Initial prediction of Total Electron Content (TEC) At a low latitude station using neural network

This paper presents the prediction of hourly Total Electron Content (TEC) values of the ionosphere using neural network by utilizing the TEC data from a GPS Ionospheric Scintillation and TEC Monitor (GISTM) receiver. There are two network configurations; the first one by using data from January-November 2005 in training and December 2005 in testing, and the second one involved data from January-October and December 2005 in training and November 2005 in testing. Overall, the testing resulting RMSE of 3-7 TECU, absolute error of 2-5 TECU and relative error of 9-20%. The prediction of TEC on November 2005 which is during lower solar activity provides better results in comparison on December 2005 which is during a relatively higher solar activity.