Real-time sensing of atmospheric water vapor from multi-GNSS constellations Article Swipe

With the modernization of GPS (Global Positioning System), recovery of GLONASS (the Russian GLObal NAvigation Satellite System), and newly emerging constellations, like the Chinese BeiDou Navigation Satellite System and the European Galileo system, the world of Global Navigation Satellite Systems (GNSS) has experienced dramatic changes within the field of multi-constellation GNSS. The rapid development of the current GNSS constellation brings a promising prospect for the real-time retrieval of tropospheric delay parameters like zenith total delay (ZTD) and precipitable water vapor (PWV), which is of great benefit for supporting the time-critical meteorological applications such as nowcasting or severe weather event monitoring. With the increasing development of the existing GNSS and the real-time PPP (precise point positioning) technique, the objective of this thesis is to develop a real-time GNSS PPP processing for time-critical meteorological applications. The core research and the contributions of this thesis are summarized as following: We develop a real-time ZTD/PWV processing based on the observations from individual system: GPS, GLONASS, and BeiDou. The performance of ZTD and PWV derived from each system using real-time PPP technique is investigated. The contribution of combining GLONASS or BeiDou with GPS for ZTD/PWV retrieving is evaluated as well. Our results show that the real-time GLONASS ZTD series agree quite well with the GPS ZTD series in general: the root mean square (RMS) values of ZTD differences are about 8 mm, which is equal to 1.2 mm in PWV. The real-time BeiDou-only ZTD series also show good agreement with the GPS-only ones: the RMS values are about 11-16 mm, of about 2-3 mm in PWV. The real-time ZTD/PWV derived from GPS-only, GLONASS-only, BeiDou-only, GPS/GLONASS and GPS/BeiDou combined solutions are compared with those derived from the Very Long Baseline Interferometry (VLBI) and radiosondes. The comparisons show that GLONASS can contribute to real-time meteorological applications, with almost the same accuracy as GPS, similar for BeiDou which is slightly less accurate than GPS. Besides, more accurate and reliable water vapor estimates can be obtained if the GLONASS or BeiDou observations are combined with the GPS observations in the real-time PPP data processing, about 1.5-2.3 mm in PWV. A multi-GNSS (GLONASS+GPS+Galileo+BeiDou) model is also presented to fully exploit all available observations from the current GNSS for deriving the real-time ZTD/PWV based on the real-time PPP processing. Observations from stations capable of tracking the multi-constellation are processed in both single- and multi-system modes. The ZTD/PWV retrieved from the individual GNSS and the combined multi-GNSS solutions are assessed by comparing with data from the nearby radiosonde stations. The benefit of the multi-GNSS combination for real-time water vapor derivation is evaluated. The results show that an accuracy at the mm-level, i.e. 1.2-1.3 mm, for the real-time PWV estimates is achievable with the multi-GNSS processing. The obtained water vapor estimates with enhanced accuracy and reliability reveal the potential benefits of multi-GNSS fusion for atmospheric monitoring systems, in particular for the time-critical meteorological applications. The tropospheric horizontal gradients with high spatiotemporal resolutions provide important information to describe the azimuthally asymmetric delays and significantly increase the ability of ground-based GNSS within the field of meteorological studies. The recent rapid development of multi-GNSS constellations has potential to provide such high-resolution gradients with a significant degree of accuracy. We develop a multi-GNSS processing for the precise retrieval of high-resolution tropospheric gradients. The tropospheric gradients with different temporal resolutions, retrieved from both single-system and multi-GNSS solutions, are validated using independent numerical weather model (NWM) and water vapor radiometer (WVR) data. The benefits of multi-GNSS processing for the retrieval of tropospheric gradients, as well as for the improvement of precise positioning, are demonstrated. The results show that the multi-GNSS high-resolution gradients agree well with those derived from the NWM and WVR, in particular for the fast-changing peaks, which are mostly associated with synoptic fronts. The multi-GNSS gradients behave in a much more stable manner than the single-system estimates, especially in cases of high temporal resolution, benefiting from the increased number of observed satellites and improved observation geometry. Furthermore, the precision of station positions can also be noticeably improved by the multi-GNSS fusion, and enhanced results can be achieved if the high-resolution gradient estimation is performed, instead of the commonly used daily gradient estimation in the multi-GNSS data processing. Precise positioning with the current BeiDou is proven to be of comparable accuracy to GPS, which is at centimeter level for the horizontal components and sub-decimeter level for the vertical component. But the BeiDou PPP shows its limitation in requiring a relatively long convergence time. Thus, we develop a NWM augmented PPP processing algorithm to improve BeiDou precise positioning. Tropospheric delay parameters, i.e., zenith delays, mapping functions, and horizontal delay gradients, derived from short-range forecasts from the Global Forecast System (GFS) of the National Centers for Environmental Prediction (NCEP) are applied into BeiDou real-time PPP. Observational data from stations that are capable of tracking the BeiDou constellation are processed with both the standard PPP and the introduced NWM augmented PPP processing. The positioning results show that an improvement of convergence time up to 60.0 % and 66.7 % for the east and vertical components, respectively, can be achieved with the NWM augmented PPP solution compared to the standard PPP solutions, while only slight improvement of the solution convergence can be found for the north component. A positioning accuracy of 2.0 cm for the north component is achieved with the NWM augmented PPP, in comparison to 3.7 cm of the standard PPP, showing an improvement of 45.9 %. Compared to the accuracy of 5.7 cm for the east component derived from the standard PPP solution, the one of the NWM augmented PPP solution is improved to 3.5 cm, by about 38.6 %. The positioning accuracy for the up component improves from 11.4 cm with the standard PPP solution to 8.0 cm with the NWM augmented PPP solution, an improvement of 29.8 %. Significant improvement on positioning accuracy, reliability, as well as convergence time with the multi-GNSS fusion can be observed in comparison with the single-system processing like GPS. In this study, a NWM augmented PPP processing system is developed to improve the multi-GNSS precise positioning. Tropospheric delay parameters which are derived from the ECMWF analysis are applied to the multi-GNSS PPP (a combination of four systems: GPS, GLONASS, Galileo, and BeiDou). Observations of stations from the IGS Multi-GNSS Experiments (MGEX) network are processed, with both the standard multi-GNSS PPP and the proposed NWM augmented multi-GNSS PPP processing. The high quality and accuracy of the tropospheric delay parameters derived from ECMWF are demonstrated through comparison and validation with the IGS final tropospheric delay products. Compared to the standard PPP solution, the convergence time is shortened by 20.0 %, 32.0 %, and 25.0 % for the north, east, and vertical components, respectively, with the NWM augmented PPP solution. The positioning accuracy also benefits from the NWM augmented PPP solution, which gets improved by 2.5 %, 12.1 %, and 18.7 % for the north, east, and vertical components, respectively.