Date of Award


Document Type

Thesis (Master's)

Department or Program

Engineering Sciences

First Advisor

Simon Shepherd


Specification of ionospheric convection over the entire high-latitude region using only measurements is desirable for space weather nowcasting but is not feasible. Instead, assimilative techniques are used to combine spatially sparse measurements with a complete background model. The Super Dual Auroral Radar Network (SuperDARN) Assimilative Mapping technique (SAM) is one such technique [Cousins et al., 2013a]. The SAM technique combines available line-of-sight velocity measurements from SuperDARN with a climatological high-latitude convection model producing a global solution of the electrostatic potential for a given period of interest (typically 1-2 minutes). The background model and velocity measurements are weighted according to each components' estimated error. While the background model error covariance matrix necessary for the SAM technique has been determined for the CS10 model [Cousins and Shepherd, 2010], a relatively recent climatological model derived using SuperDARN radar measurements, it has not been determined for the latest model, the TS18 model [Thomas and Shepherd, 2018]. The TS18 features several advancements in vector preprocessing and selection, and most importantly, included data from radars located at middle and polar latitudes that had not yet been built when earlier models (including the CS10) were constructed. To obtain the error covariance matrix for the TS18 model, the dominant modes of variability in the model are represented as a set of empirical orthogonal functions (EOFs) by fitting a set of basis functions to residuals between the model and a large set of SuperDARN observations. The procedure for obtaining these EOFs involves minimizing a non-linear cost equation to obtain multiple sets of coefficients representing both time and spatial variability. We compare the structure of the resulting EOFs to those obtained from the CS10 model and investigate differences in application output for specific time periods as well as average differences between this procedure, the previous version, and other assimilative techniques in general, and for different interplanetary magnetic field (IMF) conditions, dipole tilt angles, and universal time bins. The results presented herein indicate that the improvements in the TS18 model carry through into the SAM application, with higher potentials present at lower latitudes and higher cross-polar-cap potentials in certain conditions.