Parameter estimation for synthetic TEC surfaces by using Particle Swarm Optimization

In this study, parameter estimation is made for global ionospheric Total Electron Content (TEC) on both noiseless and noisy synthetic surfaces by using modified Particle Swarm Optimization (PSO). In addition, the improvements made in the PSO algorithm to obtain better results are presented. Trend functions that best regionally and globally represent the quiet and distorted ionosphere are given. For noisy trend surfaces, additive white Gaussian noise is added on trend surfaces according to latitude. International GPS System stations (IGS) are used for regional sampling whereas TNPGN-Active stations are used for both regional and global sampling. A brief discussion of PSO and its improvements for modified PSO is provided. Performance and error criterias are determined for the results of noisy and noiseless dual-core Gaussian trend surfaces.