Selected Topics for Further Research - Geomathematics: Theoretical Foundations, Applications and Future Developments - page 483

Fig. 12.12 Study area of 64 cells with locations of till samples, Southern Saskatchewan. Unit of
distance for east-west and north-south directions is 78.125 km (Source: Agterberg 2007a , Fig. 9)

(see Table 11.1). Original sampling design for the geochemical reconnaissance
study in Southern Saskatchewan is described by Garrett and Thorleifson ( 1995 ). An
80 80 km grid was selected as the basic sampling cell. It was successively

subdivided into 40 40 km, 20 20 km and finally 10 10 km grid cells for

sampling with randomly selected 1 1 km target cells. Compositional data for

many other elements and minerals were determined as well. Gold and arsenic were
selected for this study because their frequency distributions seem to be unimodal,
and these elements had been used extensively in other multifractal studies (Cheng
1994 ; Agterberg 2001a , b ).
For the case study described here, a modified grid was used for Southern
Saskatchewan resulting in a square-shaped study area consisting of 64 square
cells measuring slightly less than 80 km on a side (see Fig. 12.12 ). The purpose
of this re-definition of boundaries of study area was to minimize edge effects in
multifractal analysis. Data from outside the square study area were not used. In
total, 290 of the original 389 (on average, about 4.5/cell) were retained. None of the
cells inside the study area of Fig. 12.12 are empty. Figure 12.13 is a Q-Q plot of
gold cell average concentration values with best-fitting straight line.