Abstract Knowing when a volcanic eruption occurs can be effective in reducing possible damages and life losses for residents close to the event. In this study, using GPS-TEC (Global Positioning System-Total Electron Content) satellite data, potentially volcanic ionospheric anomalies are investigated in a period of about two months around the time and location of the eruption of some volcanoes including Sundhnúkur (Iceland), Kilauea (USA), Mount Etna (Italy) and Laki Laki (Indonesia). In order to reduce the uncertainty in the detection of volcano-ionospheric anomalies, five classical and intelligent anomaly detection methods including median-interquartile, Kalman filter, ANN (Artificial Neural Network), LSTM (Long Short-Term Memory) and ACO (Ant Colony Optimization) were implemented in case of Iceland volcano. All five methods detected anomalies in the time interval of 9 days before the eruption of Iceland volcano. Also, the changes and spatial extent of TEC anomalies were also investigated on abnormal days, and a number of anomalies were observed in the studied volcanoes location. The observation of TEC anomalies in the magnetically conjugate region of the volcano was also discussed. Therefore, the results of this research show investigation of the temporal and spatial range of active volcanoes with GPS-TEC data using multi-predictor analysis can be effectively useful in detecting volcano-ionospheric anomalies.
Mehdi Akhoondzadeh (2026) studied this question.