Terrestrial reference frame scale drift anomalies in VLBI and the contribution of Ny-Ålesund radio telescopes

Abstract

The scale of the last realization of the International Terrestrial Reference System, the frame ITRF2020, is derived from a combination of Very Long Baseline Interferometry (VLBI) and Satellite Laser Ranging observations. During the computations of ITRF2020, an unexpected VLBI scale drift after 2013.75 led to the exclusion of sessions beyond this epoch from the scale definition. An independent research of the origin of the suspected scale drift conducted by researchers of the Onsala Space Observatory (Chalmers University of Technology) suggested a mis-modeling of the station movement of the uniquely located VLBI antenna, NYALES20 (Svalbard, Norway), as the likely cause of these anomalies. The underlying assumption was motivated by station movement information available from the co-located Global Navigation Satellite Systems (GNSS) receiver, NYAL, aligning perfectly with the discontinuities visible in the session-wise station positions of NYALES20. In our investigations, the additional discontinuity intervals are introduced in the determination of a VLBI-only terrestrial reference frame (TRF), and their impact on the session-wise scale is analyzed in comparison to the results of a reference solution. We evaluate the importance of four additional intervals suggested by the Onsala Space Observatory and the strategy of the ITRF2020-u2023, where only two of these intervals are implemented. New in our investigations is the analysis of the scale evolution based on a TRF determination from a combination at normal equation level with our software VieCompy. Through the latter, added value is warranted by applying a different concept of TRF combination, which differs from the combination of solutions at parameter level with full covariance transfer, as applied in the ITRF computations using the CATREF (Combination and Analysis of Terrestrial Reference Frame) software. In addition, our scale computations and considerations are entirely independent of the ITRF calculations and we extended the analyzed session-wise scale time series to 2024.0 to get a better insight into the long-term development. The comparison of the time series of session-wise estimated scale factors with the reference solution reveals a significant reduction in the VLBI drift by more than 50 % when accounting for a modified station movement model for NYALES20. Therefore, we strongly advocate introducing an optimized station movement model for NYALES20 to account for climate-related processes to ensure the stability of the scale of global reference frames.