Comparative performance of PPP software packages in atmospheric delay estimation using GNSS data

Abstract This study utilizes Precise Point Positioning (PPP) for accurate GNSS-based positioning with a single receiver. Seven PPP techniques (GAPS, APPS, CSRS, Magic-GNSS, gLAB, Pride, GAMP) are evaluated for zenith tropospheric delay (ZTD) estimates, comparing them with Radiosonde (RS) and Radio Occultation (RO) data from the same day (36th in 2014). RO data is categorized based on RO Events (ROE) – GNSS observation distances (10 km, 50 km, 50:100 km), while RS data is limited to three stations (ASWN, FARF, PHLW). Generally, the differences among PPP techniques are within 5 cm, except for PRIDE-PPPAR, which exhibits semi-constant zenith wet delay (ZWD) values. Root Mean Square (RMS) and t -test comparisons with RO data reveal that APPS has the lowest RMS values for ZWD at 10 km, while CSRS-PPP performs well with increasing distance. Magic-GNSS demonstrates the highest RMS values at 10: 50 km, while APPS at 50:100 km. RMS values typically increase with greater ROE-GNSS distances, except for Magic-GNSS, which decreases. t -tests indicate statistically insignificant differences across PPP techniques, except for APPS and gLAB in the 50:100 km case. Comparisons with RS data indicate that CSRS-PPP achieves the lowest RMS values for ZWD at ASWN and PHLW, while Magic-GNSS and CSRS-PPP excel at FARF. However, Magic-GNSS, GAMP, GAPS, and Pride yield the highest RMS values at ASWN (00:00), (12:00), FARF, and PHLW, respectively. t -tests show statistically insignificant differences across PPP techniques for the three stations, except for APPS and Pride at ASWN and gLAB at PHLW. These findings demonstrate that CSRS-PPP consistently provides reliable ZWD assessments across various conditions, contributing to enhanced accuracy in GNSS-based atmospheric studies. The comparison of PPP techniques offers valuable insights into the strengths and limitations of each method, with implications for improving atmospheric modeling, positioning applications, and broader GNSS-based research in weather forecasting and climate monitoring.