Edge detection in potential filed using the correlation coefficients between the average and standard deviation of vertical derivatives

Abstract

Edge detection is an essential task in the interpretation of potential field data. Many existing edge detection methods are the functions composed by the horizontal derivative and vertical derivative of potential field data. The edges recognized by those methods are bigger than the true edges and when the depths of the geological bodies are different, the edges will be detected fuzzy and difficult to identify the edges accurately. In this paper, we present a new method to delineate the edges of the sources, which is based on the correlation coefficients between the average and standard deviation of vertical derivatives in a sliding window. The algorithm of this method is very simple. The zero value of the correlation coefficients is used to delineate the geological edges. It's clear to see the edges with the zero contours which will lead to continuous and clear detected edges and the method shows the edges more precisely. Moreover, the anomalies in different depths can be detected in the same degree and are insensitive to noise. The measure is initially applied on the synthetic gravity and magnetic data. The tests of theoretical models indicate that the new method could detect the edges of models in different depths and the position is in good accordance with the models' edges. Finally, this method is applied to gravity data from a portion of Vientiane Basin, Laos. As a result, the method is helpful in recognizing geologic fractures clearly. Moreover, it shows more geologic details with smaller window size and gives superior results with higher quality data.