The implementation of variance component estimation (vce) using MINQE approach into an adjustment of height networks

Abstract

The adjustment of height network involves a simple procedure which can be easily performed by a linear parametric adjustment. The required input data normally consists of levelling line particulars such as height differences and lines distances. The observational variance is normally introduce by the latter which then determines the weight matrix for the adjustment. Such simple variance estimation has to be reevaluated particularly in cases where the levelling data is gathered from heterogeneous sources such as GPS or from a combination of several levelling instruments. The approach of VCE using Minimum Norm Quadratic Estimation (MINQE) has been implemented in the adjustment of height network in order to examine the formulation of the stochastic model. The example of such implementation is shown using two different levelling networks – a precise levelling and a GPS levelling. The results of this study show that in some cases, the conventional way of determining weight matrix in height network has to be reconsidered.