
Minttu Uunila Baseline dependent weights in VieVS M. Uunila, Aalto University Mets"ahovi Radio Observatory J. Gipson, NVI Inc., NASA Goddard Space Flight Center H. KrŽasnŽa, Vienna University of Technology It is well known that in processing VLBI data $chi^{2}$ is usually larger than 1, typically in the range of 48. This results from either too small measurement errors or of mismodeling the data. By reweighting the data, by increasing the errors of the observation, we can make $chi^{2}$ $sim$1. In Solves operational solutions baseline dependent weights are always applied. VieVS uses a constant weight, i.e., global weighting. In order to use baseline dependent weights in VieVS we run the leastsquares adjustment a second time after calculating the reweights for each baseline in an observation. Baseline weighting reduces UT1 adjustment significantly. Discrepancy between VieVS and Solve is also reduced. 