Circular statistics for the analysis of directional geological data

Abstract

Geological data such as the strike of bedding, paleocurrent direc-ions, paleomagnetic directions, and the trend of mineral lineations are directional data defined by an angle from a reference direction. The main characteristic of such data is circularity in any reference direction (i.e. 0°=360°), which creates problems when analyzing the data statistically. In this regard, only directional statistics are ade-quate for dealing with directional data. This paper introduces the fundamental theory of circular statistics, a form of directional statistics that is restricted to data on a circumference, and presents practical methods with which to analyze circular data, including their graphical representation, descriptive statistics, the characteristics of circular uniform and von Mises distributions, tests of goodness-of-fit, and single- and multi-sample statistical tests. In addition, statistical software is considered with which to analyze circular data.