5:15 PM - 6:30 PM
[SVC31-P10] 3D crystal size distributions of pyroxene nanolites plotted against short-axis length reflect magma ascent dynamics
Keywords:crystal size distribution, X-ray CT, nanolite, morphology, magma ascent
Groundmass crystals indicate syn-eruptive magmatic conditions and thus their crystal size distributions (CSDs) reflect magma ascent histories via their crystallizing balance between nucleation and growth (e.g., Marsh 1998; Mujin and Nakamura 2014). 3D CSDs are most commonly estimated from 2D observations and plotted against long-axis length, L (e.g., Higgins 2000). L-plot CSDs of phenocrysts estimated from 2D observations were confirmed to agree well with those measured directly in 3D (e.g., Mock and Jerram 2005; Jerram et al. 2009). However, Castro et al. (2003) pointed out the possibility that largely varied shapes of groundmass crystals cause inconsistency among the CSDs in 2D and 3D.
Because L-plot CSDs require additional stereological corrections compared to CSDs plotted against S or I when they are estimated from 2D observations, the estimated L-plot CSDs should include large errors associated with the stereological corrections. Nevertheless, it has not been discussed fully whether L is the most representative size for CSDs of groundmass crystals in order to infer magma ascent conditions in conduits. Therefore, it is necessary to evaluate the influences of size criteria on the CSDs.
Here, we observed groundmass pyroxene crystals in pumices from sub-Plinian and Vulcanian eruptive phases during the 2011 eruption of Shinmoedake (Kirishima volcano group, Japan) by both a JEOL JSM-7001F field-emission scanning electron microscope (FE-SEM) at Kyoto University and synchrotron radiation-based X-ray computed nanotomography (SR-XCT) at beamline BL47XU of the SPring-8 synchrotron facility (Hyogo, Japan; Uesugi et al. 2006; Takeuchi et al. 2009). The SR-XCT observations with effective spatial resolution of 200 nm provided detailed 3D morphologies, triaxial lengths (S, I, and L), and true 3D CSDs of pyroxene crystals ranging from a few hundred nanometers to a few micrometers in width. We compared 3D CSDs estimated from 2D SEM observations (SEM-CSDs) with those measured directly by SR-XCT (CT-CSDs) to investigate whether the prevalent L-plot SEM-CSDs are accurate for groundmass crystals of various shapes. We also compared 3D CSDs plotted as a function of S, I, and L to explore which plot types are most representative of eruptive style.
The slopes of S- and I-plot SEM-CSDs were similar to the slopes of the CT-CSDs compared to L-plot SEM-CSDs. This inconsistency of L-plot CSDs results from the difficulty in estimating a representative 3D aspect ratio for fairly elongated groundmass crystals (L/S = 1–40) from 2D observations, which is critical for the accuracy of stereological conversion into 3D CSDs. Moreover, the slopes of S- and I-plot CT-CSDs differed more markedly according to eruption style (by ~20%) than those of L-plot CT-CSDs.
For estimating magma ascent histories, we propose that the optimum method for acquiring SEM-CSDs is to measure the cross-sectional widths of crystals and convert the 2D dataset into S-plot CSDs. By calculating S-plot CSD slopes from previous studies reporting L-plot SEM-CSDs and 3D aspect ratios, we show that plotting CSDs against S preserves their actual slopes, which more reliably reflect magma ascent conditions than those plotted against L.
Because L-plot CSDs require additional stereological corrections compared to CSDs plotted against S or I when they are estimated from 2D observations, the estimated L-plot CSDs should include large errors associated with the stereological corrections. Nevertheless, it has not been discussed fully whether L is the most representative size for CSDs of groundmass crystals in order to infer magma ascent conditions in conduits. Therefore, it is necessary to evaluate the influences of size criteria on the CSDs.
Here, we observed groundmass pyroxene crystals in pumices from sub-Plinian and Vulcanian eruptive phases during the 2011 eruption of Shinmoedake (Kirishima volcano group, Japan) by both a JEOL JSM-7001F field-emission scanning electron microscope (FE-SEM) at Kyoto University and synchrotron radiation-based X-ray computed nanotomography (SR-XCT) at beamline BL47XU of the SPring-8 synchrotron facility (Hyogo, Japan; Uesugi et al. 2006; Takeuchi et al. 2009). The SR-XCT observations with effective spatial resolution of 200 nm provided detailed 3D morphologies, triaxial lengths (S, I, and L), and true 3D CSDs of pyroxene crystals ranging from a few hundred nanometers to a few micrometers in width. We compared 3D CSDs estimated from 2D SEM observations (SEM-CSDs) with those measured directly by SR-XCT (CT-CSDs) to investigate whether the prevalent L-plot SEM-CSDs are accurate for groundmass crystals of various shapes. We also compared 3D CSDs plotted as a function of S, I, and L to explore which plot types are most representative of eruptive style.
The slopes of S- and I-plot SEM-CSDs were similar to the slopes of the CT-CSDs compared to L-plot SEM-CSDs. This inconsistency of L-plot CSDs results from the difficulty in estimating a representative 3D aspect ratio for fairly elongated groundmass crystals (L/S = 1–40) from 2D observations, which is critical for the accuracy of stereological conversion into 3D CSDs. Moreover, the slopes of S- and I-plot CT-CSDs differed more markedly according to eruption style (by ~20%) than those of L-plot CT-CSDs.
For estimating magma ascent histories, we propose that the optimum method for acquiring SEM-CSDs is to measure the cross-sectional widths of crystals and convert the 2D dataset into S-plot CSDs. By calculating S-plot CSD slopes from previous studies reporting L-plot SEM-CSDs and 3D aspect ratios, we show that plotting CSDs against S preserves their actual slopes, which more reliably reflect magma ascent conditions than those plotted against L.