Assessing Non-tidal Atmospheric Loading Effects on GNSS Position Time Series: A Comparison of Processing Strategies

Abstract Non-Tidal Atmospheric Loading (NTAL) plays a crucial role in the precision and reliability of GNSS-based positioning and geophysical interpretations, particularly in high-latitude regions, sensitive to atmospheric dynamics. This investigation examines the influence of non-tidal atmospheric loading on GNSS time series and velocities derived from them for high-latitude regions. With a dataset from 2020 to 2023, we process a GNSS network across northern Europe, focusing on the Finnish permanent GNSS network (FinnRef). Using GAMIT/GLOBK software, where corrections are applied at the observation level, we incorporate a new atmospheric grid model derived from the European Centre for Medium-Range Weather Forecasts (ECMWF) numerical weather data. This model provides higher spatial resolution compared to previously available models in GAMIT/GLOBK. Temporal variability of NTAL-corrected GNSS time series is reduced by 17% in the vertical component, and by 8% and 2% in the north and east components, respectively, across the FinnRef network. Additionally, our results highlight that NTAL correction lowers vertical trend uncertainty by an average of 33.5%. Besides evaluating metrics such as spectral power density (PSD) and annual amplitude variation, we observe that the spectral index of the vertical component drops from − 1.44 to − 0.9, indicating reduced long-term noise correlation. We also compare this observation-level approach with an alternative method that applies NTAL corrections at the raw-data level and find that the observation-level correction shows slightly better performance. These results demonstrate that significant improvements in the stability of GNSS time series can be expected after NTAL application, especially in the vertical component.