Page 179 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

180 Chapter 6

statistical significance of spatial association between point geo-objects of interest and
linear (or point) features. In addition, the results shown in Tables 6-III to 6-VI indicate
that the distance correlation method is sensitive to the position of an AP with respect to
line segments ⊥L i, so that not just one but at least nine APs are recommended in order to
properly characterise the optimum spatial association between P jx and L i. This is
demonstrated further with the analysis of the spatial association between the centroids of
mapped units of Nabongsoran Andesite porphyry and intersections of NNW- and NW-
trending faults/fractures in the case study area.
There are very few (i.e., four) mapped units of Nabongsoran Andesite and they are
situated only in the southwest quadrant of the study area (see Fig. 3-9). The distances
between each centroid and its nearest neighbour intersection of NNW- and NW-trending
faults/fractures vary from 230 to 700 m, with a mean of 460 m and standard deviation of
205 m. The length of each line segment ⊥L i, which passes through each of the centroids
starting from their respective nearest neighbour intersection of NNW- and NW-trending
faults/fractures, is set at 1200 m. The results of the analyses based on each of the nine
recommended APs are illustrated in Fig. 6-15. The north, northeast and southwest APs
resulted in ‘flat’ graphs of r d jx d jy versus Y j (Figs. 6-15B, 6-15C and 6-15G,
respectively), although peak values of r d jx d jy at 800 m can be observed if the
resolutions of the y-axes of these graphs are enhanced. The southeast and west APs
resulted in downward-concave graphs of r d jx d jy versus Y j with peak values of r d jx d jy
at 1000 m and 900 m, respectively (Figs. 6-15E and 6-15H, respectively). These results
are unsatisfactory because the distances Y jmax r obtained are well beyond the maximum
X j (i.e., 700 m). Nevertheless, the results obtained based on the centre, east, south and
northwest APs (Figs. 6-15A, 6-15D, 6-15F and 6-15I, respectively) are satisfactory,
because they clearly show downward-concave graphs of r d jx d jy versus Y j with peak

values of r d jx d jy mostly at 700 m.
The factors for the unsatisfactory results obtained based on the north, northeast,
southwest, southwest and west APs and the satisfactory results obtained based on the
centre, east, south and northwest APs are given in Table 6-VII. Satisfactory results are
obtained when an AP forms intermediate angles (i.e., mostly between 30º and 60º) with
⊥L i and when the distances (d jy) between an AP and every point P jy along ⊥L i have
relatively high coefficients of variation (i.e., ratio of standard deviation to mean). Note
that both of these two factors apply to the north AP, which forms angles with every ⊥L i
similar to that formed by the centre AP although distances (d jy) from the former to points
P jy along line segments ⊥L i have relatively low coefficients of variation compared to
distances from the former. Note further that APs forming, on average, low angles (<20°)
with ⊥L i result in ‘flat’ graphs of r d jx d jy versus Y j, whereas APs forming, on average,

high angles (>63°) with ⊥L i result in over-estimation of Y jmax r.
Based on the results shown in Fig. 6-15 and summarised in Table 6-VIII for the APs
providing the satisfactory results, there is positive spatial association between the four