Paleostress analysis using the Hough transform for separating stresses from heterogeneous fault-slip data

This paper shows that the Hough transform, a basic technique of image processing, is useful for separating stresses from heterogeneous fault-slip data. The present method enumerates admissible stresses by evaluating their probability for a heterogeneous dataset. Using the results of the method applied to artificial data with known responsible stresses, it is argued that the criterion that has been generally used to judge the resolution and accuracy of a numerical technique for the separation is not appropriate. The result reveals the limitation of stress inversion based on the Wallace–Bott hypothesis. No matter how good an inverse method is, it sometimes inevitably yields unexpected stresses even from artificial data if the data have heterogeneity.