Degassing of CO2-H2O from Miyakejima Volcano: Insights from Melt Inclusion Compositions and Volcanic Gas Emissions

The H2O and CO2 concentrations in melt inclusions often show a wide variation, indicating complicated alteration processes of their volatile compositions, such as combination of degassing, addition of CO2-rich fluids, and crystallization. In order to understand these complicated processes, we need to compare not only the melt inclusion compositions but also composition of the volcanic gases and degassing history of the volcano. For this purpose, we report the results of melt inclusion and volcanic gas analyses of Miyakejima volcano, and model the degassing conditions of the volcano. We analyzed melt inclusions in Mg-rich and Mg-poor olivine phenocrysts of the 18 August, 2000 eruption of Miyakejima volcano by SIMS and EPMA at the GSJ, AIST. Both melt inclusions have similar CO2 concentrations ranging from nearly zero to about 300ppm, but inclusions in the Mg-rich Ol are more H2O-rich (2-3.5 wt.%) than those in the Mg-poor Ol (1-2.5wt.%). The Mg-poor Ol melt inclusion has the same bulk composition as the magmatic ejecta of the 18 August eruption, and is considered to have been evolved by fractional crystallization of the melt in the Mg-rich Ol. The H2O-poor composition of the Mg-poor inclusions requires that the evolved melt from the Mg-rich inclusions should loose H2O by degassing at low pressure with subsequent addition of CO2. Such CO2-rich and H2O-rich melt inclusions were observed at many other volcanoes, and were often explained by buffering by large amount of CO2-rich fluid (CO2-flushing) supplied from an external source. The recent activity of the Miyakejima volcano is characterized by the intensive degassing started after the 18 August, 2000 eruption with the total emission of 24 Mt SO2 until 2009, that requires degassing of a km3 scale magma chamber. Since the composition of the volcanic gas was almost constant at about H2O/CO2 = 50 and CO2/SO2 = 1 for the nine years, indicating that degassing occurred from a homogeneous magma. If the melt in the Mg-poor Ol is the source of the intensive degassing, the km3 scale magma should have been experienced the CO2 flushing. However, the present degassing is relatively CO2-poor and there have not been any intensive degassing event in the past at Miyakejima. Therefore, the present degassing activity needs to originate from the melt in the Mg-rich Ol. Although the 18 August eruption was the main eruptive event in 2000, the amount of magmatic product was small about 5 Mt (Nakada et al., 2005). Such a relatively small amount of magma, which ascended slowly during the crater floor subsidence taking about several weeks, could have exposed to the CO2-rich gases discharged from the more primitive basalt which ascended following the erupted magma and caused the subsequent intensive degassing.