Faculty of Computing, Health and Science
School of Engineering (SOE)
Proper modelling of the temporal correlations of the zenith wet delay (ZWD) is important in some of the Global Navigation Satellite Systems (GNSS) applications such as estimation of the Perceptible Water Vapour (PWV), and methods such as Precise Point Positioning (PPP). The random walk (RW) and the first-order Gauss- Markov (GM) autocorrelation model are commonly used for the dynamic modelling of ZWD in Kalman filtering of GNSS measurements. However, it was found that the GM model consistently underestimates the temporal correlations that exist among the ZWD estimates. Therefore, a new autocorrelation dynamic model is proposed in a form similar to that of a hyperbolic function. The impact of the proposed dynamic model on the near-real time estimation of the ZWD was tested and its results were compared to that of the GM model as well as the RW model. In this test, GPS dual-frequency data collected on the 25th Jan 2010 at two Western Australian IGS stations, namely, Yarragadee and Karratha, were used. Results showed that the proposed model outperformed the GM model, and when added to hydrostatic models were able to provide near real-time (with 30 seconds intervals) ZTD estimates to within a few cm accuracy.