Articles | Volume 30, issue 9
Ann. Geophys., 30, 1379–1391, 2012

Special issue: Advanced Global Navigation Satellite Systems tropospheric...

Ann. Geophys., 30, 1379–1391, 2012

Regular paper 27 Sep 2012

Regular paper | 27 Sep 2012

Near real-time estimation of water vapour in the troposphere using ground GNSS and the meteorological data

J. Bosy, J. Kaplon, W. Rohm, J. Sierny, and T. Hadas J. Bosy et al.
  • Institute of Geodesy and Geoinformatics, Wroclaw University of Environmental and Life Sciences, Grunwaldzka 53, 50-357 Wroclaw, Poland

Abstract. The near real-time (NRT) high resolution water vapour distribution models can be constructed based on GNSS observations delivered from Ground Base Augmentation Systems (GBAS) and ground meteorological data. Since 2008 in the territory of Poland, a GBAS system called ASG-EUPOS (Active Geodetic Network) has been operating. This paper addresses the problems concerning construction of the NRT model of water vapour distribution in the troposphere near Poland. The first section presents all available GNSS and ground meteorological stations in the area of Poland and neighbouring countries. In this section, data feeding scheme is discussed, together with timeline and time resolution. The high consistency between measured and interpolated temperature value is shown, whereas some discrepancy in the pressure is observed. In the second section, the NRT GNSS data processing strategy of ASG-EUPOS network is discussed. Preliminary results show fine alignment of the obtained Zenith Troposphere Delays (ZTDs) with reference data from European Permanent Network (EPN) processing center. The validation of NRT troposphere products against daily solution shows 15 mm standard deviation of obtained ZTD differences. The last section presents the first results of 2-D water vapour distribution above the GNSS network and application of the tomographic model to 3-D distribution of water vapour in the atmosphere. The GNSS tomography model, working on the simulated data from numerical forecast model, shows high consistency with the reference data (by means of standard deviation 4 mm km−1 or 4 ppm), however, noise analysis shows high solution sensitivity to errors in observations. The discrepancy for real data preliminary solution (measured as a mean standard deviation) between reference NWP data and tomography data was on the level of 9 mm km−1 (or 9 ppm) in terms of wet refractivity.