Selection of Optimum Fractal Model for Detection of Stream Sediments Anomalies

The main purpose of this research is a comparative study among four different fractal models including Concentration-Perimeter/Area (C-P/A), Concentration-Number (C-N), Concentration-Area (C-A) and Concentration-Perimeter (C-P) for delineation of stream sediments Au anomalies based on catchment basins in Aghkand region, NW Iran. In this study, a total of 920 stream sediment samples were utilized to determine the geochemical anomalies of Au using the fractal models for selection of optimum model. As a result, the Au anomalies were correlated with geological units located in the western and SW parts of the region that mainly consist of andesite rocks and tuffs. To certify this, 78 litho-geochemical sets of data were utilized to validate the C–P/A, C–N, C–A and C–P fractal models for Au by logratio matrix. The overall accuracy rates are 0.97, 0.96, 0.95 and 0.95 for the C– P/A, C–N, C–A, C–P fractal models, respectively. It showed that the C–P/A model was the optimum fractal model in the study region.