An empirical method of estimation of the variance - Covariance matrix in GPS network design

For a priori analysis of the precision of a GPS network, we propose an empirical method to estimate the variances and covariances of components of baselines. We hypothesize that the variances are proportional to the length of the corresponding baseline components, while the covariances are computed from values of the correlation coefficients. Both the variances and covariances are evaluated in a 3D global Cartesian co-ordinate system. The method is applied to some previously measured GPS networks and the results are compared with those obtained by the methods of zero correlation and the post-processing computations.