Melting and Mantle Flow at Oblique Ultraslow-Spreading Ridges

Ultraslow-spreading ridges are a unique endmember of the mid-ocean ridge spreading system. Ultraslow-spreading ridges lack transform faults and often form segments oriented at an oblique angle to the overall spreading direction. These oblique segments are characterized by anomalously thin crust compared to predictions from 2-D models for either passive or buoyant mantle flow based on the full spreading rate, suggesting that the rate of mantle upwelling is inhibited by the oblique ridge geometry. In this study, we model mantle flow and thermal structure at oblique ultraslow-spreading ridges using a three-dimensional finite element model. The model geometry consists of an oblique segment bounded by two orthogonal segments and we account for the effects of passive flow, a simple melting law, and melt migration along the 1200 oC isotherm surface. The models predict that temperature, upwelling rate, and crustal thickness along an oblique spreading ridge are a function of only the portion of the spreading rate perpendicular to the ridge axis. Comparing our results to the 9-14oE oblique supersegment of the southwest Indian ridge (SWIR), we find that crustal thickness along a segment oriented 60o relative to the spreading direction is predicted to be ∼ 2.5 km lower than for an orthogonal segment. Including the effects of melt migration towards the bounding orthogonal segment yields a further reduction in crustal thickness with a maximum difference of ∼ 2.75 km between the center of the oblique segment and the bounding orthogonal segments. Intriguingly, the melt migration model also predicts crustal thickness anomalies of ∼ 0.25 km at either end of the orthogonal segment. These crustal thickness anomalies coincide with the locations of the magmatic Narrowgate segment and Joseph Mayes seamount at ends of the oblique supersegment of the SWIR. However, the amplitude of the predicted crustal thickness anomalies is significantly smaller than is observed at the SWIR. Further work is needed to determine whether including the effects of buoyancy, temperature-dependent viscosity, and along-axis flow may enhance the magnitude of these segment bounding crustal thickness anomalies to the proportions observed along the SWIR.