Network-Based Ionospheric Gradient Monitoring to Support GBAS

Kurzfassung

Ionospheric anomalies, like large ionospheric gradients, might produce a difference between the ionospheric error experienced by the Ground Based Augmentation System (GBAS) reference station and the aircraft on approach. This ionospheric delay difference could lead to hazardous position errors if undetected. For that reason, the GBAS Approach Service Types (GAST) C and D provide solutions against this threat, but the methods employed still face challenges by limiting the availability in certain cases, especially in regions with severe ionospheric conditions. This issue is caused by the use of very conservative ionospheric threat models derived based on the worst-ever-experienced ionospheric gradients in the relevant region. However, these worst-case gradients occur very rarely. Therefore, this paper proposes a methodology capable of detecting ionospheric gradients in real-time and estimating their parameters in near real-time by using a wide area network of dual-frequency and multi-constellation GNSS monitoring stations. Hence, the GBAS stations could use this information to update the threat model currently applied in their algorithms, which would result in an improvement of the GBAS availability in regions where it is degraded. The detection and estimation algorithm is initially theoretically explained. Then, the performance of this algorithm is evaluated with simulated gradients and with a real gradient, utilizing for both the real measurements recorded by a reference network in Alaska. The synthetic gradients are simulated over the nominal real measurements from this network and all the gradient parameters are modified within their ranges in the already existing threat models. In this way, we assess the performance of our algorithm by comparing the differences between the known simulated gradient parameters and the parameters estimated by our algorithm. Additionally, we also evaluate our algorithm with one real ionospheric gradient measured by the same network in Alaska to study the differences between using simulated gradients and real gradients. Results with both simulated gradients and the real gradient show the potential of our methodology to support GBAS stations by detecting and estimating the ionospheric gradients instead of using worst case models all the time.