Combination of Local Gravimetry and Magnetic Data to Locate Subsurface Anomalies Using a Matched Filter

Abstract

The detection of mass anomalies in the near subsurface is an important problem in many areas of interest, such as archeology, construction, and hazard analysis. Assuming that the approximate geometry of an anomaly is known, its possible location can be determined by applying a matched filter to observations of gravity, gravity gradient, and magnetic anomalies, as well as to electro-magnetic data. We analyze the specific combination of gravity, gravity gradient, and magnetic data in order to determine their relative strengths and weaknesses in the detection problem. Poisson’s Relation is used to model the magnetic signals generated by the source to be detected, and the mutual covariances of the background geologic noise that may contaminate the observations. Simulations show that the magnetic data can improve the detection using the matched filter, especially with limited gravity gradients from a typical ground gravity gradiometer. Further analyses using actual data over a known local anomaly illustrate the enhancements as well as limitations of the gravimetry, gradiometry, and magnetic data combinations.