Decay of Marine Magnetic Anomalies by Ferrous Ion Diffusion

Abstract

THE linear marine magnetic anomalies are well understood in terms of the hypothesis of Vine and Matthews¹ of normal and reversely magnetized dykes on the seafloor. But it is curious that their amplitudes decay as a function of lateral distance from the middle of the mid-oceanic ridge. This requires the central anomaly to be due to a dyke whose intensity of magnetization, M, has sometimes to be taken as twice that of the dykes on either side². Matthews and Bath³ suggested that the dykes causing the central anomaly are not injected exactly at the centre of the ridge, but with a finite standard deviation. Normally magnetized dykes 200 m wide with a standard deviation of 5 km were thus shown to contaminate the reversely magnetized flanking dykes, causing the variation in anomalies as a function of lateral distance. But Harrison⁴ has shown that such large deviations are unacceptable because they would obliterate the sharp magnetic events, such as the Olduvai event. Harrison concluded that the explanation has to be in terms of a stronger dipole magnetic moment of the Earth in recent times, as against the past and/or a recent secular decay of the frozen natural remanent magnetization (NRM) of the dykes. As supporting proof of the first hypothesis, Harrison quoted palaeointensity work^5–7 on rocks and archaeological samples. But, as Stacey⁸ has pointed out, a secular intrinsic decay of NRM in these samples would appear to lend weight to the hypothesis that the Earth's magnetic field was stronger in the past than now. Harrison's alternative suggestion was a secular decay of NRM because of the demagnetizing action of the past reversals of the Earth's magnetic field. The effect would be somewhat similar to that of demagnetization by a slowly alternating weak magnetic field. But the efficiency of such a demagnetization process would be no different from that of the viscous magnetization acquired at room temperature in the Earth's field, and that has been shown to be unimportant by Nagata⁹.