Relative Error Analysis of Geodetic Networks

Abstract

Positional accuracy estimators for all the points in a primary control network are derived from the covariance matrix. Relative error ellipses in scale and in orientation are proposed as objective descriptors of the positional accuracy of every point with respect to a ring of its neighbors. The ring size is chosen as a function of the average distance between points in the network. A free‐net error ellipse is considered as another alternative, for describing the positional accuracy of a point relative to its neighbors. A few examples of the analysis of a primary horizontal network are given as an illustration.