Single Parameter-Based Neural Networks Prediction of TEC

Abstract:

Article Preview

Total electron content (TEC) is a highly non-linear ionospheric parameter that is strongly affected by changes in the solar activity. Reliability of space-based systems is affected by TEC and hence requires the development of a good model to nowcast and forecast its variability. Therefore, the aim of this paper is to present a Neural Networks (NN) model to predict the TEC variability. Slant TEC (sTEC) from the GPS Ionospheric Scintillation and TEC Monitor (GISTM) receiver is calculated and converted to vertical TEC (vTEC) based on the Single Layer Model (SLM). As a preliminary study, only one variable is considered as an input parameter for the NN, which is the hourly vTEC values and this is used to train the network with feed forward back propagation (BP) algorithm. The data set is the observed TEC for April and May 2005. Absolute maximum error and root mean square error (RMSE) are then calculated for each trained NN to evaluate the performance of the model. Analysis of the results reveals that, on average, the NN model is capable to predict TEC values but however extended input data set is needed to improve the generalization capability and reliability of the model. The highest absolute maximum error is 17 TECU and the smallest absolute maximum error is 6 TECU. The RMSE values for May 27 to May 30 are 1 TECU, 4 TECU, 5 TECU and 1 TECU, respectively.

Info:

Periodical:

Main Theme:

Edited by:

R. Varatharajoo, E. J. Abdullah, D. L. Majid, F. I. Romli, A. S. Mohd Rafie and K. A. Ahmad

Pages:

481-485

Citation:

V. Jayapal and M. Z. Ahmad Faizal, "Single Parameter-Based Neural Networks Prediction of TEC", Applied Mechanics and Materials, Vol. 225, pp. 481-485, 2012

Online since:

November 2012

Export:

Price:

$38.00

[1] Rama Rao, P. V. S., Gopi Krishna. S, Vara Prasad, J., Prasad, S.N.V.S. , Niranjan, K., and Prasad, D. S. V. V. D. Geomagnetic Storm effects on GPS based navigation. Ann. Geophys., 27 (2009) 2101-2110.

DOI: https://doi.org/10.5194/angeo-27-2101-2009

[2] Kouris, S.S., Xenos, T.D., Polimeris, K.V., and Stergiou, D. TEC and foF2 variations: preliminary investigations. Ann. Geophys. 47 (4) (2004).

[3] Tulunay, E., Senalp, E.T., Radicella, and S.M., Tulanay, Y. Forecasting total electron content maps by neural network technique. Radio Sci. 41, doi: 10. 1029/2005RS003285 (2006).

DOI: https://doi.org/10.1029/2005rs003285

[4] John A. Klobuchar. Ionospheric Time-Delay Algorithm for Single Frequency GPS Users. IEEE Transactions on Aerospace And Electronic Systems Vol. Aes-23 May (1987), No. 3.

DOI: https://doi.org/10.1109/taes.1987.310829

[5] J.B. Habarulema, L. -A. McKinnell, B.D.L. Opperman. A recurrent neural network approach to quantitatively studying solar wind effects on TEC derived from GPS; preliminary results. Ann Geophys., 27 (2009) 2111–2125.

DOI: https://doi.org/10.5194/angeo-27-2111-2009

[6] Williscroft L.A. and Poole A.W. Neural Networks, foF2, sunspot number and magnetic activity. Geophysical Research Letters, 23(24) (1996), p.3659–3662.

DOI: https://doi.org/10.1029/96gl03472

[7] Leandro R.F. and Santos M.C. A neural network approach for regional vertical total electron content modeling. Studia Geophysica et Geodaetica, 51(2) (2007), p.279–292.

DOI: https://doi.org/10.1007/s11200-007-0015-6

[8] Hernandez-Pajares M., Juan J. and Sanz J. Neural network modelling of the ionospheric electron content at global scale using GPS. Radio Science, 32 (1997), p.1081–1090.

DOI: https://doi.org/10.1029/97rs00431

[9] Satish Kumar. Neural Network, A Classroom Approach, Twelfth ed., Tata McGraw- Hill Education Private Limited, New Delhi, (2011).