Prediction Of High Latitude Ionospheric Scintillation Based On Deep Learning

The rate of the total electron content(TEC)change index(ROTI)can be regarded as an effective indicator of ionospheric scintillation,in particular in low and high latitude regions.An accurate prediction of the ROTI is essential to reduce the impact of the ionospheric scintillation on earth observation systems,such as the global navigation satellite system.However,it is difficult to predict the ROTI with a high accuracy using conventional methods because of the sophisticated ionosphere.Therefore,it is necessary to further study the ionospheric scintillation in order to achieve highprecision prediction of ionospheric scintillation.In this study,the ROTI prediction of single station and multiple stations is carried out respectively based on the analysis of the temporal and spatial characteristics of ionospheric scintillation in the high latitude region of the northern hemisphere.A big data-driven method is used to model and predict the periodic ionospheric scintillation which is affected by solar and geomagnetic activities etc.The single station ROTI prediction results show that the accuracy of the deep learning method Long Short-Term Memory Network(LSTM)is more than 90% and the Total Skill Score(TSS)index is more than 0.17.The accuracy of predicting the scintillation in the next 5 minutes is even over 94%,and the TSS index is over 0.5.This LSTM model also has a high predictive ability during the period of high solar activity and geomagnetic storms.Deep learning method Artificial Neural Network(ANN)is used to model and predict the ROTI of multiple stations in the high latitude region of the northern hemisphere.The inner coincidence accuracy is 0.08TECU/min,the accuracy is above 94%,and the TSS index is above 0.5;The outer coincidence accuracy is 0.119TECU/min,the TSS index is above 0.36,and the accuracy is above 90%.This ANN model has a better performance in the spatial prediction of high latitude ionospheric scintillation with ROTI index.This study has important reference value and guiding significance for exploring the application of deep learning methods in solving the problem of ionospheric ROTI prediction.