Laserfiche WebLink
Raw Data Background <br />Filtered Data <br />Figure 6. An illustration of the rolling ball algorithm for masldng the effect of terrain on conductivity <br />measurements (from (Bashed, 2016)) <br />The rolling ball algorithm determines a local background value for each element of the gridded data. <br />The data are processed as if they are terrain heights, and the background value for each element is <br />determined by averaging a very large ball moving over the whole data grid. This terrain mask, or <br />background, value is then subtracted from the raw data, leaving a filtered, or residual, dataset. <br />Geophysical anomalies, in the form of conductivity values lower than the background, will appear as <br />negative departures from this background in the residual dataset. <br />Results and Discussion <br />The FDEM results have been overlain atop satellite and aerial imagery and are presented in a series <br />of two figures, Figures 7 and 8, attached to the end of this report. The figures are 11 x 17 inches in <br />landscape format. Figure 7 presents the corrected quadrature component, as electrical conductivity <br />in milliSiemens per meter (mS/m), before filtering for the terrain background removal. Low <br />conductivity values are in cool colors (blues), while high conductivity values are in warm colors <br />(reds). Figure 8 presents the residual anomaly from the conductivity mapping, with terrain <br />background removed. The color scale on Figure 8 is qualitative, with red symbolizing readings <br />equivalent or near -equivalent to the terrain mask, or background, and blue values symbolizing low <br />conductivity anomalies filtered from the background. The FDEM control dataset acquired at the <br />nearby property is shown on Figure 9, as both a corrected conductivity dataset, and a residual dataset <br />from the terrain filtering. The color scale for terrain conductivity at the control site is different than <br />at the primary site due to a generally lower conductivity at the site. In addition, the overall variation <br />due to terrain is of a much lower magnitude. Rolling ball filtering was applied at the control siteusing <br />the same gridding and filtering parameters at theas at the primary investigation site. <br />Sinkole/Subsidence feature locations surveyed in previous investigations, provided to Olson by <br />Fontanari, are annotated on the figures. Low conductivity anomalies interpreted from both the <br />conductivity data, and from the residual filtered data, are coincident with these features observed on <br />the surface. Additional low conductivity anomalies, highlighted in particular by the filtering method <br />applied, may be interpreted along a similar orientation to the observed surface anomalies, and are <br />indicated on Figure 8 by dashed black lines. Linear anomlales which trend North-South parallel to <br />the survey direction are also observed in the residual data. These anomalies are interpreted to be <br />artifacts introduced in the data gridding process (see Figure 7), and highlighted by the rolling ball <br />filtering (Figure 8). <br />Project No. 6422A 6 Geophysical Report <br />