Multi-GNSS time transfer with different modeling of station coordinates and troposphere delays

In GNSS solutions, the receiver clock parameter (Clc), the station height component (Up), and the Zenith Tropospheric Delay (ZTD) are correlated. However, some of these parameters are considered known values to avoid correlations and reduce time convergence. In Precise Point Positioning (PPP), the receiver clock parameter is estimated at each measurement epoch together with the coordinates and potentially ZTD. However, coordinates can be taken from daily solutions, whereas ZTD can be taken from an external model and fixed. We study twelve PPP strategies using International GNSS Service (IGS) stations equipped with external hydrogen masers, employing two different ZTD models, and various configurations for estimating or fixing Up component and ZTD. Each day is analyzed independently utilizing multi-GNSS PPP, post-processing, and Forward–Backward–Forward estimation techniques. The most accurate results are obtained in four strategies, two of which estimate all three parameters: Clc, Up, and ZTD; whereas in the other two strategies, the station coordinates are derived from the last epoch of the daily pre-processed PPP solution. Fixing station coordinates to an a priori position from ITRF2020 leads to inferior results. Fixing troposphere parameters to GMF or VMF3 models also leads to worse results than estimating corrections to wet delay. Solutions using VMF3 are only slightly better than solutions based on GMF. Estimating all parameters simultaneously, however, provides the most comprehensive solution despite existing correlations. The best results for parameter stability and Time Transfer are achieved in strategy estimating Up, Clc, and ZTD using the VMF3 model updated every 6 h for ZTD.