Analysis on the performances of the GNSS tropospheric delay correction models

Abstract

Tropospheric delay is one of the important factors affecting GNSS positioning accuracy, and there are different ways to deal with the multiple measurement situations. In short baseline measurements, the difference method is commonly used to eliminate tropospheric errors. However, it cannot be used in long baseline measurements or complex weather since it still has great influences on precision measurement after difference calculation. Therefore, modelling method is usually used to reduce tropospheric delay. As it is well known, there are three types of commonly used tropospheric delay correction models, which are suitable for different situations. When any model is used to solve the tropospheric delay in a large scale, there is always an error between the model value and the actual one. In order to investigate the applicability of the three models in different atmospheric conditions, we actually used the measured meteorological data provided by IGS (International GNSS Service) stations as a reference, and then calculated the ZTD (Zenith Tropospheric Delay) with the different models, including Hopfield model, Saastamoinen model and Black model. The calculation results indicate that Saastamoinen model is the most robust and practical model.