Accuracy Assessment and Rainfall Intensity Response Mechanisms in GNSS Real-Time PPP Water Vapor Retrieval: Coupling Effects of Multi-Source Real-Time Products and Meteorological Conditions

Abstract

The Global Navigation Satellite System (GNSS) real-time Precise Point Positioning (PPP) technology offers a highly efficient approach for atmospheric Precipitable Water Vapor (PWV) monitoring and extreme weather warning. However, its retrieval accuracy is limited by the performance of orbit and clock products, as well as the coupling effects of water vapor phase transitions and signal attenuation under precipitation conditions. This study utilized observation data from the Hong Kong Continuously Operating Reference Stations network between 2016 and 2022, employing real-time products from the International GNSS Service (IGS) and the Center National d'Études Spatiales (CNES) for PPP solutions. Validated against IGS post-processed products and radiosonde data, the PWV retrieval accuracy of IGS real-time products outperformed that of CNES products. The rainfall conditions exhibited systematic impacts on retrieval errors: compared to non-rainfall conditions, the root mean square error of PWV increased from 1.54 to 1.89 mm during rainfall events, with retrieval errors escalating as rainfall intensity increased. Analysis of the response relationship between varying rainfall intensities and PWV identified a distinct three-step correlation between PWV and precipitation processes, that is the accumulation of water vapor before rainfall, the rapid release during precipitation and the dissipation process after rainfall. The magnitude of PWV variations demonstrated a positive correlation with rainfall intensity. This research reveals the coupling influence mechanisms of real-time product selection and meteorological conditions on GNSS PWV retrieval, providing a theoretical foundation for optimizing real-time monitoring models under heavy rainfall scenarios and enhancing the reliability of extreme weather early warning systems.