GFZ
Home
go to search page

Exploitation of GNSS tropospheric gradients for severe weather Monitoring And Prediction | EGMAP

Global Navigation Satellite Systems (GNSS) have become vital to our daily lives. Atmospheric water vapor monitoring using GNSS regional ground networks has improved existing meteorological observation systems through its widely distributed ground station network. GNSS stands out from other observation systems because of its low operating costs, all-weather availability, and excellent spatio-temporal resolution. High-quality observations are significant for accurately representing variables in a weather model. Through data assimilation (DA), we can combine observations and numerical weather models optimally.

Our research project, "Exploitation of GNSS Tropospheric Gradients for Severe Weather Monitoring and Prediction" (EGMAP), focuses on the impact of GNSS tropospheric gradients and their effective use for operational forecasting of severe weather. EGMAP is a research collaboration between Technische Universität Berlin (TUB) and the Helmholtz Centre for Geosciences (GFZ), funded by the German Research Foundation (DFG). As a part of this project, at GFZ, we have developed a new operator (Zus et al., 2023) to assimilate Tropospheric Gradient (TG) observations in the Weather Research and Forecasting (WRF) model. TGs are retrieved from the GNSS station network along with Zenith Total Delay (ZTD) observations.

Integrated water vapor
Figure 1: Left: Integrated water vapor [IWV, mm] calculated in near real time by the GFZ for about 104 GNSS stations on July 14, 2021 (flood disaster in the Ahr valley). Right: Improvement of the specific humidity distribution by GNSS using the six-hourly WRF forecast for June/July 2021 (Thundathil et al., 2024). (GFZ)

The operational use of GNSS ZTDs from GFZ by the Deutscher Wetterdienst (DWD) began on April 21, 2020. In parallel, novel observation operators were developed to assimilate improved GNSS products and impact studies were carried out to quantify the forecast improvement (Thundathil et al., 2024). For the study, data from around 104 ground stations in Germany were evaluated over a period of two months, including during the flood disaster in the Ahr Valley in July 2021. For the first time, it was shown that, together with the innovative use of horizontal moisture information (TGs), significant improvements in the predicted moisture fields can be achieved (up to 10% in the lower troposphere). This enables, in particular, more precise predictions of the location and extent of precipitation areas. The new observation operator will be integrated into the DWD's operational forecast system.

[Translate to English:] Empfindlichkeitsexperiment
Figure 2: Left: Spatial impact of ZTDs and TGs with sparsely distributed (1-degree) GNSS stations and only ZTDs with densely distributed (0.5-degree) GNSS stations. Right: Root Mean Square Error profile comparison of the ZTD (dense network) and ZTDGRA (sparse network) runs (Thundathil et al., 2025 in prep.). (GFZ)

Through a TG sensitivity experiment conducted by de-densifying the GNSS network, it was found that the assimilation of TGs in addition to ZTDs from a sparse station (1-degree) network provides an equivalent improvement as that from the assimilation of only ZTDs from a dense station (0.5-degree) network (Thundathil et al., 2025 AMT preprint). This highlights the potential of TGs to improve humidity fields at places with a few GNSS stations.

[Translate to English:] Reflektivitätskarte für Supertaifun Koinu, 2023
Figure 3: Reflectivity (dBZ) map of Super Typhoon Koinu, which made its landfall on 4 October 2023, derived from WRF DA analysis of gradient-only assimilation run. (a) shows the analysis valid at 3 October 12 UTC, and (b), (c) are those valid at 4 October 00 and 12 UTCs, respectively (Thundathil et al., 2025 in prep.). (GFZ)

Assimilation of TGs aims to improve severe weather predictions. The GNSS meteorology group recently conducted a GNSS DA study on Super Typhoon "Koinu" over Taiwan. The simulation ran from 1 September to 30 October 2023, with the typhoon forming on 28 September. Although the DA was initiated a month earlier, the NWP system successfully captured the typhoon's genesis and structure, including its "eye" with calm, humid conditions and its wall with high reflectivity and heavy rain. Assimilation of TGs significantly reduced the track error to 43 km from 54 km in the Control run with no GNSS DA. Around 250 stations across Taiwan and parts of Japan were used.

 

References:

Zus, F., Thundathil, R., Dick, G., & Wickert, J. (2023). Fast Observation Operator for Global Navigation Satellite System Tropospheric Gradients. Remote Sensing, 15(21), 5114.

Thundathil, R. M., Zus, F., Dick, G., & Wickert, J. (2023). Assimilation of GNSS Tropospheric Gradients into the Weather Research and Forecasting Model Version 4.4. 1. Geoscientific Model Development Discussions, 2023, 1-38.

Thundathil, R., Zus, F., Dick, G., & Wickert, J. (2025). Assimilation of GNSS Zenith Delays and Tropospheric Gradients: A Sensitivity Study utilizing sparse and dense station networks. EGUsphere, 2025, 1-24. (AMT preprint – under review)

Link to GEPRIS project data base of DFG:
gepris.dfg.de/gepris/projekt/443676585

back to top of main content