Crustal magnetization and accretion at the Southwest Indian Ridge near the Atlantis II fracture zone, 0–25 Ma

Abstract

We analyze geophysical data that extend from 0 to 25‐Myr‐old seafloor on both flanks of the Southwest Indian Ridge (SWIR). Lineated marine magnetic anomalies are consistent and identifiable within the study area, even over seafloor lacking a basaltic upper crust. The full spreading rate of 14 km/Myr has remained nearly constant since at least 20 Ma, but crustal accretion has been highly asymmetric, with half rates of 8.5 and 5.5 km/Myr on the Antarctic and African flanks, respectively. This asymmetry may be unique to a ∼400 km wide corridor between large‐offset fracture zones of the SWIR. In contrast to the Mid‐Atlantic Ridge, crustal magnetization amplitudes correlate directly with seafloor topography along the present‐day rift valleys. This pattern appears to be primarily a function of along‐axis variations in crustal thickness, rather than magnetic mineralogy. Off‐axis, magnetization amplitudes at paleo‐segment ends are more positive than at paleo‐segment midpoints, suggesting the presence of an induced component of magnetization within the lower crust or serpentinized upper mantle. Alteration of the magnetic source layer at paleo‐segment midpoints reduces magnetization amplitudes by 70–80% within 20 Myr of accretion. Magnetic and Ocean Drilling Program (ODP) Hole 735B data suggest that the lower crust cooled quickly enough to lock in a primary thermoremanent magnetization that is in phase with that of the overlying upper crust. Thus magnetic polarity boundaries within the intrusive lower crust may be steeper than envisioned in prior models of ocean crustal magnetization. As the crust ages, the lower crust becomes increasingly important in preserving marine magnetic stripes.