Comparison of GPS precipitable water vapor and meteorological parameters during rainfalls in Tehran

In this study, the sensing of water vapor using the global positioning system (GPS) was investigated in Tehran. Water vapor mean temperature Tm is a key parameter in conversion of GPS zenith wet delay to precipitable water vapor (PWV). Then, by using 8 years of radiosonde and surface temperature data, we achieved a new site-specific Tm model in Tehran. After the comparison of Bevis et al. (J Geophys Res 97(D14):15787–15801, 1992) and the site-specific models, a mean bias error of −1.3 K was found for Bevis model, while this is less than 0.1 K for the new local model. Therefore, PWV time series were generated for the ground-based GPS site in Tehran from the beginning of 2005 to the end of 2013. Comparing the GPS with the radiosonde PWV, it was shown that the GPS can be used to measure the PWV with high precision. The root mean square error for differences between the GPS and radiosonde was derived as 1.5 mm. The PWV has significant relationship with precipitation in our case study. The thresholds of the relative humidity anomaly with different PWV anomaly related to rainfall occurrences are also considered in this research. The analyses show that use of PWV anomaly condition together with surface meteorological parameters reduces the number of false rainfall recognitions significantly.