Page 101 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

100 Chapter 4

uni-element concentrations in the associated sample. A case example of fractal modeling
of geochemical anomalies using discrete surfaces based on stream sediment sample
catchment basins is demonstrated by Shen and Cohen (2005). So, one purpose of this
case study is to compare and contrast results of using continuous and discrete surfaces of
stream sediment uni-element concentrations in the concentration-area fractal method for
separation of background and anomaly. To ensure proper comparison and analysis of the
results, the continuous and discrete geochemical surfaces are classified uniformly using
5-percentile intervals of data distributions. For classification of the discrete geochemical
surfaces, the 5-percentile intervals are determined from the original point data values
rather than from the map (pixel) values in order to respect the empirical density
distributions of the data and because the variations in the sizes of the stream sediment
sample catchment basins (Fig. 4-11) can introduce artificial data distributions and thus
artifacts in the concentration-area relations.

Analysis of uni-element threshold values and anomalies
Based on the continuous geochemical surfaces, the log-log plots of the area-
concentration relations for all elements, except As, can be fitted with more than two
straight lines indicating the presence of at least three populations in the uni-element data
sets (Fig. 4-12). Based on the discrete geochemical surfaces, the log-log plots of the
area-concentration relations for all elements can be fitted with more than two straight
lines indicating the presence of at least three populations in the uni-element data sets
(Fig. 4-13). For each element, the log-log curves of the concentration-area plots based on
the continuous geochemical surfaces and on the discrete geochemical surfaces have very
similar shapes (Figs. 4-12 and 4-13). This suggests that, in this case study, either the
continuous geochemical surfaces or discrete geochemical surfaces can be used in the
concentration-area fractal analysis of geochemical anomalies.
The breaks in slopes of the straight lines fitted to the log-log plots of the
concentration-area relations represent thresholds that can be used to classify the uni-
element data sets into background and anomalous populations. The numbers of
thresholds defined per element based on the continuous and discrete geochemical
surfaces are equal except for Cu, Mn and As. For each of these three elements, the
number of thresholds obtained from the discrete geochemical surfaces is greater than the
number of thresholds obtained from the continuous geochemical surfaces. Nevertheless,
the values of the thresholds defined using the continuous and discrete geochemical
surfaces are closely similar (Table 4-I), especially for Zn, Ni, Mn and As, although for
each of the last two elements a third threshold and a second threshold was defined,
respectively, using the discrete geochemical surfaces. For Cu, the higher threshold based
on the continuous geochemical surface is closely similar to the highest threshold based
on the discrete geochemical surface. In addition, for Cu, the lower threshold based on the
continuous geochemical surface is roughly equivalent to the average of the lowest and
intermediate thresholds based on the discrete geochemical surface. For Co, the two
higher thresholds based on the continuous geochemical surface are respectively