An improved atmospheric weighted mean temperature model and its impact on GNSS precipitable water vapor estimates for China

The atmospheric weighted mean temperature, Tm, is an important parameter for retrieving precipitable water vapor (PWV) from global navigation satellite system (GNSS) signals. There are few empirical, high-precision Tm models for China, which limit the real-time and high-precision application of GNSS meteorology over China. The GPT2w (Global Pressure and Temperature 2 Wet) model, as a state-of-the-art global empirical tropospheric delay model, can provide values for Tm, surface temperature, surface pressure, and water vapor pressure. However, several studies have noted that the GPT2w model has significant systematic errors in the calculation of Tm for China, mainly due to the neglect of the Tm lapse rate. We develop an improved Tm model for China, IGPT2w, by refining the Tm derived from GPT2w using both gridded Tm data and ellipsoidal height grid data from the Global Geodetic Observing System (GGOS) Atmosphere. Both gridded Tm data from the GGOS Atmosphere and radiosonde data from 2015 are used to test the performance of IGPT2w in China. The results are compared with the GPT2w model and the widely used Bevis formula. The results show that IGPT2w yields significant performance against other models in Tm estimation over China, especially in western China, where the significant systematic errors of the GPT2w model are largely eradicated. IGPT2w has sigma PWV and sigma PWV/PWV values of 0.29mm and 1.38% when used to retrieve GNSS-PWV, respectively. Thus, the IGPT2w has significant potential for real-time GNSS-PWV sounding in China, especially when used to retrieve GNSS-PWV values for the study of PWV transportation in the Tibetan Plateau.