Statistical Analysis of Directional Features - Geomathematics: Theoretical Foundations, Applications and Future Developments - page 285

Fig. 8.7 Lineations ( dots ) and s -poles ( circles ) plotted on equal-area Schmidt net, lower hemi-
sphere, for domains A and B in central part of Fig. 8.6 . Averages plotted in Fig. 8.6 are indicated
by triangles and open circles , respectively (Source: Agterberg 1974 , Fig. 111)

Fig. 8.8 San Stefano area. Unit of distance for coordinates of the grid is 10 km. Length of arrow is
proportional to cosine of angle of dip (Maximum arrow length is 1 km). Left side : B-axes and unit
vector means in (100 m 100 m) cells also shown in Fig. 8.6 . Right side : Unit vector means of
B-axes located within circles with 1 km radius (solid arrow heads) and 2 km radius (open arrow
heads) (Source: Agterberg 2012 , Fig. 4)

cos 2

ʸ i is optimal, is not
the Fisher distribution but the so-called Scheidegger-Watson distribution (Watson
1985 ). In Sect. 1.4 it was discussed that this method involves extracting the
dominant eigenvector from a (3 3) matrix with elements determined by the
direction cosines of the observed unit vectors. With respect to shapes of frequency
distributions for unit vectors, it should be kept in mind that they may not be
isotropic. This is illustrated in Fig. 8.7 where it can be seen that the B-axes,

Σ

distribution of the measurements, for which maximizing