Assessment of observations using minimum norm quadratic unbiased estimation (minque)

Abstract

Estimation of the variances and covariances of observations in geodetic and engineering surveys of high precision is of great importance for proper weighting of the observations in their subsequent processing and for better understanding of sources of errors and their modeling. The authors have adopted from modern statistics the Minimum Norm Quadratic Unbiased Estimation (MINQUE) method for the variance-covariance component estimation. The method has successfully been implemented in modeling the error propagation in geodetic leveling, in determination of error components and weighting of measurements with the Global Positioning System, and in an assessment of a triangulateration network observed with different EDM instruments.