Empirical model for total precipitable water retrieval from ground-based GNSS observations in Thailand

Retrieval of Total Precipitable Water (TPW) from ground-based Global Navigation Satellite System (GNSS) observations is a challenging task due to its real-time and high temporal resolution nature. In this paper, we present a method for establishing an analytical model for retrieving the TPW based on the GNSS observations over 1-year period from 12 stations distributed across Thailand. The derived Zenith Total Delay (ZTD) at all stations agrees well with the TPW data available from the Global Data Assimilation System (GDAS) Numerical Weather Prediction (NWP) model. First, a unique relationship between the ZTD and the TPW was established by taking into account the variations in station altitudes. Then, a bias correction technique with Probability Distribution Function (PDF) matching was applied to improve the final model. The inversion model of the TPW from ZTD was then easily obtained by using a numerical technique. The performance of our method has been successfully evaluated on an independent test data. This model can be useful for near real-time TPW measurements from globally available GNSS receivers.