Integrating New GNSS Methods into 3D Ionospheric Imaging: A Simulation Study

This presentation focuses on the incorporation of GNSS reflectometry (GNSS-R), signals obtained by low Earth orbiting (LEO) CubeSats, and low-elevation GNSS signals into 3D ionosphere imaging using simulated data. 3D images are useful in observing structures and abnormalities in the ionosphere to help us to better understand their evolution with space or time. The spatial and temporal resolution of tomographic images of the ionosphere is often limited due to the sparsity of ground-based receivers over oceans, polar regions, and restricted territories, as well as the low number and coverage of space-based observations such as GNSS radio occultation (GNSS-RO). Ground-based data alone cannot discern patterns with altitude, while radio occultation signals are quite sparse and provide low horizontal temporal resolution. Some imaging methods attempt to account for the lack of data by using interpolation of total electron content (TEC) data through the employment of basis functions or parameterized voxel-based techniques.
The goal of this project is to create an imaging method that will improve 3D imaging at high latitudes in both hemispheres using spaceborne GNSS-R and low-elevation satellite signals. We simulate ground-based, GNSS-RO, and GNSS-R signals based on real orbital parameters for GNSS constellations and Spire CubeSats (such as orbit inclination, altitude, and mean anomaly) with actual high-latitude ground receiver locations. The specular points of reflected signals are calculated based on satellite positions and reflection surface and determined whether it is a coherent (usable) signal using a statistical mask. Ionospheric image reconstruction is achieved using a voxel-based inversion technique. We simulate the 3D ionosphere using the International Reference Ionosphere (IRI) model as a base. A Voronoi forward model is used to construct a matrix of ray path intersection coordinates used in a simultaneous multiplicative algebraic reconstruction technique (SMART) of a base ionospheric image with given data. In a previous pilot study, this method was used to reconstruct a simulated 3D image of the ionosphere electron density using ground- and space-based radio occultation satellite ray paths. The method has since been updated to account for the geometry of reflected signals. TEC values derived from Precise Orbit Determination (POD) and low-elevation (between 5o and 20o) satellite signals, which have recently been shown to provide precise TEC measurements, have been included as well to maximize the number of ray paths. The images resulting from our newly developed method are then analyzed for performance quality to evaluate how the number of voxel intersections for different satellite geometries improve the image. We quantify reconstruction error as the difference between the base model and reconstructed image to assess how error improves with new signal types.
Inversion method generated image accuracy depends on observation geometry and data density distributions. We show that incorporating GNSS-R and POD measurements increases the number of signal paths and coverage at high-latitudes, improving the resolution and accuracy of the resulting tomographic images. With this, we can observe the relative impact of GNSS-R vs GNSS-RO and ground-based on image quality, as well as explore other geometries resulting from changes in LEO constellation size or orientation. Results so far are promising for the future addition of real GNSS-R data to a tomographic program as ours can successfully reconstruct images using simulated TEC. By testing the new inversion equation first with accurately simulated signals, we demonstrate the ability to easily integrate GNSS-R geometry with ground-based, POD, and GNSS-RO data into the algorithm.