Investigating the independently derived tropospheric parameters at GNSS-VLBI co-located sites

This paper presents the next step of the recently undertaken comparison of the tropospheric parameters independently derived from the analysis of Very Long Baseline Interferometry (VLBI) and Global Navigation Satellite Systems (GNSS) data. We established a concept of cross-validation of the tropospheric parameters which allows us to focus now on the investigation of the discrepancies detected in the cross-validation. In particular, the dataset and its parameterization are examined in view of their impact on the obtained biases. The most recent 5 years of observations are considered, during which the sufficiently dense time series of observations are available for the legacy VLBI stations as well as for the new generation antennas following the VLBI Global Observing System (VGOS) specifications. In our study the co-location sites are limited to those observatories, where the legacy VLBI antenna, VGOS antenna and GNSS receiver are present. Furthermore, the tropospheric parameters are compared only at the common epochs, which are chosen to include the independently conducted observations of the legacy VLBI antenna, VGOS antenna and GNSS receiver during a defined time span. The corresponding solution parametrization is varied between intervals of VGOS, legacy and GNSS solution parametrizations in order to identify the most optimal duration. While the parameterization of the GNSS solution considers the tropospheric variations stable over 2 hours, the parameterization of VLBI solutions for legacy antennas is known to be reliable over a shorter 1-hour interval. The highly dense VGOS observations allow us to reach up to 20-minute duration without losing performance.