On the accuracy assessment of input gravity data in local gravity field modeling

The subject of this study is an assessment of the accuracy of gravity data used in local gravity field modeling when the a priori standard deviations of input gravity data are not available. To assess the accuracy of gravity data sets, Variance Component Estimation (VCE) technique for the observation group weighting is applied. The parameterization of gravity field is realized in terms of the spherical radial basis functions (SRBF). The unknown parameters are estimated by a least-squares technique. The performance of VCE technique for the observation group weighting is demonstrated using real data, providing that the parameterization of gravity field is chosen optimally by means of modeling the gravity signal and not the observation data noise.