Modelling the solar, geomagnetic and periodic variations of the ionosphere over Ethiopia

A careful and continuous ionospheric modelling can significantly influence the performance of activities such as Positioning, Navigation and Timing services related with the Global Navigation Satellite System applications as well as the Earth Observations System, satellite communication and Space weather forecasting applications. In this paper, the linear time-series modelling that consists of the solar, geomagnetic and periodic components has been carried out on the daily ionospheric vTEC at two different Ethiopian GPS locations, at Arbaminch, ARMI (geographic 6.06ºN, 37.56ºE) and Bahir Dar, BDMT (geographic 11.60ºN, 37.38ºE), for the year 2012, 2014 and 2016 in the 24th solar cycle. The variations of vTEC due to the solar activities, geomagnetic activities and periodic oscillations have been explicitly investigated. The results confirmed that the correlation coefficient of the linear model based estimated vTEC and the observed GPS-vTEC is around 80% in the year 2014. Besides, solar activity is identified as the key component for the 27 days period variations of vTEC whereas geomagnetic activity is identified as the key component that influences the short-period variations of the daily average vTEC. In addition to the correlation analysis, the accuracy of the model has been assessed by comparing the International Reference Ionosphere (IRI 2016) model based vTEC and GPS-vTEC measurements as well as with the quadratic model based vTEC. Consequently, the linear model formulated with the solar, geomagnetic and periodic components significantly captured the variations (78-80%) of the observed vTEC compared with both the IRI 2016 and the quadratic models during the years 2012, 2014 and 2016. The comparison of the observed and predicted vTEC variations has also been examined using the continuous wavelet transform. The decomposed waves from the wavelet analysis have revealed that the predicted and observed vTEC have had simultaneous periods of variations specifically with the period of 27 days whereas the IRI 2016 could capture the short-period variations of vTEC. Moreover, the analysis from the transformed data in the year 2014 over both Arbaminch and Bahir Dar has indicated that the linear model based vTEC and the observed GPS-vTEC have had common pattern of variations with the period of 27 days that had lasted for 150 days (from day of the year 100 to 250).