Improved curvature gravity gradient tensor with principal component analysis and its application in edge detection of gravity data

Edge detection is one of the common processes in the interpretation of gravity data, and it gives a view of the earth which highlights boundaries of geological bodies. Many mathematical techniques have been proposed for edge detection including the curvature gravity gradient tensor method (CGGT). Eigenvalues of the curvature gravity gradient tensor can well delineate the edges of some geological bodies. However, it cannot be applied to complex gravity data with both positive and negative anomalies. Moreover, the CGGT method is also very sensitive to noise. In view of limitations of the method, we proposed an improved CGGT method by incorporating the principal component analysis (PCA) into the CGGT formula. The improved method can be utilized to outline the edges of causative bodies in more general cases and is insensitive to noise. The method was tested on synthetic gravity data and actual gravity data recorded from a metallic mineral deposit area in the middle-lower reaches of the Yangtze River in China. All of the results have shown effectiveness of the proposed method.