Ambiguity in depth resolution in potential field inversion

Inverse potential field problems have ambiguous solutions, and a unique solution can only
be computed if additional a priori information is incorporated into the solution process. We
discuss several kinds of ambiguity, and show how each type of ambiguity occurs and is handled
differently. We demonstrate that potential-field data provide depth information when used in
connection with certain discretizations of the problem, and that the role of the regularization
term in the Tikhonov formulation is primarily to filter out noise. In addition we show how the
inspection of two graphical tools, based on the singular value decomposition, can guide the
regularization and also reveal how much depth information can be achieved for a given noisy
problem.