عنوان مقاله [English]
نویسندگان [English]چکیده [English]
Propagation of radio waves through the troposphere is based on minimum travel time between transmitter and receiver. The troposphere layer affects the propagation of GNSS signals and causes errors in point positioning due to (1) delay of the signal, and (2) change of the curvature of the signal path. For this reason GNSS users apply models for tropospheric error corrections. One of the commonly used models, which is applied in both scientific and commercial software packages is the Saastamoinen model. Since the tropospheric error, unlike the ionosphereic error, cannot be removed via dual frequency receiver observations, application of the tropospheric models based on meteorological information can always improve point positioning accuracy. This becomes more essential in the case of Precise Point Positioning (PPP), which has real-time applications. Atmospheric delays have two components. Wet delay is caused by atmospheric water vapor and dry delay by all other atmospheric constituents. The dry delay can be predicted to better than 1 mm with surface pressure measurement. Wet GNSS delay is highly variable and cannot be accurately predicted from surface observations. In this paper the wet Saastamoinen model is evaluated based on Zenith Total Delay (ZTD) computed from one year (2005) of permanent GNSS observations of the National Cartography Center (NCC) of Iran. The wet Saastamoinen model is dependent on constants, which are estimated from local observations. Our evaluation procedure can be summarized as follows:
1- GNSS ZTD is estimated from GNSS observation equations.
2- ZTD is computed from surface pressure, temperature and humidity observations.
3- Root Mean Square (RMS) between the two above solutions is determined.
4- The value of RMS is used as an efficiency test of constant coefficients of the Saastamoinen model.
The results of our test computations show that the constant coefficients of the Saastamoinen model have enough accuracy for the tropospheric corrections needed for PPP applications and as such there is no need to estimate those constants based on local atmospheric observations.