Tomographic imaging and characterization of ionospheric equatorial plasma irregularities with the Global Ultraviolet Imager

This dissertation develops a technique for reconstructing multidimensional images of the electron density of the Earth's ionosphere using spectroscopic measurements obtained from the Global Ultraviolet Imager (GUVI). This work combines a novel image processing approach, models of GUVI observations as tomographic reconstruction problems, and new studies of equatorial plasma irregularities and the global F-region ionosphere.; The image processing component of the research involves the development of a computationally efficient, edge-preserving regularization technique incorporating projection on convex sets (POCS) in order to solve the limited-angle tomographic image reconstruction problem. The algebraic reconstruction technique was specifically tailored to reconstruct the shapes of plasma irregularities while also recovering the altitude and longitude profile of the background ionosphere, all using data with a low signal-to-noise ratio.; GUVI observation geometry is discretely modeled and a linear algebraic relationship is derived between GUVI brightness measurements and ionospheric electron density values. The specific formulation of the problem varies to accommodate the physics of different latitude regions of the ionosphere in order to allow the three-dimensional observation geometry to be cast as a two-dimensional limited-angle tomography problem.; The experimental aspect explores the equatorial plasma bubble imaging capabilities of this technique and its scientific impact. Retrievals of altitude and longitude profiles allow for characterization of plasma bubbles based on their structure and depth of depletion. This information can then be used in coordinated studies of plasma bubbles with groundbased imaging systems. These coordinated studies both serve to validate the reconstructed images and to provide complementary information for a more complete understanding of plasma bubble events.