Effects of Meteorological Input Data on the VLBI Station Coordinates, Network Scale, and EOP

Abstract

The requirement for the quantification of sea level rise and other global change phenomena are precise and stable reference frames. To accord with these science driven requirements, GGOS, the Global Geodetic Observing System, aims for reference frame accuracy of 1 mm, stability of 0.1 mm/year, and Earth Orientation Parameters accuracy of ∼ 30 μas. At this level of precision a variety of otherwise neglected effects has to be taken into account. In our article we investigate whether the use of meteorological data for the analysis of space geodetic techniques could be one of these effects that need to be considered. The article moves forward in finding an optimized meteorological data set in order to provide results at the required accuracy level. At the Very Long Baseline Interferometry (VLBI) stations of the International VLBI Service for Geodesy and Astrometry meteorological data are observed and are customarily applied for analyses. Observed data can contain outliers and inhomogeneities that have to be appropriately accounted for. We test the two blind models GPT and GPT2 as a replacement for locally observed data. We find that both cause small seasonal signals in the celestial pole offsets with an amplitude of ∼ 15 μas. We also assess whether the mean values as obtained by extrapolation of surface or near-surface data of ERA-Interim reanalysis product or of the World Meteorological Organization can be used as reference values for homogenization of the observed meteorological data. In particular at sites where the model surface height and the height of the space geodetic reference point differ significantly, very large differences were found. Hence, we cannot recommend shifting the mean value of the observed data to the mean values derived by hypsometrically adjusted surface data.