Machine Learning Prediction of Highlatitude Ionospheric Irregularities from GNSS-derived ROTI Maps

Lei Liu, Y. Jade Morton, Yunxiang Liu

Abstract: This paper applies a convolutional long short-term memory (convLSTM) model with a custom- designed loss function Lc (convLSTM-Lc) to forecast high-latitude ionospheric irregularities. The data used for this study are GNSS-derived rate of TEC change index (ROTI) maps collected in 2014-2016. The convLSTM-Lc is trained with real ROTI data above 51┬░magnetic latitude (MLat) during January 1, 2014 to October 21, 2015. Test results show that the convLSTM-Lc model can predict large-scale features and overall evolution of the high-latitude ionospheric irregularities occurrence with a lead time of up to one day.
Published in: Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021)
September 20 - 24, 2021
Union Station Hotel
St. Louis, Missouri
Pages: 3870 - 3877
Cite this article: Liu, Lei, Morton, Y. Jade, Liu, Yunxiang, "Machine Learning Prediction of Highlatitude Ionospheric Irregularities from GNSS-derived ROTI Maps," Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021), St. Louis, Missouri, September 2021, pp. 3870-3877.
https://doi.org/10.33012/2021.18046
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In