Use of GPS radio occultation data for validation and improvement of ionospheric assimilation models

With the development of a new ionospheric assimilation model comes a need to evaluate the quality of that model's predictions. The availability of GPS radio occultation (RO) data allows us to accomplish such an assessment. Through an application of statistical methods, we conduct an analysis of observed and simulated values of slant total electron content (TEC). With the support of this comparison, we are able to generate a skill score parameterizing the goodness of fit between the two data sets. Furthermore, an analysis of Abel-derived electron density profiles (EDPs) shows that our skill score similarly transforms to a measure of the quality of the overall predicted ionospheric structure. Preliminary results indicate that a model which includes GPS RO data as part of its assimilation significantly out-performs a model which assimilates only ground-based data, or a model based solely on physics. More specifically, the GPS RO assimilation model predicts EDP structure and layer height with greater accuracy than other alternatives. These findings suggest there is a considerable benefit to the predictive value of an assimilation model associated with the introduction of space-based GPS RO data.