4:00 PM - 4:15 PM
[SCG51-08] Mantle processes heterogeneity in the Northern and Southern Oman ophiolite: Insights from Independent Component Analysis and Open-System Melting model
Keywords:Oman, Independent Component Analysis, geochemistry, Mantle
Oman ophiolite records the spatial compositional variability of the oceanic lithosphere and mantle. Recent studies suggest that its tectonic setting was a Fore-arc during subduction-initiation like the Izu-Bonin-Mariana (IBM) (Guilmette et al., 2018). Therefore, the Oman ophiolite preserves the geochemical processes (e.g., partial melting, melt-rock reactions) of the upper mantle and their spatial distribution during subduction. Additionally, the heterogeneity of mantle processes between the northern and southern massifs has also been discussed. While the mantle section of the northern massif preserves more hydrous signatures (e.g., fluid-fluxed melting) (Kanke & Takazawa, 2014), the southern massif primarily records anhydrous signatures (Hanghøj et al., 2010). However, it is under debate whether these processes occurred in different tectonic settings or continuously over a ~600 km scale in the ophiolite.
In this study, we quantitatively compared the mantle processes of the northern and southern massifs using Independent Component Analysis (ICA) and an open-system mass balance equation model (OSM) (Ozawa, 2001). Moreover we also distinguished two processes involving fluid and/or melt. ICA reduces high-dimensional data into low-dimensional data, the components of which are statistically independent (Iwamori and Albarède, 2008). The independent components (ICs) extracted by ICA represent independent mantle processes. We analyzed 19 trace element contents of 194 peridotites in the northern (Fizh) and southern (Wadi Tayin) massifs and obtained four ICs (IC1-IC4). We also used the OSM with the following parameters: (1) Partial melting: nearly fractional melting (melt fraction = 0.008-0.1), (2) Flux melting: slab-derived fluid fluxed melting (Ishikawa et al., 2005) with low to high influx rate (= 5x10-5-0.5), (3) MORB-like melts–peridotite reactions: very low to high melts influx rate (= 10-6-0.5).
ICA reveals four independent geochemical processes: (1) IC1: positively correlated with incompatible major and trace elements contents, however, negatively correlated with compatible elements contents, Olivine Fo, Spinel Cr#, (2) IC2: positively correlated with compatible major elements contents and light rare earth elements (LREEs) contents, however, negatively correlated with heavy rare earth elements (HREEs) contents, (3) IC3: only positively correlated with Sr and Ba contents, however, (4) IC4: only strong positive correlations with large ion lithophile elements (LILEs) contents, such as Ba, Cs and Rb. The above observations suggest that IC1–IC4 corresponds to closed partial melting, slab fluid-fluxed melting, serpentinization, and metasomatism, respectively. To evaluate quantitatively igneous processes, we project modeled peridotite compositions onto IC1–IC2 spaces.
The interpretations of ICs for each region are identical, suggesting that mantle processes in the northern and southern massifs are similar. However, slight differences in ICs values were observed: (1) IC1: north < south, (2) IC2: south < north, indicating that the northern mantle section has (1) more slightly partial melted peridotites and (2) more hydrous signatures than those of southern. In the correlations between IC1–2 and the model results, nearly fractional melting (melt fraction=0.008-0.015) in the garnet field (up to 3-17%) can account for the IC1 chemical trends. We can also distinguish two processes involving fluid and/or melt: (1) positive IC2: slab derived fluid-fluxed melting with a low to high influx rate (<0.5), (2) negative IC2 and positive IC1: a reaction between MORB-like melt and peridotite with a low influx rate (<0.1).
In conclusion, there are no significant differences in processes between the regions; however, the northern massif exhibits slightly more hydrous signatures. Additionally, we have specified the parameters of the igneous processes responsible for mantle compositional variations using ICA and OSM.
In this study, we quantitatively compared the mantle processes of the northern and southern massifs using Independent Component Analysis (ICA) and an open-system mass balance equation model (OSM) (Ozawa, 2001). Moreover we also distinguished two processes involving fluid and/or melt. ICA reduces high-dimensional data into low-dimensional data, the components of which are statistically independent (Iwamori and Albarède, 2008). The independent components (ICs) extracted by ICA represent independent mantle processes. We analyzed 19 trace element contents of 194 peridotites in the northern (Fizh) and southern (Wadi Tayin) massifs and obtained four ICs (IC1-IC4). We also used the OSM with the following parameters: (1) Partial melting: nearly fractional melting (melt fraction = 0.008-0.1), (2) Flux melting: slab-derived fluid fluxed melting (Ishikawa et al., 2005) with low to high influx rate (= 5x10-5-0.5), (3) MORB-like melts–peridotite reactions: very low to high melts influx rate (= 10-6-0.5).
ICA reveals four independent geochemical processes: (1) IC1: positively correlated with incompatible major and trace elements contents, however, negatively correlated with compatible elements contents, Olivine Fo, Spinel Cr#, (2) IC2: positively correlated with compatible major elements contents and light rare earth elements (LREEs) contents, however, negatively correlated with heavy rare earth elements (HREEs) contents, (3) IC3: only positively correlated with Sr and Ba contents, however, (4) IC4: only strong positive correlations with large ion lithophile elements (LILEs) contents, such as Ba, Cs and Rb. The above observations suggest that IC1–IC4 corresponds to closed partial melting, slab fluid-fluxed melting, serpentinization, and metasomatism, respectively. To evaluate quantitatively igneous processes, we project modeled peridotite compositions onto IC1–IC2 spaces.
The interpretations of ICs for each region are identical, suggesting that mantle processes in the northern and southern massifs are similar. However, slight differences in ICs values were observed: (1) IC1: north < south, (2) IC2: south < north, indicating that the northern mantle section has (1) more slightly partial melted peridotites and (2) more hydrous signatures than those of southern. In the correlations between IC1–2 and the model results, nearly fractional melting (melt fraction=0.008-0.015) in the garnet field (up to 3-17%) can account for the IC1 chemical trends. We can also distinguish two processes involving fluid and/or melt: (1) positive IC2: slab derived fluid-fluxed melting with a low to high influx rate (<0.5), (2) negative IC2 and positive IC1: a reaction between MORB-like melt and peridotite with a low influx rate (<0.1).
In conclusion, there are no significant differences in processes between the regions; however, the northern massif exhibits slightly more hydrous signatures. Additionally, we have specified the parameters of the igneous processes responsible for mantle compositional variations using ICA and OSM.