Fusion of radio occultation into computerized ionospheric tomography

Signals passing through the ionosphere are subjected to effects such as attenuation, scattering, polarization loss, time delay, Faraday rotation, and phase shift. The cause of the ionosphere's effect on signals is the frequency-dependent refractive index, with electron density being the most significant variable influencing the refractive index. Therefore, reconstructing the electron density of the ionosphere is crucial for examining the ionosphere's structure and its impact on signals passing through it. In this study, a 4-D Computerized Ionospheric Tomography (CIT) algorithm, IONOLAB-Fusion, is developed to reconstruct electron density using both real and virtual vertical and horizontal paths for both quiet and disturbed days of the ionosphere. The user-friendly algorithm only requires input of the coordinates of the region of interest and the desired spatio-temporal resolution interval. The model ionosphere is constructed using spherical voxels arranged in a lexicographical order to map the 4-D ionosphere into a 1-D vector. The model matrix is automatically generated using an optimized background ionosphere model either retrospectively or in near real-time. Singular Value Decomposition (SVD) is applied to estimate a subset of significant singular values and the corresponding signal subspace basis vectors. The measurement vector and sampling matrix is automatically created using optimized ground-based and satellite-based paths. Reconstruction is achieved in closed form using Least Squares estimation. When comparing IONOLAB-Fusion's performance over Europe to vertical electron density profiles obtained from ionosonde measurements, improvements of 26.51% and 32.33% are observed for quiet and disturbed days, respectively, relative to the background ionosphere model. When comparing Total Electron Content (TEC) maps calculated by IONOLAB-Fusion with GIM-TEC maps, IONOLAB-Fusion showed 37.89% and 31.58% better agreement for quiet and disturbed days, respectively, compared to the background ionosphere model. Additionally, IONOLAB-Fusion's results are compared with four other function-based methods in the literature, demonstrating performance improvements of 28.24% up to the height of hmF2 and 33.59% up to the Chapman height.