On the possible contribution of ionospheric vertical drifts to TEC modelling in low latitudes

Abstract

A single station empirical total electron content (TEC) model based on the Global Navigation Satellite System (GNSS) data during 2001–2015 is presented over the African low latitude region. For the first time, we have investigated the contribution of ionospheric vertical drifts to TEC modelling by including vertical E_B drift generated from Communication/Navigation Outage and Forecasting System (C/NOFS) satellite data as an input along with diurnal variation, seasonal variation, solar activity and geomagnetic activity representations. The inputs were used to develop the neural network (NN) TEC model over Mbarara, MBAR (0:6_S; 30:74_E; 10:22_S geomagnetic) which was later validated on independent dataset that was selected from different solar activity periods. The model without vertical E_B drift as an input gave mean absolute error (MAE) of 4.013 TECU. Inclusion of the vertical E_B drift input reduced the MAE to 3.757 TECU equivalent to an average improvement of 6.4% in TEC modelling. The maximum improvement attained was around 13% during the high solar activity period. The correlation coefficient (R) values for models without and with vertical E_B drift input were 0.954 and 0.958 respectively. Our results show that availability of ionospheric vertical drift data has the potential of improving future TEC modelling in low latitude regions. _ 2020 COSPAR. Published by Elsevier Ltd. All rights reserved.