GNSS water vapor inversion and spatiotemporal variation analysis considering altitude influence

This study aims to investigate the influence of altitude on the accuracy and spatio temporal variation of atmospheric precipitable water vapor (PWV) derived from the GNSS. By comparing GNSS-derived PWV data with radiosonde and ERA5 data, and utilizing singular spectrum analysis and fast Fourier transform, the accuracy of PWV retrieval in regions with varying altitudes was systematically assessed, and the comprehensive effects of altitude on the spatiotemporal distribution of PWV were explored. The results indicate that GNSS-derived PWV exhibits high reliability across different altitude regions, with a better agreement with radiosonde data than ERA5, and shows more stable accuracy in high-altitude areas. The study reveals a significant trend of PWV variation, with a nonlinear increase observed between 2011 and 2020 at an average rate of 0.12 mm/year, particularly accelerating between 2015 and 2017. PWV demonstrates distinct annual, semi-annual, and 4-month periodic variations, with amplitudes of 8 mm, 2.2 mm, and 0.7 mm, respectively. The impact of altitude on the spatiotemporal patterns of PWV was analyzed, showing that annual cycles are more pronounced in high-altitude areas, while long-term trends dominate in low-altitude regions, with seasonal variations being especially prominent in high-altitude regions. Spatially, PWV values are highest in the southeastern low-altitude regions and lowest in the southwestern high-altitude regions, reflecting the significant influence of topography on PWV distribution. This research provides scientific support for optimizing PWV retrieval models in high-altitude areas and offers new perspectives for understanding regional climate change and its mechanisms.