The master thesis combines the fields of geodesy in meteorology, which both in various specific ways enable the determination of the zenith tropospheric delay of the signals. The troposphere refracts the signal and therefore prolongs the range between a receiver and a satellite according to the propagation in vacuum. The delay of the incoming signal, due to the influence of the troposphere, can be calculated directly from the meteorological data obtained from the radiosonde measurements. There is also an indirect way of estimating the impact, i.e. from GNSS carrier-phase processing. Each of the approach has distinctive pros and cons, so the main objective of the research was to compare the results as well as to combine them for the further impact assessments. For this, the 2017 year-round data of the zenith tropospheric delay was acquired from the Slovene continuously operating reference station network, SIGNAL. Zenith tropospheric delays from radiosonde were improved to be comparable with GNSS processing results, so in only some rare cases differences exceeded 5 centimetres. From the radiosonde data, changes in dry and wet components of the tropospheric zenith delays of the selected one-week periods in different seasons were analysed. Secondly, from the GNSS processing results in the SIGNAL network, daily changes in the zenith tropospheric delay, i.e. combined for dry and wet component, were acquired for the very same intervals. Eventually, tropospheric impacts for the specific baseline endpoints were modelled to show that in the situations with significant altitude differences the tropospheric impacts cannot be easily eliminated by the use of tropospheric models. The conclusion is, that in such situations, as well as in the absolute technique of precise point positioning, the influence of the troposphere should be assessed as an additional unknown in the carrier-phase processing.