Automated ambiguity estimation for VLBI Intensive sessions using L1-norm

Abstract

Very Long Baseline Interferometry (VLBI) is a space-geodetic technique that is uniquely capable of direct observation of the angle of the Earth's rotation about the Celestial Intermediate Pole (CIP) axis, namely UT1. The daily estimates of the difference between UT1 and Coordinated Universal Time (UTC) provided by the 1-hour long VLBI Intensive sessions are essential in providing timely UT1 estimates for satellite navigation systems and orbit determination. In order to produce timely UT1 estimates, efforts have been made to completely automate the analysis of VLBI Intensive sessions. This involves the automatic processing of X-and S-band group delays. These data contain an unknown number of integer ambiguities in the observed group delays. They are introduced as a side-effect of the bandwidth synthesis technique, which is used to combine correlator results from the narrow channels that span the individual bands. In an automated analysis with the c5++ software the standard approach in resolving the ambiguities is to perform a simplified parameter estimation using a least-squares adjustment (L2-norm minimisation). We implement L1-norm as an alternative estimation method

Ultra-rapid earth rotation determination with VLBI during CONT11 and CONT14

“…There are some small deviations of the ultra-rapid results probably caused by failure of the automated ambiguity resolution that is part of the automated processing. For the ultra-rapid CONT11 and CONT14 analysis, robust methods (Kareinen et al 2016) were not part of the automated analysis chain yet. Table 1 summarizes statistical information of these comparisons in terms of root-mean-square differences (rms), on the level of 9-12 µs.…”

“…It is anticipated that the currently existing strategies for automated analysis Kareinen et al 2015) will be extended in the near future from single baseline observations to network observations. In this context, it is highly desirable to also include robust analysis strategies, as described by Kareinen et al (2016).…”

Estimation of large-scale deformations in VLBI radio source catalogs with mitigation of impact of outliers: A comparison between different L1- and L2-norm-based methods

“…Nevertheless, the L1 norm is considered to be more robust than the L2 norm, as the squared values in the definition of the L2 norm increase sensitivity to outliers, while in the L1 norm, absolute values are only linearly dependent to extreme values. Thus, L1-norm minimization could have some advantages at some stages of analysis, in particular for outlier detection (e.g., see Gontier et al 2001;Kareinen et al 2016, who both used comparisons between residuals of L1 and L2 minimization as a criterion to detect outliers and evaluate the goodness-of-fit).…”