COMPARATIVE ANALYSIS OF ZTD OBTAINED FROM DIFFERENTIAL AND ABSOLUTE APPROACH BASED ON LONG TIME SERIES

Over the years, processing strategies adopted for GNSS (Global Navigation Satellite Systems) observations have developed. One of the commonly used approach for coordinate estimation is based on double difference observations (DD) and it allows to obtain submillimeter precision. However, in recent years, there was also a marked development of precise absolute methods, called PPP (Precise Point Positioning). On the one hand, PPP method is more complicated than DD, because it requires to take into account much more effects, than in case of differential approach. On the other hand, thanks to the progress in the field of algorithms and modeling methods, PPP allows to obtain results with an accuracy of more and more close to the accuracy of DD method. Moreover, in contrast to the differential approach, absolute positioning is not affected by network adjustment. Despite of differences in data processing, troposphere has major influence on both of these methods. Its exact impact on position determination is called ZTD (Zenith Tropospheric Delay) and depends on such meteorological parameters like air pressure, temperature and humidity. While the delay caused by hydrostatic part of the troposphere (ZHD, Zenith Hydrostatic Delay) is possible to model with sufficient accuracy, the delay caused by wet part of troposphere (ZWD, Zenith Wet Delay) is difficult to model, due to the large variability of water vapor content. Therefore, in order to ensure high precision of coordinates, value of ZWD has to be estimated in positioning process regardless used method, DD or PPP. Consequently, for the same stations troposphere influence should be the same. However, in practice ZTD obtained according to various approaches may be different in terms of ZTD and its variability. In this paper a comparative analysis of ZTD estimated for selected 27 EPN (EUREF Permanent Network) stations, according to Double Difference and Precise Point Positioning approach, has been carried out. In order to ensure fully coherence between this two methods, all calculations were conducted in Bernese 5.2 software, with using the same models and parameters, which are recommended in EPN guidelines. Based on the 19-year ZTD time series (1996-2014), short and long term differences between these two methods were investigated, as well as statistical parameters of ZTD differences.