New oceanic crust is being formed along active segments of the global midocean ridge (MOR) system. The presence of an axial magma chamber and associated zones of partial melt and hydrothermal activity located up to several kilometres beneath the seafloor is central to almost all recently proposed theories of crustal formation. Seismic images of the top few kilometres beneath the fast spreading East Pacific Rise (EPR) have already been obtained. Reflection profiles place strong constraints on the geometry of the axial magma chamber but refraction data provide only coarse estimates of the subsurface temperature, distribution of partial melt and porosity, parameters required to distinguish between the proposed petrological models. Electrical conductivity is directly related to all these critical parameters and therefore electromagnetic experiments should be designed to help characterize the ridge environment. We determine the magnetic field B(t) on the seafloor caused by a sudden change in current in a 2D electric dipole aligned perpendicular to the strike of the ridge. The finite element technique is used to solve the governing differential equation numerically in the Laplace sdomain. The transformation to the time domain is by the GaverStehfest method. We show that the extraction of two basic parameters from the response curve (t) can provide sufficient information to identify the more important features of the petrology. The parameters are the response amplitude max, which is the maximum derivative of tB(t), and the diffusion time T, the time at which this maximum occurs. The behaviour of as a function of distance from the source is analogous to that of first arrival time in refraction seismology. The value of is a weighted integral of the conductivity along the most resistive path between the source and the receiver. A highly conductive, partially molten magma chamber beneath the ridge axis slows the rate of diffusion of electromagnetic fields across the ridge, increasing but also reducing max at sites on the side of the ridge opposite the transmitter. A melt lens ponding as a thin layer on top of the chamber increases max at the ridge crest and increases at sites on the far side. Hydrothermal fluid circulation in the uppermost 2 km of the crust reduces max everywhere across the ridge but increases only at sites within 03 km of the ridge crest. Electromagnetic energy in this case can reach the more distant points via paths which bypass the fluids. Inferences made from the results of 2D modelling indicate that a practical experiment would require a 104 A m horizontal electric dipole (HED) transmitter located 5 km offaxis and receivers with a sensitivity of at least 1 pT s1 over a time window up to 10 s. Copyright 1991, Wiley Blackwell. All rights reserved
