A grid interpolation technique for anomaly separation of stream sediments geochemical data based on catchment basin modelling, U-statistics and fractal

Stream sediment geochemical survey is an efficient tool for regional prospecting of ore minerals. However, complex processes of dilution and enrichment made it difficult to be studied by conventional methods. One of the most important methods to separate anomalous areas are fractal and multifractal analysis; however, some techniques of the fractal analysis which are more popular and applicable, such as concentration-area, suffer from lack of a reliable interpolated map of stream sediments data. In this study, we introduce a simple workflow to separate anomalous areas in three steps: 1. a grid interpolation method is developed to produce continues geochemical surface based on geometry of Sample Catchment Basin (SCB), 2. Although U-statistics, a powerful spatial based method, has no permission to work with Stream sediment data, the grid interpolation prepares a convenience (by converting stream sediment data from vector nature into point or pixel-based data) to apply it for this type of geochemical data, 3. Finally, the fractal analysis is applied for anomaly separation. In other words, the transformation into pixel-based nature in the first step is followed by using U-statistics (or any spatial method) to compensate spatial relation among the pixels. Once an accurate interpolated map is computed in the first two steps, fractal analysis is capable to effectively separate different populations (i.e. background and anomaly) with the highest accuracy and validity. This workflow is implemented on 305 stream sediment samples from Saqez area located on western part of Iran to delineate Au mineralization. Results show that anomalous areas of Au are successfully separated in the final categorical map, which is highly matched with the location of 5 high potential zones of Au known in this area.