| Forecasting Of Ionospheric Electron Content (TEC) Using a Nonlinear Regression Timeseries Network over the northern crest region of the equatorial anomaly in Egypt |
| Paper ID : 1074-ICRSSSA-FULL |
| Authors |
|
Wellen Rukundo *1, Shiokawa Kazuo2, Elsaid Ahmed3, Ola Ahmed Abuelezz4, Ayman Mahrous5 1Space Environment, Basic and Applied Sciences, Egypt Japan University of Science and Technology, Egypt 2Division for Ionospheric and Magnetospheric Research, Institute for Space-Earth Environmental Research, Nagoya University, Japan 3Faculty of Engineering, Egypt Japan University of Science and Technology 4Space Weather Monitoring Center, Helwan University, Cairo, Egypt 5Space Environment, Basic and Applied Sciences, Egypt Japan University of Science and Technology |
| Abstract |
| Egypt lies in the crest region of the northern equatorial anomaly which experiences high spatial and temporal ionospheric variations resulting from the equatorial irregularities. This poses a challenge to standard empirical ionospheric models like the International Reference Ionosphere (IRI) to accurately forecast and predict TEC. Several neural network techniques developed from diurnal, seasonal, solar and geomagnetic activity indices have been applied to improve the empirical TEC modeling in the Indian crest region. In this paper, we developed a nonlinear autoregressive (NARX) TEC model that accurately and consistently forecasted ionospheric TEC variation over the northern equatorial anomaly crest region of Egypt during the high solar activity year, 2014. The key factor for model development was the addition of estimated parameters representing equatorial electrojet (EEJ) and ExB drift velocity. The EEJ was estimated from the difference in the horizontal component of the magnetic field for two magnetometer stations in the low latitudes (off dip equator and dip equator) in the same longitudinal band, ExB drift velocity was estimated from the empirical model [1] while TEC was derived from a SCINDA GPS receiver installed at Helwan University, Cairo. The NARX TEC model forecasted diurnal TEC variation in 2014, with a root mean square error of 2.11 TECU, and was highly correlated (⁓0.99) with GPSTEC. The developed model also forecasted the solstices better than the equinox. However, the IRI-2016 model overestimated GPSTEC for both diurnal and seasonal variation. |
| Keywords |
| Equatorial ionization anomaly, Ionospheric TEC, ExB drift, Equatorial electrojet |
| Status: Accepted (Oral Presentation) |
