Imaging meso-scale ionospheric structures

The accuracy and capacity to resolve meso-scale structures of a four dimensional ionospheric imaging algorithm in the circumstance of data from dense networks of permanent GNSS ground receiver stations were investigated. Simulation studies were conducted in order to be able to assess the performance of the algorithm over the entire imaged region. The Multi-instrument Data Assimilation Software (MIDAS) algorithm was used for this purpose. Simulated input data in Receiver Independent Exchange Format (RINEX) were produced by calculating slant Total Electron Content (sTEC) values for satellite to receiver raypaths through an artificially generated ionosphere. Modeling these signals including Differential Code Biases (DCBs) and noise had negligible impact on the output from the imaging algorithm when compared with modeled signals that included neither. Comparing the output from MIDAS using a range of grid definitions show that finer grids have improved capacity to resolve meso-scale structures in the input model but over all are less accurate than coarser grids. The greatest errors occur in low-data regions of the grid and where structures in the input have the greatest gradients in vertical Total Electron Content (vTEC). A good compromise between the conflicting needs of resolution and accuracy is given by a grid defined with 2 degrees x 2 degrees latitude by longitude local horizontal grid divisions.