Characterization and prediction of the GPS derived total electron content over the Eastern part of the African sector.

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
University of Dar es Salaam
The purpose of this study was to investigate the behavior of the ionosphere over the eastern part of the African sector and develop a model for TECv prediction over the region. Data used were from the Africa array and IGS network of ground based dual-frequency GPS receivers from eighteen (18) stations within the eastern part of the African sector for 2012 and 2013, and for NN modeling, data for 2014 to 2016 were also included in the study. The network array covers approximately the area from 17°S to 12°N in geographical latitude (magnetic lat. ~2.6oN to ~26.9oS) and from 26°E to 40°E in geographical longitude (magnetic long. ~95oE to ~112oE). Stations at equatorial region were used for comparison purposes. The results showed that, the variations of TECv over the region were complex and irregular especially during afternoon hours and varied from day to day without remarkable variation pattern, and were more evident during equinox months than during solstice months. The post noon maxima vary differently from one season to another and the period of time occupied by the day maximum is larger with increasing latitude. TECv varied semiannually, and the day maximum value increases from the stations in low latitudes and decreases significantly with the increasing latitudes. Generally, the IRI-2012 model was observed to overestimate the GPS-TECv, and IRI-2001 overestimated GPS-TECv more compared with other options. GPS-TECv values and IRI-TECv values showed good correlation (above 0.8) at most of time and the TECv using IRI-neq and IRI-01- corr had small PRMSE from the GPS measured TECv compared to the IRI-2001. It was further observed that, TECv was enhanced on the storm day reaching its maximum value at around the mid-day, and strong reductions in TECv were also depicted from the stations that are located far from the equatorial region from around 13 – 20 hours (UT). For the developed model based on NN, it was established that, when the NmF2 is added as an input neuron to the model, the optimal results was obtained. This was archived when the hidden layer of 18 neurons was used. The developed model predicted the GPS TECv more accurate compared to IRI-2012 model with NeQuick topside Ne option.
Available in print form, East Africana Collection, Dr. Wilbert Chagula Library, Class mark (THS EAF QP881.2.I6.A354D328)
Ionospheres, Research, Eastern part of Africa, Geomagnetism, Diural variation
Daudi, E. (2018). Characterization and prediction of the GPS derived total electron content over the Eastern part of the African sector. Doctoral dissertation, University of Dar es Salaam.