A new high-precision short-term ionospheric TEC prediction model based on the DBO-BiLSTM algorithm: A case study of Europe

In order to achieve high accuracy of ionospheric total electron content (TEC) short-term prediction for Europe, a hybrid novel deep learning model was established applying the dung beetle optimizer (DBO) algorithm to optimize the bidirectional long short-term memory (BiLSTM) neural network, named DBO-BiLSTM. For evaluating the TEC prediction accuracy of DBO-BiLSTM model, the TEC predicted by this model was compared with TEC computed using GPS observation released by the European Permanent Global Navigation Satellite System network (EPGNSS), and with those predicted by the sparrow search algorithm-based BiLSTM (SSA-BiLSTM), BiLSTM, and long short-term memory (LSTM) neural network models. The test results indicate that the predicted TEC by DBO-BiLSTM has the closest agreement with those solved by GPS data compared with those predicted by the other three models, and the prediction accuracy achieved by DBO-BiLSTM model is the highest with the root mean square error (RMSE) values of 1-h and 2-h predictions reaching 0.57 TECU and 0.92 TECU, respectively. What's more, the optimized hybrid DBO-BiLSTM model can effectively capture the ionospheric characteristics with the spatial-temperal changes, under quiet and moderate disturbed geomagnetic conditions, and during moderate solar activity period. This research provides a valuable hybrid DBO-BiLSTM model for high accuracy short-term prediction of ionospheric TEC for Europe, and gives an important reference for the further comprehensive TEC prediction under more sever disturbed geomagnetic conditions and more violent solar activity periods.