Abstract:
A geomagnetic variation study on mature oceanic lithosphere in the North Atlantic just south of Bermuda has
revealed the presence of at least one layer of low electrical resistivity. The low resistivity layer of approximately 10
ohm-m has been found at three widely spaced seafloor sites with crustal ages of 85, 110 and 150 million years. There
is also evidence that the resistivity increases to greater than 20 ohm-m below about 100 km.
Apparent resistivity and phase versus period are calculated using the vertical gradient of the horizontal magnetic
field variations to estimate the seafloor electric field. The vertical gradient method assumes that the seasurface
magnetic variations can be estimated from a nearby land station and that no local magnetic induction occurs at either
reference or seafloor site. Both assumptions are critically evaluated during the analysis.
Seafloor observations are modeled using the Monte Carlo technique. Estimates of the smoothed resistivity structure as
well as the resolution and precision of the estimates are made using the Backus-Gilbert method. Models are shown to be
severely data limited. Resolution is found to be poor in the upper 30-40 km of the lithosphere due to the lack of reliable
data at periods shorter than 30 minutes. The uncertainty involved in estimating the magnetic field at the seasurface
and the large error estimates combine to give low overall precision. The diurnal results do not agree with the continuum
results if the continuum is corrected for latitudinal variations of the source field between the reference station
and seafloor sites. Data at periods as short as 10 minutes are required to resolve structures in the upper 30 km of the mantle. Artificial source fields may be necessary to obtain periods short enough to resolve crustal features. Periods longer
than diurnal will be required to study sub-lithospheric resistivity variations.
