Consistent Weather-Dependent Corrections for Space Geodesy; Validation via GNSS and VLBI Analysis

Whether utilizing consistent models to describe weather-dependent effects on geodetic observations or a collection of models that yield accurate results individually, has remained an unanswered question in space geodesy. We study the superimposed effect of atmospheric refraction and environmental loading on GNSS and VLBI data analysis. Variable atmospheric refraction and site displacements induced by non-tidal geophysical loading constitute a large contribution to the modern space geodetic data analysis error budget. State-of-the-art weather models such as ECMWF‘s operational analysis and the atmospheric reanalysis ERA5 have proven to be an accurate forcing data set to drive relevant measurement corrections. Since the effects of these phenomena (refraction and loading) on geodetic observables exhibit non-trivial correlations with each other at a multitude of spatio-temporal scales, employing inconsistent data sets may deteriorate the geodetic results, such as the station coordinates and the Earth rotation parameters. The purpose of this contribution is twofold: (i) present our strategy towards consistent weather-dependent models and explore the merits stemming from the adoption thereof, and (ii) evaluate atmospheric delay and geophysical loading models consistently derived from ERA5 via the reanalysis of GNSS and VLBI data. To identify the extent to which the application of inconsistently forced reduction models causes discrepancies in the geodetic adjustment, we carried out a series of Monte Carlo runs. GNSS and VLBI observations were simulated employing ERA5-driven data (ray-traced delays and loading displacements), but reduced by applying a version thereof subjected to systematic and random noise driven from the performance of state-of-the-art models, at the observation equation level. To evaluate the model coupling with real data, we conduct a GNSS and VLBI repro and compare our new solutions to the GFZ‘s contribution to ITRF2020.