GNSS TEC-Based Earthquake Ionospheric Perturbation Detection Using a Novel Deep Learning Framework

In this article, a new method for seismic ionospheric Global Navigation Satellite System (GNSS) total electron content (TEC) based anomaly detection using deep learning framework is proposed. The performance of the proposed encoder–decoder long short-term memory extended model compared with those five other benchmarking predictors. achieves best (R2 = 0.9105 and root-mean-square error (RMSE) 2.6759) in predicting TEC time series data, demonstrating 20% improvement R2 39.1% RMSE over autoregressive integrated moving average model. To detect pre-earthquake disturbances more accurately, reasonable strategy determining limits also Finally, applied to analyze disturbance 2016 Xinjiang Ms 6.2 earthquake. By excluding effects solar activity geomagnetic activity, obvious anomalies could be observed, occurring during 4–8 days prior to, on 1 day before, Negative tended occur earlier period, whereas positive occurred closer earthquake time, increasing intensity temporal proximity. Furthermore, confusion analysis used article verify reliability significant improvements GNSS shown advance technology.