A report on some magnetic modelling experiments for image processing applications

Abstract

Interpretation of aeromagnetic data is greatly aided by the use of synthetic images generated from three-dimensional models. A study of relatively simple models, varying parameters such as source geometry, height above the body, and magnetic field values, highlights some well-known characteristics of magnetic anomalies. The study also clearly demonstrates that resolution of anomaly sources and their shapes is difficult except for elevations less than the source dimensions. A complex geological model was constructed, comprising a set of ring-structures (the host geology) and a set of random prisms (the target ore-bodies). The magnetic field was calculated at locations typical of modern aeromagnetic surveys and the resulting data sets interpolated using a variety of algorithms. These included methods based on distance-weighting, spline interpolation, minimum curvature and triangulation techniques. The gridded data sets were then subjected to downward continuation and latitude reduction procedures. The resulting images demonstrate typical gridding problems, including: inadequate representation of the anomalies between flight lines; introduction of trends and features not present in the original data; and, distortion of the shapes of anomalies. Thus, where geophysical data are to be presented as enhanced images, care is needed with the choice of gridding technique. Furthermore, survey specifications currently employed by the industry, may need to be revised.