ED?ConvLSTM: A Novel Global Ionospheric Total Electron Content Medium?Term Forecast Model

In this paper, we proposed an innovative encoder-decoder structure with a convolution long short-term memory (ED-ConvLSTM) network to forecast global total electron content (TEC) based on the International GNSS Service (IGS) TEC maps from 2005 2018 1-hr time cadence. The ED-ConvLSTM model is used 1–7 days in advance through iterations. To investigate model's performance, compared Reference Ionosphere (IRI2016) 2014 and 2018, 1-day Beijing University of Aeronautics Astronautics (BUAA) 2018. results show that our 7-day (ED-ConvLSTM forecasts 7 advance) outperforms IRI2016 5-day 5 BUAA model. Furthermore, root mean square error (RMSE) respect IGS decreases by 51.5% 43%, respectively, RMSE 20.3% lower than addition, has highest Equatorial Ionospheric Anomaly (EIA) region, but can roughly predict features locations EIA. However, fails localized enhancement sudden ionospheric response geomagnetic storms. Overall, shows competitive performance medium-term prediction during quiet periods.