This report investigates the viability of using Gaussian Ball Filters as described by Aldridge et al. to account for variance in cavity radii with charge size in the Sharpe Hollow Cavity Model. The dominant frequency of these filters were observed to decrease with increasing width of the Gaussian Ball; if it is assumed that this width corresponds to the charge size then the results are similar to Sharpe's predictions. A cubic relationship between cavity radius and dominant frequency of the resulting spectra was estimated by Petten et al. in Sharpe's model. The Gaussian Ball model however appeared to display a square relationship between the dominant frequency and the width of the Gaussian filter. Several different wavelets were used in conjunction with the Gaussian Ball model to observe the effects of different wavelets. The results of this test showed that both minimum phase and Ricker wavelets produced frequency spectra that were similar in form to those produced by Sharpe's model as well as the Priddis and Hussar data examined by Petten et al. in 2012.
View full article as PDF (0.42 Mb)