The development of the International Global Navigation Satellite System (GNSS) Service (IGS) Real-Time Pilot Project shows promise for real-time GNSS-based precipitable water vapor (GNSS-PWV) retrieval. For better applications of real-time GNSS meteorology over China, a method is proposed to establish a real-time GNSS-PWV monitoring system using the national GNSS network of China. The prototype system generating the zenith tropospheric delay (ZTD) is based on the real-time precise point positioning technique, in which the real-time data streams and state-space-representative satellite orbit and clock corrections are processed. The parallel computing technology is embedded in the system for decoding and processing the real-time data streams, which enables the multistation processing mode and improves the computing efficiency. At the initial phase, a total of 215 global positioning system (GPS) stations from the crustal movement observation network of China are used to generate the real-time GPS-ZTDs. The pressure data for separating the zenith wet delay (ZWD) from the ZTD are obtained from the collocated meteorological sensors. An advanced weighted mean temperature model, namely, Gridded-Mixed Tm, is adopted in the system to determine the conversion factor “Π” (converting GPS-ZWD to GPS-PWV) in real-time mode with the measured temperatures input. The generated real-time GPS-PWV products have a time resolution of 5 min. To validate the system performance, comparisons between the real-time GPS-PWV and the PWVs derived from nearby radiosonde data (RS-PWV) and reanalysis data from the National Centers for Environmental Prediction (NCEP)-Department of Energy Reanalysis II (NCEP-II-PWV) are conducted over a period of 83 days. The comparisons show a mean bias of −0.1 mm with a root mean square (RMS) of 1.7 mm between the real-time GPS-PWV and RS-PWV. The agreement between the real-time GPS-PWV and NCEP-II-PWV is approximately 2.0 mm in terms of RMS and has a mean bias of −0.8 mm. These results confirm that the established system can be used for real-time PWV monitoring across China.
Tropospheric delays are one of the main sources of errors in the Global Navigation Satellite System (GNSS). They are usually corrected by using tropospheric delay models, which makes the accuracy of the models rather critical for accurate positioning. To provide references for suitable models to be chosen for GNSS users in China, we conduct herein a comprehensive study of the performances of the IGGtrop, EGNOS and UNB3m models in China. Firstly, we assess the models using 5 years’ Global Positioning System (GPS) derived Zenith Tropospheric Delay (ZTD) series from 25 stations of the Crustal Movement Observation Network of China (CMONOC). Then we study the effects of the models on satellite positioning by using various Precise Point Positioning (PPP) cases with different tropospheric delay resolutions, the observation data processed in PPP is from 21 base stations of CMONOC for a whole year of 2012. The results show that: (1) the Root Mean Square (RMS) of the IGGtrop model is about 4.4 cm, which improves the accuracy of ZTD estimations by about 24% for EGNOS and 19% for UNB3m; (2) The positioning error in the vertical component of the PPP solution obtained by using the IGGtrop model is about 15.0 cm, which is about 30% and 21% smaller than those of the EGNOS and UNB3m models, respectively. In summary, the IGGtrop model achieves the best performance among the three models in the Chinese region.
For the timely retrieval of global positioning system based precipitable water vapor (GPS-PWV) over complex topography when there is no in situ pressure-derived zenith hydrostatic delays (ZHDs), this article evaluates three types of ready-made ZHD products, including two newly released products from the Vienna University of Technology (TU Wien, TUW-VMF3) and GeoForschungsZentrum Potsdam (GFZ-VMF1) and a legacy product TUW-VMF1. We implement them for the first time for GPS-PWV retrieval in regions with highly variable topography, such as the Tibetan plateau (TP). First, we present a refined method [model-assisted ZHD height adjustment, model-assisted ZHD height adjustment (MAZHA)] for implementing the above VMF1/VMF3-like ZHD products. The results reveal that the MAZHA method improves the implementation accuracy over the TP by 70% compared to the conventional method. Then, the above multisource VMF1/VMF3-like ZHD products are evaluated using the in situ pressure-derived ZHDs across the TP. The results show that the GFZ-VMF1 outperforms the TUW-VMF1 mainly because a more advanced underlying numerical weather model (ERA5) are adopted. TUW-VMF3 achieves the best performance mainly because it is provided with improved horizontal resolution. Finally, we find that implementing the state-of-the-art TUW-VMF3 ZHD product for GPS-PWV retrieval yields no statistically significant errors, even under extreme weather conditions, and the retrieved GPS-PWVs can well characterize the PWV diurnal variations in all seasons except winter. Additionally, considering the fact that the operational and forecasted TUW-VMF3 ZHDs present almost equivalent performance, this article is also beneficial to real-time GPS-PWV retrieval using forecasted TUW-VMF3 ZHDs, thus allowing for the ease of real-time GPS-PWV retrieval.
1 September 2001 in Advances in Atmospheric Sciences