmas re-equilibrated with the original rhyolites, resulting in the crystallization of Low-Ca and Low-Mg broad rims in phenocryst. Re-heating of the magma led to amphibole breakdown during the Nizaka magma generation, which served as a source of halogens in the melt.
The generation of Yanagisawa magma probably took place in the same magmatic system due to the gradual stable heating of residual partly solidified Nizaka magmas. This process was not accompanied by magma mixing, but only heat and volatiles were injected into the original Nizaka mush to form the Yanagisawa magmas. Mineral zoning in Yanagisawa phenocrysts reflects the multiple injections of hotter magma below rhyolites that lead to thermal convection and remobilization of the resident magma. These conditions led to the formation of an association of phenocrysts with broad inverse zoning and did not induce an eruption long enough to generate 10 km3 of chemically homogeneous magma (given that the entire range between the two eruptions is 27 thousand years). In the Yanagisawa magmas, hornblende again became a stable phase, associated with more stable generation conditions (without sporadically intense heating), a slight enrichment of magmas with MgO and an increase in the oxygen fugacity by 0.5-0.8 log units.
We also considered the theoretical scenarios of the Naruko rhyolite formation through batch crystallization of Naruko andesitic basalt (most primitive samples) to estimate potential water excess during fractional crystallization under shallow crustal conditions. One of the Rhyolite-Melts modeling results suggest that the Naruko rhyolites can be obtained by 70% crystallization of andesitic basalt with initial H2O=4.2 wt%, CO2=0.001 wt% at a pressure of 150 MPa. Based on the Rhyolite-Melts results and mass balance calculation, we estimated the mass of water that could be dissolved in the parental magma in the final Naruko rhyolite and the part that was not dissolved in the final rhyolite and probably was separated as a free fluid. We estimate that the parental melt contained 3.86 Gt bulk water content when the final rhyolite dissolved only 1.36 Gt. Thus, approximately 2.5 Gt of water (65 wt%) has not been dissolved in the rhyolitic endmember but became excessive due to pressure limitation of water dissolution in the melt.