Laserfiche WebLink
limitations caused by their sensitivity to the selection of <br />• appropriate parameters for use in the analyses. Although on a <br />case-by-case basis they have been manipulated to provide a <br />reasonable fit to empirical data, difficulty associated with the <br />selection of para meters in the absence of calibration data, have <br />given them limited value as a predictive tool. Similarly, the <br />finite element and distince element methods, have merit only in <br />conditions where a large volume of monitoring data is available <br />for calibration of the models. This can then provide a fine- <br />tuned model for future predictive use in the same mine. In the <br />absence of such data, however, the accuracy of predictions would <br />be no better than other methods making the extreme amount of time <br />and effort required to set up such models unnecessary and <br />• impractical. <br />For the purposes of this study, highly accurate predictions of <br />the final subsided profile were not considered to be necessary. <br />Rather, we wished to employ a method which was relatively easy to <br />use and would allow us to accurately compare the relationship <br />between alternative mine plans, methods, etc. Therefore, we <br />selected the use of influence functions for subsidence <br />prediction. Influence functions would also appear to be among <br />the most widely utilized subsidence predictive methods at the <br />present ti me. <br />We did wish, however, to explore the sensitivity of the results <br />• to the specific function/method utilized. Two computer codes <br />15 <br />