Tomographic Reconstruction of Ionospheric Electron Density Using Altitude-Dependent Regularization Strength over the Eastern Africa Longitude Sector

Abstract

A linear and iterative inversions with second order Tikhonov Regularization for 3-D tomography is employed to reconstruct ionospheric electron density (IED) during the solar minimum 2008 period. However, application of 3-D regularization is not straightforward due to spatial and temporal inhomogeneities in signal and noise along the ray paths that contribute to the observed ionospheric total electron contents (TECs). A regularization scheme is considered that takes into account the inhomogeneity, such that regions with weak signal contributions to TEC measurements are heavily penalized through a weighting factor incorporated with the discrete operator. The a priori IED from the NeQuick model and slant TECs at a chain of GPS stations over Eastern Africa are used in the analysis. The application of the new regularization scheme with iterative inversion on simulated TECs with actual observation geometry has recovered the true IED with lowest reconstruction error of 0.38% and 0.39% from measurement contaminated by 4% and 15% random noise, respectively. In contrast, the reconstruction with commonly used regularization factor has error of 0.79% and 0.85% from the same measurement. Moreover, various diagnostic tools applied to reconstructions from both synthetic and actual observations as well as comparison of vertical TEC (VTEC) from reconstruction and JPL VTEC have shown that the new altitude-dependent regularization performs well. This is reflected in high correlation of 0.87 to 0.91 between VTEC from reconstruction and JPL VTEC under iterative formulation. The tomographic reconstruction results under even bad a priori information show that the algorithm can recover up to 70% of the true IED over dense observations.