Approximation of average course of measured subsidences of mining area by smooth splines

Abstract

The article presents the way of determining an average course of subsidences of mining area measured after the termination of the subsequent mining stages run in the south of Poland by "Chwałowice" coalmine. The mean square approximation was conducted via smooth spline with the use of R - project programme. The problem of ambiguity of parameter value of smoothing of approximation function spar (in case when the function illustrating dependence among parameter values of smoothing and values obtained by generalised cross validation reaches its minimum value in a certain range of its values) was solved with the use of mean square minimalisation criterion of loss function between approximated and forecast subsidence values of Bialek's formulas and the Budryk-Knothe theory. The efficiency of both forecasts was verified as well as their accuracy was estimated based on average course of measured subsidences.