CRITERION FOR COMPARING THE VARIANCE COMPONENT ESTIMATION METHODS USED FOR ADJUSTING HORIZONTAL CONTROL NETWORKS
Determination of which stochastic model taken in hand for this study is suitable under conditions, when comparing stochastic models, used for adjusting horizontal control networks, is the aim of this study. Some well-known variance component estimation methods like Conventional, Helmert, MTNQUE, AUE, and Förstner, which have been developed to determine the stochastic model, necessary to be formed in a real way for adjusting geodetic nets, have been compared. For comparing the models mentioned above, concrete deciding criteria, using statistical tests, have been defined and the determination of which model is superior has been studied. For comparison of the models, numerical experiment using data, which belong to the part of Istanbul Metropolitan Triangulation Network (Asiatic side of Istanbul), has been performed.