Precursors Identification for Forecasting UHF‐Band Ionospheric Scintillation Events Over Chinese Low‐Latitude Region by Deep Learning. Issue 9 (8th September 2022)

• Record Type: Journal Article
• Authors: Zhang, Hongbo; Wang, Feifei; Sheng, Dongsheng; Ban, Panpan; Liu, Yumei
• Is Part Of: Earth and space science. Volume 9:Issue 9(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-09-08
• Subjects: ionospheric scintillation -- forecasting -- deep learning -- precursors; Space sciences -- Periodicals; Geophysics -- Periodicals; 500.5
• DOI: 10.1029/2021EA002164
• Languages: English
• ISSNs: 2333-5084

Abstract

Abstract: The occurrences of the UHF‐band ionospheric scintillation events caused by Equatorial Plasma Bubbles (EPBs) in the low‐latitude region depend on many factors. Nowadays, it is also a hard work for researchers to analyze and find out the useful precursory signatures of the generation of EPBs from several kinds of observations, which is a necessary work to build a forecasting model base on the priori knowledge. Many studies have revealed that there are distinctive features of daytime background ionosphere characteristics on scintillation and nonscintillation days. Deep learning technique (DLT) could automatically discover the distinctive variations from the daytime background ionospheric observations of scintillation and nonscintillation days. It is found that the forecasting problem of postsunset ionospheric scintillation events could be converted to a classification problem, and easily solved by DLT. The effectiveness of DLT for building a forecasting model of the UHF‐band ionospheric scintillation events over Chinese low‐latitude region is studied and evaluated. By analyzing the performance metrics of different combinations of related factors as the input parameters of DLT, a new important precursor for forecasting the UHF‐band ionospheric scintillation events over Chinese low‐latitude region is identified for the first time. It is suggested that the latitudinal and diurnal variations of the transequatorial Total Electron Content (TEC) profile in the east of the forecasting area is one of the key precursors. Key Points: The forecasting of postsunset UHF‐band ionospheric scintillation events is converted to a classification problem solved by deep learning Using deep learning to identify precursors of the UHF‐band ionospheric scintillation events over Chinese low‐latitude region Latitudinal and diurnal variation of transequatorial electron density profile is one of the key precursors for forecasting scintillations