Ionospheric tomography: an algorithm enhancement

Abstract The Applied Research Laboratories, The University of Texas at Austin (ARL: UT), conducted a computerized ionospheric tomography (CIT) experiment called the “Mid-America CIT Experiment”, or MACE'93, from June through December, 1993. Portions of the data from this experiment have been input into various CIT algorithms to reconstruct a two-dimensional electron density image of the ionosphere. One of the CIT algorithms relies on model ionospheres to incorporate a priori information into basis functions that are used to represent the unknown ionospheric electron density structure. When global ionospheric models are used to generate the basis model ionospheres for this algorithm, the reconstructed ionospheres often underestimate the F2 peak density and some reconstructions display incorrect latitudinal electron density variation. These erroneous reconstructions are mainly caused by the global ionospheric models incorrectly modeling the ionosphere in terms of the F2 region thickness and latitudinal variations. This paper identifies an alternate set of basis ionospheres that improves the accuracy of the reconstructed ionospheres. In general, this enhancement has reduced the average per cent difference between actual and CIT reconstructed ionograms to 6% and the average absolute error between measured and CIT reconstructed f0F2 to 2.6%.