Morphology and crustal structure of a small transform fault along the Mid-Atlantic Ridge: The Atlantis Fracture Zone

The Atlantis Fracture Zone (30° N) is one of the smallest transform faults along the Mid-Atlantic Ridge with a spatial offset of 70 km and an age offset of ~ 6 Ma. The morphology of the Atlantis Fracture Zone is typical of that of slow-slipping transforms. The transform valley is 15–20 km wide and 2–4 km deep. The locus of strike-slip deformation is confined to a narrow band a few kilometers wide. Terrain created at the outside corners of the transform is characterized by ridges which curve toward the ridge-transform intersections and depressions which resemble nodal basins. Hooked ridges are not observed on the transform side of the ridge-transform intersections. Results of the three-dimensional inversion of the surface magnetic field over our survey area suggest that accretionary processes are sufficiently organized within 3–4 km of the transform fault to produce lineated magnetic anomalies. The magnetization solution further documents a 15-km, westward relocation of the axis of accretion immediately south of the transform about 0.25 Ma ago. The Atlantis Transform is associated with a band of high mantle Bouguer anomalies, suggesting the presence of high densities in the crust and/or mantle along the transform, or anomalously thin crust beneath the transform. Assuming that all the mantle Bouguer anomalies are due to crustal thickness variations, we calculate that the crust may be 2–3 km thinner than a reference 6-km thickness beneath the transform valley, and 2–3 km thicker beneath the mid-points of the spreading segments which bound the transform. Our results indicate that crustal thinning is not uniform along the strike of the fracture zone. Based on studies of the state of compensation of the transform, we conclude that the depth anomaly associated with the fracture zone valley is not compensated everywhere by thin crust. Instead, the regional relationship between bathymetry and gravity is best explained by compensation with an elastic plate with an effective thickness of ~ 4 km or greater. However, the remaining isostatic anomalies indicate that there are large variations away from this simple model which are likely due to variations in crustal thickness and density near the transform.