Comparison of advanced troposphere models for aiding reduction of PPP convergence time in Australia

Abstract

This paper first analyses the precision of tropospheric zenith total delay (ZTD) values obtained from the empirical models GPT2 and GPT2w, and the numerical weather models (NWM) from Australian Bureau of Meteorology (BoM), and European Centre for Medium-Range Weather Forecasts (ECMWF). Comparison of these ZTD values with IGS ZTD product at four sites showed that the ZTDs from NWM datasets were more precise than the empirical models. The ZTD from BoM data gave the best results, with mean errors between -0.034 m to 0.029 m and standard deviations better than 0.045 m. Next, the PPP convergence time and achievable accuracy using the BoM NWM constrained ZTD by including them as pseudo-observations with a pre-set precision was compared to the case of estimating the troposphere. This resulted in a slight enhancement in convergence time, and improvements in vertical positioning accuracy was found at all the four tested sites at 0.036–0.058 m after 2 min, 0.023–0.038 m after 3 min and 0.013–0.020 m after 5 min of PPP initialisation.