10:00 AM - 10:15 AM
[SVC28-05] Decoding pre-eruptive process at mushy magma reservoir: case studies of Unzen 1991-95,1792 and 1663 eruptions
Keywords:mushy magma reservoir, crystal clots, amphiboles, pre eruptive process
1. Introduction
One of the causes of volcanic eruption is magma mixing at magma reservoir beneath volcano. For example, intrusion of hot magma into the magma reservoir can be the trigger of eruption. Recently, some studies have shown that the condition of magma reservoir, for example, composition and H2O contents of magma control the eruption style (e.g., Popa et al., 2021). Therefore, it is important to reveal the condition of magma reservoir and magmas involved in the magma mixing.
In this study, we focused on pre-eruptive process of Unzen eruptions at historical times. These eruptions are thought to experience magma mixing before eruption(e.g., Sato et al., 2017). However, the temperatures, compositions of erupted products are different, and some studies (e.g., Saito et al., 2008) showed that these differences determined whether erupted lava could flow the flank or make lava lobes and domes. Therefore, pre-eruptive process that affect the conditions of mixed magma would play the important role to determine the eruption styles.
2. Samples and analytical methods
We use lava blocks of Unzen 1991-95, 1792 and 1663 eruptions as samples. At the 1991-95 eruption, erupted lava formed many lava domes and they repeatedly collapse and flow as pyroclastic flows. On the other hand, At both 1792 and 1663 eruptions, erupted magma flow as lava flow. Crystal clots in products of each eruption are mainly composed of plagioclases, amphiboles, and Fe-Ti oxides. Some crystal clots contain biotite, quartz and pyroxenes. Between crystals, there are interstitial melts. We analyzed phenocrysts (amphiboles, plagioclases, pyroxenes, Fe-Ti oxides) and crystals (plagioclases, amphiboles) and interstitial melts of crystal clots.
Our main analyses were conducted by EPMA at GSJ (Geological Survey of Japan) and ERI (Earthquake and Research Institute, The University of Tokyo), and some trace elements are analyzed LA-ICP-MS, equipped in GSJ. In addition, we analyze the bulk chemistry of 1663 mafic enclave by XRF at GSJ.
3. Results and Discussions
Our results & discussion parts are divided into three main topics. First, we talk about endmember magmas involved in each eruption. Then, we look magma reservoir by crystal clots and reveal the detailed picture of magma reservoir. Finally, we propose a model of pre-eruptive process pf each eruption and compare each process to see what causes the difference of eruption styles.
For 1991-95 eruption and 1792 eruption, we estimate the composition and temperature of the magmas involved in the pre-eruptive magma mixing by using amphibole phenocrysts. On the other hand, for 1663 eruption, because amphiboles are almost totally broken, we use Fe-Ti oxides and pyroxenes for temperature estimation. As a result, form the chemical composition and/or temperature of the melts from which the crystals grew, we reveal the end member magmas of each eruption.
Moreover, our textual and compositional analysis of crystal clots show the detailed picture of magma reservoirs. These crystal clots contain crystals that have grown at several circumstances. This indicate that several different conditions by former intrusions exist in the magma reservoir. In addition, from the diffusion analysis using trace elements of plagioclases, we reveal the differentiation processes at main magma reservoir.
By summarizing these results above, we construct the view of pre-eruptive processes of each eruption. And show that how the conditions of magma reservoirs and magmas involved in the magma mixing affect the magma mixing processes and ascend of mixed magma.
One of the causes of volcanic eruption is magma mixing at magma reservoir beneath volcano. For example, intrusion of hot magma into the magma reservoir can be the trigger of eruption. Recently, some studies have shown that the condition of magma reservoir, for example, composition and H2O contents of magma control the eruption style (e.g., Popa et al., 2021). Therefore, it is important to reveal the condition of magma reservoir and magmas involved in the magma mixing.
In this study, we focused on pre-eruptive process of Unzen eruptions at historical times. These eruptions are thought to experience magma mixing before eruption(e.g., Sato et al., 2017). However, the temperatures, compositions of erupted products are different, and some studies (e.g., Saito et al., 2008) showed that these differences determined whether erupted lava could flow the flank or make lava lobes and domes. Therefore, pre-eruptive process that affect the conditions of mixed magma would play the important role to determine the eruption styles.
2. Samples and analytical methods
We use lava blocks of Unzen 1991-95, 1792 and 1663 eruptions as samples. At the 1991-95 eruption, erupted lava formed many lava domes and they repeatedly collapse and flow as pyroclastic flows. On the other hand, At both 1792 and 1663 eruptions, erupted magma flow as lava flow. Crystal clots in products of each eruption are mainly composed of plagioclases, amphiboles, and Fe-Ti oxides. Some crystal clots contain biotite, quartz and pyroxenes. Between crystals, there are interstitial melts. We analyzed phenocrysts (amphiboles, plagioclases, pyroxenes, Fe-Ti oxides) and crystals (plagioclases, amphiboles) and interstitial melts of crystal clots.
Our main analyses were conducted by EPMA at GSJ (Geological Survey of Japan) and ERI (Earthquake and Research Institute, The University of Tokyo), and some trace elements are analyzed LA-ICP-MS, equipped in GSJ. In addition, we analyze the bulk chemistry of 1663 mafic enclave by XRF at GSJ.
3. Results and Discussions
Our results & discussion parts are divided into three main topics. First, we talk about endmember magmas involved in each eruption. Then, we look magma reservoir by crystal clots and reveal the detailed picture of magma reservoir. Finally, we propose a model of pre-eruptive process pf each eruption and compare each process to see what causes the difference of eruption styles.
For 1991-95 eruption and 1792 eruption, we estimate the composition and temperature of the magmas involved in the pre-eruptive magma mixing by using amphibole phenocrysts. On the other hand, for 1663 eruption, because amphiboles are almost totally broken, we use Fe-Ti oxides and pyroxenes for temperature estimation. As a result, form the chemical composition and/or temperature of the melts from which the crystals grew, we reveal the end member magmas of each eruption.
Moreover, our textual and compositional analysis of crystal clots show the detailed picture of magma reservoirs. These crystal clots contain crystals that have grown at several circumstances. This indicate that several different conditions by former intrusions exist in the magma reservoir. In addition, from the diffusion analysis using trace elements of plagioclases, we reveal the differentiation processes at main magma reservoir.
By summarizing these results above, we construct the view of pre-eruptive processes of each eruption. And show that how the conditions of magma reservoirs and magmas involved in the magma mixing affect the magma mixing processes and ascend of mixed magma.