Inverse Filtering for Gradiometry Data

The convolutional model is considered for a particular type of magnetic gradiometry data. On the basis of a model, its anomaly is considered as the convolution of a shape and an amplitude function. The first one is inverted and forms the filter, which is next convolved with the data to recover the distribution of the subsurface magnetization. Tests on synthetic and real data explore the applicability and efficincy of the scheme.