Automatic Detection of Large‐Scale Traveling Ionospheric Disturbances Using GNSS Data and Image Processing Techniques

Large-scale traveling ionospheric disturbances (LSTIDs) are wave-like structures propagating through the ionosphere from auroral to lower latitudes, affecting radar-based technologies and satellite navigation. Due to their disruptive effects, timely detection is crucial. This work presents an autonomous detection technique based on travel-time diagrams of detrended Total Electron Content (TEC). Our method employs a network of GNSS receivers across Europe, selected using a K-Means clustering algorithm to ensure comprehensive spatial coverage. Applying image processing techniques, we identify propagating features of positive and negative amplitude in detrended TEC keograms. These features are fitted using the RANdom SAmple Consensus method to estimate propagation speed along the North-South direction. Following feature identification, we refine our results by merging positive and negative perturbations and filtering out false positives based on physical constraints. This methodology allowed us to compile a catalog of LSTID with periods ranging from 30 to 90 min, spanning 2021–2024. To validate our approach, we conducted case studies and statistical analyses, confirming accuracy and reliability. Additionally, testing in North America and Japan demonstrated the algorithm broader applicability. Our findings, obtained with user-defined parameters tailored to examine LSTIDs with periodicities in the 30–90 min range, show an average speed of 629 ± 215 m/s and period of 62 ± 14 min. A direct relation with auroral electrojet indices confirms that stronger auroral currents produce faster, higher amplitude and shorter period LSTIDs. Furthermore, our analysis reveals a link between LSTID occurrences and geomagnetic activity, reinforcing previous findings.