Gridding aeromagnetic data using longitudinal and transverse horizontal gradients with the minimum curvature operator

Abstract

The measurement of gradients of the earth's magnetic field from an airborne platform has been advocated for many years. In 1954 Wickerham proposed the use of what is essentially a gradient along the flight direction (longitudinal gradient). Hood's 1965 suggestion was to measure the vertical gradient using a total-field magnetometer mounted in a tail stinger and a second magnetometer in a towed bird below the tail stinger. In fact, the system finally built by Hood's group at the Geological Survey of Canada used two tail stingers separated by 2 m. The advantages of this type of total-field vertical gradiometry were advocated so convincingly that vertical gradiometers was also developed for helicopter surveying. One of the key developments that allowed the collection of high-quality vertical-gradient data from a fixed-wing aircraft was the development of aircraft compensation schemes for gradiometer arrays. Subsequent installations generally measured a vertical gradient and a transverse gradient (perpendicular to the flight direction). The longitudinal gradient is either calculated by taking the spatial derivative of a single sensor traverse, or by averaging two wing tip sensors and subtracting this from a tail sensor. Marcotte et al. (1992) give a number of examples where having both horizontal gradients can help in the interpretation of magnetic data. They suggest that both gradients can be used to (1) determine whether the geology is two-dimensional; (2) estimate the strike direction, and (3) help calculate the vertical gradient. Interestingly, Marcotte et al. demonstrate that the vertical gradient derived from the horizontal gradients gives data with noise levels one tenth of that seen on the measured vertical gradients. Other advantages of measured horizontal gradients are that they can be used: (1) to derive a total field that is free of diurnal effects; (2) to derive the magnetic gradient tensor; (3) to estimate the depth …