A major source of error in space-based geodesy is the tropospheric delay, which results in excess path length of the Global Navigation Satellite Systems (GNSS) signal as it passes through the neutral atmosphere. Tropospheric models that use real, predicted or empirical meteorological data have been applied to account for this error. The models widely used in geodesy are Hopfield, Saastamoinen, Black, GPS-Code-Analysis –Tool (GCAT) and Minimum Operational Performance Standards (MOPS). In current study, a methodology that utilized real meteorological data obtained by Global Forecast System (GFS) digital weather model in GRIB format was defined and used. Based on data obtained for different months in 2019, which coverage all seasonal variations of ZTDs, it was found that the accuracy obtained by using of real meteorological data of GFS was 1.5 to 2 times higher than that obtained from standard or empirical models. The RMS of Hopfield, Saastamoinen, and Black’s using predicted meteorological parameters showed less than 5% difference from one another (0.063 m, 0.061 m, and 0.062 m, respectively).
Real GPS data of nine GPS stations in 7-day period were used for evolution the zenith part obtained by using of real meteorological data of GFS against that used standard or empirical models. The data were processed using Bernese software version 5.0. The closure error results prove that the model used real meteorological data is the best model in all session.