Establishing a high-precision real-time ZTD model of China with GPS and ERA5 historical data and its application in PPP

A high-precision real-time troposphere model is constructed by combining ground-based GNSS observation data and the latest European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5). First, the zenith tropospheric delay (ZTD) is extracted in real time with high accuracy by combining the data of more than 500 GNSS stations in the Crustal Movement Observation Network of China (CMONOC) and national reference station network (NRSN); second, a grid model of the elevation normalization model (ENM) in China using ERA5 data is constructed, which takes into account the annual, semiannual and daily cycles. The ZTD estimated by GNSS stations at different heights based on precise point positioning (PPP) is normalized to a uniform height based on ENM; in addition, the optimal smoothing factors of the Gauss distance weighting function in different seasons are determined based on ERA5, which contributes to improved accuracy of ZTD interpolated from GNSS-derived ZTD to ZTD at grid points; finally, a real-time 1° × 1°ZTD grid model of China is created; the broadcast interval is extended to 6 min from few seconds. The new ZTD model has been evaluated using the data of 15 GNSS stations in China in 2020. The test results show that the new ZTD model deviates from the reference value with a mean value better than − 0.09 cm and RMSE, better than 1.44 cm compared with the ZTD estimated by post-processing GNSS, while the mean value of the deviation is -0.13 cm, and the RMSE is approximately 3.11 cm compared with radiosonde-derived ZTD. The new ZTD grid model can be used to enhance GNSS/PPP. Two weeks of GNSS observations, one week in winter and another in summer, were randomly collected for PPP processing. The statistical results show the convergence time in the vertical directions is shortened by 37.4% and 38.6% at the 95% and 68% confidence levels after ZTD constraints are applied to the float PPP solution, respectively.