Abstract

This paper describes GNSS tropospheric impacts, acquired from radiosonde data in Ljubljana over the autumn-winter period in 2017. The tropospheric effect is the same for code and carrier-phase observations and is not frequency dependent. It appears as a bending path of the signal, which results in a propagation delay of the refracted GNSS-signal acquisition. Precise modelling of tropospheric slant delay requires knowledge of the meteorological parameters from the full signal’s path, which is difficult to realize. In GNSS, we use standard atmosphere or other meteorological models as a substitute for the real meteorological data throughout the signal propagation although there is a lack of information on humidity in most cases. However, occasionally we can verify how the impacts from models match the actual impacts. For this, we used radiosonde measurements of the Slovenian Environment Agency to determine the hydrostatic and wet component of the effect. In the first component, the results differed from that of the standard atmosphere in a range of several decimeters. Daily and monthly changes in a much smaller wet component, which were up to 25 cm, were apparent as well. If in GNSS processing only the model removes the impact, the defective part will cause errors in coordinates. Therefore, in high-accuracy positioning requirements we have to estimate the tropospheric impact as an additional parameter.

Key words: GNSS observations, tropospheric impacts, standard atmosphere, radiosonde measurements