The tropospheric refraction is one of the major biases affecting GPS observations. It is dealt with by means of tropospheric models and mapping functions, which allow the computation of corrections to be added to the original measurements. Tropospheric models and mapping functions have been developed based upon data collected by stations mostly located in the Northern Hemisphere. This paper presents a comparative study on the effect of using combinations among models and mapping functions on geodetic positioning with GPS. GPS data collected by the Brazilian RBMC network have been used in this study. This network is composed of 9 permanent GPS tracking stations, which are distributed in such a way that every geographic region has at least one of them. This distribution also allowed a station to be located in every one of the Brazilian climate regions. This fact allowed the development of a comparative study on the behaviour of some tropospheric models, such as Saastamoinen, Hopfield and Ifadis, and the mapping functions Ifadis, Hopfield, Saastamoinen, Herring and Lanyi, using RBMC data. The study was based on 3 baselines in which the solutions, for each component, were analysed by means of comparison with SIRGAS values (South American Geocentric Reference System). Results indicate that the best performance was achieved using Ifadis, Herring and Lanyi mapping functions.
Global Positioning System; Tropospheric Refraction; Satellite Positioning
