Geodynamic and Geochemical Modeling of Mantle Processes along the Southwest Indian Ridge at 35°-40°E: A Hotspot-Mid-Ocean Ridge Interaction Region

Mantle convection can be regarded as the superposition of two convective models: aplate mode and a plume mode. Geodynamic modeling of these regimes has grantedinsight into surface features, and tells us about the mantle processes in a system largelydevoid of observables. Our study of the 35°-40°E segment of the Southwest Indian Ridge(SWIR) seeks to link geochemical and geological observations with the underlying mantleprocesses.Both plate and plume modes interact and combine at the SWIR 35°-40°E segment. Themid-ocean ridge itself is a manifestation of the plate tectonics mode of mantle convection.The slow opening rate and obliquity of this segment should lead to low volcanic activityalong this segment. However, this segment is the point along the SWIR closest to theMarion hotspot, a manifestation of the plume mode of mantle convection. When interactingwith the mid-ocean ridges, hotspots like the Azores, Iceland, Galápagos, and Rodriguezproduce distinctive patterns, such as propagating rifts, triple junctions, and enriched MORBsignatures. The Marion hotspot does not have a similar effect on the SWIR even thoughit is associated with a bathymetric high and residual mantle Bouguer anomaly low. Anotable feature along the ridge is a V-shaped bathymetric anomaly around one of the non-transform discontinuities (NTD).As for the SWIR 10°-16°E area (Montési et al., 2011) geodynamics modeling predictsmagma focusing to highly segmented non-transform oblique segments (NTOS) along theridge. However, geophysical observations show a thinning crust at these regions. Modelingwithout the segmentation along the oblique segments shows much better agreement withthe observations. So either the NTOS are a crustal structure that does not influence mantleupwelling, melt extraction parameters vary along the ridge, or the density of the crust isanomalous in NTOS due to a different fractionation history.We will incorporate whole rock chemistry (including trace element, & REEs) constraintsto the evaluation of our hypotheses by modeling melt evolution and crystallization underdifferent conditions and comparing model predictions with collected samples.