应用GPT2+Saastamoinen模型对我国IGS测站天顶方向对流层时延误差的研究

he GPS system pinpoints the user by measuring the distance between the satellites and the receiver. However, the refraction in the atmosphere bends the propagation path of the signal and as a result, the distance measured is no longer precise. The prolongation of the satellite signal caused by the troposphere is called the tropospheric delay. A calibration of this effect based on a tropospheric model is necessary in the receiver. GPT2 model is an empirical model to estimate the tropospheric delay based on the global data of temperature and pressure. Nevertheless, its precision and accuracy in real application is uncertain. In this context, this paper tries to apply the GPT2 model, along with the Saastamoinen model, on several IGS stations, randomly selected in China and lay out the errors at different times and stations. Then we will find out the main causes of such errors and discuss its precision in different conditions such as the data used to fit the coefficients, so that we can optimize the GPT2 model accordingly.