9:15 AM - 9:30 AM
[SVC44-02] Temporal evolution of proto-Izu-Bonin-Mariana arc volcanism: Constraints from statistical analysis of melt inclusion composition
Keywords:Amami Sankaku Basin, International Ocean Discovery Program, Izu-Bonin-Mariana arc, Melt inclusion, Statistical analysis
Recently, we have extended the dataset of Brandl et al. (2017) by (i) additional analysis of four volatile elements (H2O, S, Cl and F) and non-volatile P2O5 for 55 selected melt inclusions by Secondary Ion Mass Spectrometry (SIMS) at Kochi Institute for Core Sample Research of JAMSTEC and (ii) statistical analysis on the geochemical composition of 236 selected melt inclusions in order to separate them into several, petrologenetically distinct groups. Among methods of statistical analysis, we performed Principal Component Analysis and K-means Cluster Analysis on the 236 melt inclusions to make full use of the geochemical data of the major elements (10 elements) and the volatiles (S and Cl), following the procedures of Iwamori et al. (2017).
Combined with (i) and (ii), the melt inclusions can be grouped into seven clusters termed Clusters 1 to 7, mainly based on characteristics of elements of lower concentration, such as K2O, TiO2, S and Cl (Fig. a-d), indicating seven distinct magmatic activities. The cluster numbers are assigned in the order of the mean values of SiO2; the mean value of SiO2 is the lowest for Cluster 1 and that is the highest for Cluster 7. Cluster 1 melt inclusions (n=63) are medium-K series rocks characterized by relatively high S concentrations (500-3,000 ppm) and define calc-alkaline trends. Cluster 2 melt inclusions (n=64) are medium-K tholeiites charecterized by high TiO2 (>0.8 wt.%). Cluster 3 melt inclusions (n=9) contain extremely high Cl concentrations (up to 1 wt.%). Cluster 4 melt inclusions (n=31) are high-Mg andesites. Melt inclusions assigned to Cluster 6 (n=19) and Cluster 7 (n=5) are silicic. Identification of subgroups of melt inclusions as summarized here cannot be made by conventional graphical approach using two-dimensional diagrams, demonstrating the usefulness of introducing statistical approaches into geochemistry.
The Cluster 5 melt inclusions (high-Mg andesites) emerge from 40 Ma and fades out at 35 Ma, while melt inclusions assigned to Cluster 2 (medium-K tholeiites) and Cluster 4 (low-K tholeiites) emerge from 38 Ma and last until 30 Ma (Fig. e). We interpret that the disappearance of Cluster 5 melt inclusions is linked to the occurence of melt inclusions assigned to Cluster 2 and Cluster 4 which represent normal arc volcanism of tholeiitic magmas. Assuming such a transition in volcanism at age interval of 35 to 38 Ma would be attributed to crustal thickening as the proto-IBM arc evolves by continuous addition of deep basaltic magmas (e.g., Tamura et al., 2016).
Brandl, P.A., Hamada, M., Arculus, R.J., Johnson, K., Marsaglia, K.M., Savov, I.P., Ishizuka, O. and Li, H. (2017) The arc arises: The links between volcanic output, arc evolution and melt composition. Earth Planet. Sci. Lett. 461, 73-84.
Iwamori, H., Yoshida, K., Nakamura, H., Kuwatani, T., Hamada, M., Haraguchi, S. and Ueki, K. (2017) Classification of geochemical data based on multivariate statistical analyses: complementary roles of clustering, principal component and independent component analyses. Geochem. Geophys. Geosyst. 18, doi:10.1002/2016GC006663.
Tamura, Y., Sato, T., Fujiwara, T., Kodaira, S and Nichols, A. (2016) Advent of continents: A new hypothesis. Sci. Rep. 6: 33517, doi: 10.1038/srep33517.