Optimized Parameterization of VLBI Auxiliary Parameters in Least-Squares Adjustment: Preliminary Results

Abstract

In a general parameter estimation model, a priori information is used to linearize the system of equations being solved so that just offsets to the a priori values need to be estimated. A priori information used in Very Long Baseline Interferometry (VLBI) data analysis is additionally needed for modeling and constraining the auxiliary parameters, i.e. zenith wet delays, clocks, and troposphere gradients, in order to stabilize the parameter estimation. In our study we investigate the modelling of the auxiliary parameters.

In order to improve the currently used parameterization in the VLBI software VieVS used by many International VLBI Service for Geodesy and Astrometry – IVS Analysis Centers we consider three different approaches. In the first one the choice of the length of the time interval is made relatively to the temporal behavior of parameters (approach1), in the second one we do not estimate parameters in gaps in the observation sessions (approach2), and in the third one the time interval is choosen relatively to the number of observations (approach3). The preliminary results show that approach2 and approach3 provide results better than the standard approach (currently used parameterization in VieVS software) for VLBI single session analysis with the least squares solution of the Vienna VLBI Software (VieVS). The impacts of approach2 and approach3 on various VLBI solutions with VieVS were assessed by descriptive statistics and remarks for future studies are given. The optimization depending on approach1 will be investigated in future.