A method for inverting magnetic anomalies in the presence of topography (Parker and Huestis, 1974) is used to analyze and describe crustal magnetization along 12 profiles in the South Atlantic. The profiles are approximately 2000 km long and are almost perpendicular to the trend of the magnetic anomalies. A bandpass filter is used to remove long-wavelength anomalies unassociated with seafloor-spreading processes. A filter that passes only wavelengths between 200 and 3 km produces a magnetization that approximates the positive and negative aspects of the geomagnetic reversal sequence, the positive aspects originating from normally magnetized portions of the crustal source and negative aspects originating from reversely magnetized portions of the crustal source and negative aspects originating from reversely magnetized portions of the crustal source. A 6-km thick crust is assumed in order to produce magnetizations consistent with dredged oceanic rocks. A long-wavelength pattern is evident in the magnetizations solutions.

The highest magnetizations are associated with the ridge crest (1.35--1.74 A/m). Within the western section of the study area, a sharp decrease in magnetization occurs within 100 km of the ridge crest (4.5 Ma). A minimum in magnetization is reached at 400 km from the axis or at 18 Ma (0.51--0.57 A/m). The magnetization then gradually increases out to 1100 km from the ridge axis or 50 Ma (0.91--0.93 A/m) and dramatically decreases to 1900 km or 86 Ma (0.34--0.43 A/m). Within the eastern section of the study area, a sharp decrease in magnetization occurs withing the first 100 km of the ridge crest (5 Ma). A minimum in the magnetization occurs at approximately 400 km from the ridge axis or at 20 Ma (0.54--0.56 A/m). The magnetization then gradually increases out to 1200 km from the ridge axis or to 60 Ma (1.15--1.24 A/m). An overall pronounced decrease occurs to 1900 km from the ridge axis or to 95 Ma (0.36--0.40 A/m). In order to explain this behavior, the field intensity may have changed with a time scale of several millions of years. Alternatively, the higher magnetizations at 50--60 Ma could be explained by a uniform intensity of magnetization in a thicker crust. ¿ American Geophysical Union 1989